Complete Solubilization and Purification of Recombinant

Human Growth Hormone Produced in Escherichia coli

Min-Ji Kim1,2., Hyun Soo Park1., Kyung Hye Seo1, Hyo-Jin Yang1,2, Sook-Kyung Kim1,2, Jun-Hyuk Choi1*
1 Center for Bioanalysis, Department of Metrology for Quality of Life, Korea Research Institute of Standards and Science, 1 Doryong-dong, Youseong-gu, Daejeon, South
Korea, 2 Department of Bio-Analytical Science, University of Science & Technology, Daejion, South Korea

Abstract
High-level expression of recombinant human growth hormone (hGH) in Escherichia coli (E. coli) leads to the formation of
insoluble aggregates as inclusion bodies devoid of biological activity. Until recently, significant efforts have been made to
improve the recovery of active hGH from inclusion bodies. Here, we developed an efficient procedure for the production of
completely soluble hGH by minimizing the formation of inclusion bodies and optimizing protein purification conditions.
Under the newly established conditions we were able to obtain most of the total hGH in the soluble fraction. We show that
the soluble protein can be efficiently purified in high yield by a series of chromatographic procedures. We analyzed the
resulting hGH using various analytical techniques such as reversed-phase high-performance liquid chromatography (RP-
HPLC), size-exclusion chromatography (SEC), matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass
spectrometry, and circular dichroism (CD). These multiple analyses support the conclusion that we obtained highly pure
hGH with the expected molecular mass and intact secondary structure. The biological activity of purified hGH was also
confirmed by evaluating its growth-promoting effect using a cell proliferation assay. Taken together, we describe a
straightforward strategy for the production of completely soluble and biologically active hGH in E. coli.

Citation: Kim M-J, Park HS, Seo KH, Yang H-J, Kim S-K, et al. (2013) Complete Solubilization and Purification of Recombinant Human Growth Hormone Produced
in Escherichia coli. PLoS ONE 8(2): e56168. doi:10.1371/journal.pone.0056168
Editor: Jason Mulvenna, Queensland Institute of Medical Research, Ustralia
Received August 3, 2012; Accepted January 7, 2013; Published February 7, 2013
Copyright: ß 2013 Kim et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the Korea Research Institute of Standards and Science under the project "Development of Protein Measurement
Standards", Grant 12011021. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: [email protected]
. These authors contributed equally to this work.

Introduction developed in order to prevent the formation of inclusion bodies

relies on the secretion of hGH into the bacterial periplasmic space
Human growth hormone (hGH) is a single-chain polypeptide using a signal peptide [11,12,13]. The yields associated with
containing 191 amino acid residues and is synthesized in the periplasmic expression using different expression constructs have
pituitary gland. It is one of the most important hormones in the been extensively investigated under various conditions [14].
human body due to its pivotal role in a variety of biological In the present study, we describe a much simpler approach to
functions, including cell proliferation and metabolism [1]. maximize the solubility of hGH expressed in bacterial cytoplasm
Endogenous hGH is a non-glycosylated protein [2], and conse- by optimizing the conditions for protein expression, extraction and
quently recombinant forms of hGH have been extensively purification.
produced in prokaryotic expression systems [3,4]. However,
high-level expression of hGH in E. coli results in insoluble protein
Materials and Methods
aggregates as inclusion bodies that are often observed with many
other eukaryotic proteins [3,4,5]. Therefore, the aggregated hGH Gene cloning of hGH
requires solubilization and refolding steps prior to purification by The hGH gene (GenBank Accession No. K02382.1) construct
chromatography, and the overall protein recovery is significantly was amplified from human cDNA by the polymerase chain
affected by the efficiency of these pre-purification steps. In general, reaction (PCR) using forward primer 5̀-gcggctagcatgttcccaac-
the aggregated proteins need to be solubilized using high cattcccttatcc-3̀ and reverse primer 5̀-gcgctcgagctagaagcca-
concentrations of denaturants such as urea or guanidine hydro- cagctgccctc-3̀ (NheI and XhoI restriction enzyme sites underlined,
chloride (GnHCl), followed by removal of denaturants for protein respectively). The hGH gene was cloned into pET-28b (Novagen,
refolding. In many cases, however, the overall yield of biologically Madison, WI, USA) expression plasmid to produce recombinant
active protein from inclusion bodies is very low, and these hGH with a hexahistidine tag and a thrombin cleavage site at the
purification procedures are often costly and time–consuming [6]. N-terminus (His-hGH). For untagged hGH, the gene construct
Therefore, considerable efforts have been made to increase the was amplified by PCR with forward primer 5̀-gcgccatggcgatgttcc-
efficiency of solubilization and refolding as a means of improving caaccattcccttat-3̀ and reverse primer 5̀-gcgctcgagctagaagcca-
the overall recovery of biologically active hGH from inclusion cagctgccctc-3̀ (NcoI and XhoI restriction enzyme sites underlined,
bodies [7,8,9,10]. An alternative approach that has been respectively). The PCR product was digested with NcoI and XhoI,

PLOS ONE | www.plosone.org 1 February 2013 | Volume 8 | Issue 2 | e56168

 Solubilization and Purification of Recombinant hGH

and ligated into NcoI and XhoI restriction sites of pET-28a of 0 – 1 M NaCl gradient. Fractions containing hGH were pooled
(Novagen) expression vector to produce recombinant hGH and dialyzed again. Dialyzed fractions were purified by anion-
without a hexahistidine tag at the N-terminus (untagged hGH). exchange and gel-filtration chromatography under the same
The nucleotide sequences of inserts were verified by automatic conditions as described above. Purified proteins were dialyzed
sequencing. against storage buffer (10 mM Na2HPO4, pH 7.4, 0.5% glycine,
2.25% mannitol). Note that all purification procedures were
Solubility assay performed at 4uC. Protein concentration was determined by the
Plasmids pET28-His-hGH and pET28-hGH were transformed Bradford and bicinchoninic acid (BCA) assays using BSA as a
into E. coli BL21 (DE3) cells for protein expression. A 10 ml standard [15]. Protein purity was initially monitored by SDS-
aliquot of an overnight culture was seeded into 500 ml of fresh LB PAGE and silver staining, and later by more advanced methods as
(Luria-Bertani) medium (10 g Bacto tryptone, 5 g yeast extract specified below.
and 10 g NaCl per liter of solution) containing 50 mg/ml
kanamycin and grown to an OD600 of about 0.6 at 37uC. Characterization of purified proteins
Recombinant protein expression was induced with 1 mM Protein purity was examined by RP-HPLC using a Kinetex C18
isopropyl-b-D-thiogalactopyranoside (IPTG), and cells were fur- column (2.6 mm, 15062.10 mm; Phenomenex, Torrance, CA,
ther grown at different temperatures where indicated. The cells USA) [16]. Analytes were eluted using a linear gradient from 28%
were then harvested by centrifugation at 6,000 g for 20 min, and in buffer A (0.1% trifluoroacetic acid in H2O) to 100% in buffer B
cell pellets were suspended in lysis buffer (50 mM Tris-HCl, (0.1% trifluoroacetic acid in acetonitrile) at 40uC. The flow rate
pH 8.0, 0.5 mM EDTA, 1 mg/ml lysozyme) containing 16 was 0.2 ml/min and detection was achieved by monitoring the
protease inhibitor cocktail (Roche, Barcelona, Spain). During UV absorbance at 220 nm.
protein extraction, solubility assays were performed using different Analytical size exclusion chromatography was performed on an
buffer additives and buffer volumes (up to 2 ml for 30 mg cell AKTA FPLC system using a Superdex 75 10/300 GL column
pellets). Following incubation on ice for 30 min, cells were (GE Healthcare). The purified hGH was injected into the column
sonicated at 200 W (10 times for 10 s) using an Uibra cell TM using a 100 ml sample loop with an injection volume of 100 ml.
sonicator (Sonics & Materials, Newtown, CT, USA) and then Protein was eluted isocratically with 50 mM Tris-HCl (pH 8.0),
centrifuged at 10,000 g for 20 min. The soluble (supernatant) and 150 mM NaCl and 10% glycerol at a flow rate of 0.8 ml/min.
insoluble (pellet) fractions were analyzed by SDS-PAGE and The reliability of the analysis was confirmed by at least two
Coomassie blue staining. The level of protein was quantified using independent measurements under the same conditions. Detection
ImageQuantTM TL 5.2 analysis software (GE Healthcare, Piscat-
was achieved by monitoring the UV absorbance at 280 nm.
away, NJ, USA) after scanning gels on a V799 scanner (EPSON,
MALDI-TOF mass spectrometry was performed on an Autoflex
Long Beach, CA, USA). Averages from at least three independent
III Smartbeam device (Bruker Daltonics, Billerica, MA, USA)
experiments were graphed and presented as mean 6 standard
[16]. Sample was mixed with the same volume of MALDI matrix
deviation, except for induction at 37uC.
(10 mg/mL of a-cyano-4-hydroxycinnamic acid) and spotted on a
MALDI target plate. External calibration was performed with a
Purification of recombinant hGH Peptide and Protein MALDI-MS Calibration Kit (Sigma, St.
Through extensive solubility assays, the protein expression and Louis, MO, USA). Mass spectra in the m/z range of 15000–45000
extraction conditions were optimized as follows. E. coli BL21 (DE3) were acquired in the positive ion mode.
cells expressing recombinant hGH (untagged hGH and His-hGH) Circular dichroism (CD) measurements were performed at
were grown in a 1 L baffled flask containing 250 ml of culture
25uC on a J-815 circular dichroism spectropolarimeter (Jasco,
medium and induced at 16uC for 16 h. Harvested cells were
Tokyo, Japan) using a quartz cuvette with path length 0.1 mm to
suspended in 25 ml of lysis buffer (50 mM Tris-HCl, pH 8.0,
measure the far-UV range [17]. Spectra were recorded over a
0.5 mM EDTA, 0.1% Triton X-100, 1 mg/ml lysozyme, 16
wavelength range of 200–250 nm with bandwidth 0.1 nm,
protease inhibitor cocktail), disrupted by sonication, and then
scanning speed 50 nm/min, and 10-s response time. Control
centrifuged at 10,000 g for 20 min. For purification of His-hGH,
hGH was purchased from LG Life Science (Daejeon, South
the supernatant was applied onto a 5 ml Ni-NTA agarose column
Korea) and analyzed as a control under identical conditions [18].
(Qiagen, Valencia, CA, USA). The column was washed with 3
column volumes of washing buffer (16 PBS, pH 7.4, containing
150 mM NaCl, 10 mM imidazole, 0.1% Triton X-100, 10% Nb2-11 cell culture
glycerol) and eluted with elution buffer (16 PBS, pH 7.4, The PRL-dependent rat T-lymphoma cell line, Nb2-11 cells
containing 150 mM NaCl, 200 mM imidazole, 0.1% Triton X- (Sigma) were cultured in RPMI 1640 medium (Gibco/Invitrogen,
100, 10% glycerol). Fractions containing His-hGH were pooled Grand Island, NY, USA) supplemented with 10% fetal bovine
and dialyzed against buffer containing 50 mM Tris-HCl (pH 8.0) serum (FBS) (Gibco/Invitrogen), 10% horse serum (HS) (Gibco/
and 10% glycerol. Dialyzed fractions were further purified by Invitrogen) and 1% penicillin-streptomycin (Gibco/Invitrogen) at
anion-exchange chromatography on a 1 ml Mono Q 5/50 GL 37uC in a humidified atmosphere containing 5% CO2 [19]. Cell
column (GE Healthcare), and eluted with a 10-column volume proliferation was determined using MTS assays [20]. Briefly, cells
linear gradient of 0–0.5 M NaCl. Fractions forming the major were harvested, rinsed in culture medium without FBS, and then
peak containing His-hGH were pooled and finally purified by gel- incubated for 48 h in 96-well plates at 20,000 cells/ml (100 ml/
filtration using a HiLoad 26/30 Superdex 200 column (GE well) in the presence of various amounts of purified hGH.
Healthcare) equilibrated with buffer containing 50 mM Tris-HCl Following incubation, 20 ml of the MTS reagent (Promega, WI,
(pH 8.0), 150 mM NaCl and 10% glycerol. For purification of USA) was added to each well, and cells were incubated for 2 h.
untagged hGH, the supernatant was dialyzed against binding The absorbance was recorded on a microplate reader (Bio-Rad,
buffer containing 50 mM Tris-HCl (pH 8.0) and 10% glycerol. CA, USA) at a wavelength of 490 nm. Cell numbers were
Dialyzed lysates were loaded onto a 20 ml Hiprep DEAE FF 16/ determined using a standard curve plotted from a linear
10 column (GE Healthcare) and eluted with a 10-column volume relationship between cell number and absorbance.

PLOS ONE | www.plosone.org 2 February 2013 | Volume 8 | Issue 2 | e56168

 Solubilization and Purification of Recombinant hGH

Results As a substantial amount of hGH was still found in the insoluble

fraction, we examined whether the pelleted hGH could be
Solubilization of recombinant hGH under various recovered by optimizing protein extraction. Although the forma-
conditions tion of inclusion bodies may be further reduced by optimizing
In agreement with previous studies, we confirmed significant environmental factors such as culture media [21] or genetic
accumulation of insoluble protein aggregates as inclusion bodies engineering [23,24], we reasoned that the observed protein
upon high-level expression of hGH in E. coli (Figure 1). We aggregation may involve some other factors such as non-specific
observed that most of the hGH was found in the insoluble fraction protein interactions. To reduce such interactions and facilitate
(lane P; pellet fraction), and there was a very small amount of protein solubility, a number of useful buffer additives have been
soluble hGH (lane S; soluble fraction). To increase the yield of identified and efficiently implemented [25]. We first tested the
soluble protein, the protein expression conditions can be optimized effects of mild non-ionic detergents including Triton X-100 and
by varying induction temperatures, inducer concentrations and Tween 20 since they are unlikely to disrupt protein structure. As
media compositions [21]. In particular, protein expression at low shown in Figure 2A, the use of Triton X-100 significantly
temperature often significantly improves the solubility of recom- increased the solubility of hGH expressed at low temperature. This
binant proteins [22]. Indeed, we observed that induction at lower non-ionic detergent was very efficient at concentrations as low as
temperatures (16–20uC) significantly increased the amount of 0.1%. The addition of Tween 20 also increased the solubility, but
soluble hGH without modifying our protein extraction method appeared less efficient than Triton X-100 (Figure 2B). Next, we
(Figure 1A), indicating that a slow induction of hGH efficiently determined the effect of salts on hGH solubility as the solubility of
reduced the formation of inclusion bodies. We observed efficient protein is often dependent on the salt concentration. Interestingly,
expression of hGH with a 5-fold increase in solubility upon the results show that the addition of NaCl stimulated protein
induction at 16uC for 16 h (Figure 1B), and thus those conditions aggregation to some extent (Figure 2C). The observed effect was
were used in subsequent experiments. We also determined not specific to NaCl, as even more dramatic effect was observed
whether the formation of inclusion bodies could be reduced by with KCl (Figure 2D). Although 150 mM salts appeared to have
lowering IPTG concentration (to 0.1 mM). The results indicate marginal effects (especially NaCl, the more commonly used salt),
that reducing IPTG concentration is not as effective as lowering the solubility of hGH seems to be sensitive to such additives at
induction temperature at preventing the formation of inclusion higher concentrations. Since protein concentration is another
bodies under our conditions (Figure S1). important factor in protein aggregation, we assessed protein

Figure 1. Solubility comparison of recombinant hGH expressed at different temperatures. (A) The levels of insoluble versus soluble hGH
upon induction at various temperatures. E. coli BL21 cells expressing recombinant hGH were grown to an OD600 of 0.6 at 37uC and then induced at
the indicated temperatures. Cells were lysed, and the pelleted (P) and soluble (S) fractions were analyzed by SDS-PAGE and Coomassie blue staining
(upper panel). The levels of soluble hGH were quantified and graphed by a densitometry assay using ImageQuantTM TL 5.2 analysis software (bottom).
The values are plotted as mean 6 standard error. (B) Increased solubility of hGH expressed at reduced temperature. Protein samples obtained with
induction at either 37uC (upper) or 16uC (bottom) were separated on a 4–12% SDS-PAGE and visualized by Coomassie blue staining. U, uninduced; I,
induced; L, lysis; P, pelleted; S, soluble; asterisk; lysozyme.
doi:10.1371/journal.pone.0056168.g001

PLOS ONE | www.plosone.org 3 February 2013 | Volume 8 | Issue 2 | e56168

 Solubilization and Purification of Recombinant hGH

solubility using different buffer volumes. As expected, the risk of To ascertain that the established procedure is not specific to the
protein aggregation was significantly decreased by appropriate tagged version of hGH that harbors some additional amino acid
dilution (Figure 2E). Although reducing agents often prevent residues including the hexahistidine sequence and a protease
protein aggregation by inhibiting the formation of non-native cleavage site, we produced an untagged form of hGH and
intermolecular disulfide bolds, we observed no significant effect of confirmed the high level of protein solubility under identical
b-mercaptoethanol (Figure 2F). conditions (Figure 3C). While optimizing purification steps for
It should be noted that we observed essentially no effect of the untagged hGH, we found that the consecutive application of a
buffer additives on the solubility of hGH expressed at higher weak followed by a strong anion exchange resins (DEAE and
temperature (37uC), indicating that irreversible formation of MonoQ respectively) and size-exclusion chromatography pro-
inclusion bodies resulted in insoluble pellets (Figure 2A–F). duced very high yields of pure hGH (Figure 3D). The results for
the purification are summarized in Table S2.
Purification of soluble hGH Taken together, these results indicate that we established an
Based on the results described above, we optimized the efficient purification procedure for soluble hGH in E. coli using
conditions for protein expression and the buffer solution for optimized conditions in which protein aggregation was minimized.
protein extraction. The protein was induced at 16uC for 16 h and
extracted using the optimized buffer solution (50 mM Tris-HCl, Characterization of purified hGH
pH 8.0, 0.5 mM EDTA, 0.1 % Triton X-100, 1 mg/ml lysozyme, Upon purification, the purified proteins were analyzed by
16protease inhibitor cocktail) and buffer volume (1 ml for 30 mg several analytical methods, including RP-HPLC, size-exclusion
of pellets). Under these conditions, we observed greatly improved chromatography, MALDI-TOF mass spectrometry, and circular
solubility of recombinant hGH (Figure 3). Significantly, the yield dichroism (CD). RP-HPLC analyses of the two purified proteins,
of soluble hGH was estimated to be up to 95% of the total hGH His-hGH and untagged hGH, showed that they were eluted as a
expressed in E. coli. single peak near 8 min and there was no detectable contaminant
As we obtained highly soluble hGH, the protein was (Figure 4A). The purity of His-hGH and untagged hGH was
subsequently purified by Ni-NTA affinity, anionic exchange, and estimated to be 97.6% and 98.7%, respectively. The minor peaks
size-exclusion chromatography (Figure 3A). SDS-PAGE analysis near 2.5 min are thought to have been caused by the buffer
shows a clear, single band with a molecular mass of about 25 kDa solution. Next, we performed analytical size-exclusion chroma-
when eluted from the final column (Figure 3B). The results for the tography (SEC) to determine the molecular mass of the purified
purification indicate that we achieved efficient purification of hGH proteins (Figure 4B). The results show that their molecular weights
with high yield (.40%) and high purity (.97%) under our were approximately 21.3 kDa (His-hGH, expected: 24.612 kDa)
conditions (Table S1). and 20.3 kDa (untagged hGH, expected: 22.246 kDa). Once
again, we observed only one major peak for both proteins with no

Figure 2. Solubilization of hGH under various extraction conditions. The solubility of hGH, expressed either at 37uC or 16uC, was determined
by evaluating the effects of Triton X-100 (A), Tween 20 (B), NaCl (C), KCl (D), buffer volumes (E), and b-mercaptoethanol (F). The levels of soluble hGH
were quantified and graphed as described in Figure 1.
doi:10.1371/journal.pone.0056168.g002

PLOS ONE | www.plosone.org 4 February 2013 | Volume 8 | Issue 2 | e56168

 Solubilization and Purification of Recombinant hGH

Figure 3. Purification of recombinant hGH under optimized conditions. (A) Scheme of His-hGH purification from E. coli. The arrow indicates
the overall yield of hGH at each step in the purification process. (B) Purification of His-hGH. Elutions from each column were separated on a 4–12%
SDS-polyacrylamide gel and analyzed by Coomassie blue staining. (C) Scheme of His-hGH purification from E. coli. (D) Purification of untagged hGH.
Elutions were analyzed as in (B).
doi:10.1371/journal.pone.0056168.g003

other significant protein peaks, indicating their high purity. As the Biological activity of purified hGH
molecular weights determined by SEC appeared slightly lower To determine whether the purified hGH proteins were
than the expected values, we wanted to determine their molecular biologically active, we performed a rat Nb2-11 lymphoma cell
masses more accurately. To this end, we analyzed the proteins proliferation assay [26]. Cells in starvation media were treated
using MALDI-TOF mass spectrometry. As shown in Figure 4C, with different concentrations of purified His-hGH or untagged
the measured molecular masses of His-hGH and hGH were hGH, and then their growth-promoting activity was evaluated. As
24,565 Da and 22,262 Da, respectively, and these values are very shown in Figure 6, both proteins greatly stimulated the prolifer-
close to the expected masses. Finally, to determine the structural ation of Nb2-11 cells in a dose-dependent manner in concentra-
integrity of the purified proteins, structural characterization was tions ranging from 0.4 to 10 ng/ml. Moreover, the stimulation
performed by circular dichroism (CD) analysis. As shown in Figure levels of both proteins were comparable to that of control hGH. In
5, we found that the two CD spectra of the purified proteins were contrast, no growth stimulation was observed with BSA, which
very similar to that of the control hGH. In agreement with was used as a negative control. Therefore, the hGH proteins
previous reports [11,18], we clearly detected a positive peak at purified under our established conditions retained their biological
195 nm and a negative peak with a minimum at 208 nm and a activity.
shoulder at 222 nm, which are typical characteristics of a structure
with high a-helical content. The results suggest that our optimized Discussion
procedure for protein expression, extraction and purification did
not disrupt the structural integrity of the proteins. E. coli has been efficiently used for the high-yield production of
many recombinant proteins. However, high-level expression often
leads to insoluble protein aggregates such as inclusion bodies, and

PLOS ONE | www.plosone.org 5 February 2013 | Volume 8 | Issue 2 | e56168

 Solubilization and Purification of Recombinant hGH

Figure 4. Characterization of purified hGH. (A) RP-HPLC chromatograms of purified hGH. Separation was performed by isocratic elution using
trifluoroacetic acid-water-acetonitrile. Vertical axis units represent mV detector output. (B) Analytical size exclusion chromatography (SEC) of purified
hGH. (C) MALDI-TOF mass spectrometry analysis of purified intact hGH. Spectra were acquired over the m/z range 15000–45000 Da in the positive ion
mode.
doi:10.1371/journal.pone.0056168.g004

Figure 5. Circular dichroism (CD) analysis of purified hGH. The CD spectra of control hGH, His-hGH and untagged hGH were scanned in the
UV range 190–250 nm.
doi:10.1371/journal.pone.0056168.g005

PLOS ONE | www.plosone.org 6 February 2013 | Volume 8 | Issue 2 | e56168

 Solubilization and Purification of Recombinant hGH

Figure 6. NB2-11 cell proliferation assay of purified hGH. Nb2-11 cells were arrested by serum deprivation and then incubated for 48 h in the
presence of control hGH (%), His-hGH (m), untagged hGH (#), or BSA (X) at the indicated concentrations. Cell numbers were determined as
described in Material and Methods, and averages from three independent experiments are graphed and presented as mean 6 standard error.
doi:10.1371/journal.pone.0056168.g006

thus requires additional steps including solubilization and refolding observation may be explained by the fact that addition of salt leads
prior to purification. Although the formation of inclusion bodies to increased hydrophobicity on the protein surface, which results
has certain advantages such as convenient isolation and protection in protein aggregation through hydrophobic interactions, a
from proteolysis, the recovery of biological activity has often been process referred to as ‘salting out’ [27].
unsatisfactory. Furthermore, the resolubilization of protein aggre- It should be noted that the soluble fraction of hGH can be
gates can be problematic, requires the use of high concentrations destabilized and aggregated during purification. This observation
of denaturants, and the subsequent refolding process generally often occurred with buffer exchange, which is necessary for
requires extensive optimization. subsequent purification steps. As glycerol is an efficient stabilizing
In this study, we demonstrated the efficient preparation of agent for many proteins [28], we included 10% glycerol in buffers
soluble hGH directly from the E. coli cytoplasmic space under during purification and found no detectable protein aggregates.
optimized conditions where protein aggregation was minimized. It is known that hGH contains two intramolecular disulfide
Although we found that a lowering of the induction temperature bonds, one between Cys-53 and Cys-165 and the other between
significantly decreased the formation of inclusion bodies, consid- Cys-182 and Cys-189 [29]. Although for some proteins disulfide
erable amounts of recombinant hGH were still found in the bonds are crucial for their folding as reducing disulfide bonds can
insoluble pellet fraction. Unlike inclusion bodies produced by result in inactive protein, previous studies indicate that the
induction at high temperature, we were able to solubilize most of disulfide bonds in hGH do not appear to be necessary for
the pelleted hGH under our optimized protein extraction biological activity [30,31,32,33]. Although hGH was expressed as
conditions. This observation suggests that they are reversible a reduced form in the E. coli cytoplasmic space, we found that the
protein aggregates or clusters, rather than inclusion bodies, cystein residues of hGH were spontaneously oxidized and formed
probably resulting from non-covalent protein interactions. Nu- intramolecular disulfide bonds under our purification conditions.
merous detergents are known to disrupt such protein interactions. Therefore, we are currently exploring whether those disulfide
While anionic and cationic detergents are likely to disrupt protein bonds are formed among correctly paired cysteins.
structure, non-ionic detergents including Triton X-100 generally It was previously reported that the auto-induction process can
do not disrupt protein structure. We showed that the proper use of be used as an efficient protein expression method [34]. Using this
non-ionic detergents significantly increased the solubility of hGH method, high-level expression of bubaline growth hormone was
during protein extraction. The solubilizing effects of such achieved, while most of the protein was found in the form of
detergents can be attributed to an association with hydrophobic inclusion bodies [35]. Therefore, we plan to determine whether
parts of hGH, disrupting non-specific intermolecular hydrophobic the production of hGH can be further improved by the auto-
interactions. Since detergent-dependent protein solubility varies induction method combined with our established procedure for
with the nature of the protein and experimental conditions, an high recovery of soluble protein.
appropriate detergent should be determined empirically. In conclusion, we describe an efficient strategy that allowed us
While salt solutions (e.g., 150 mM NaCl) are commonly used to maximize the overall recovery of soluble hGH, which was
for protein extraction, under our conditions the solubility of hGH structurally intact and biologically active. We showed that the
was somewhat decreased by the addition of NaCl or KCl. Our solubility of hGH expressed in bacterial cytoplasm can be

PLOS ONE | www.plosone.org 7 February 2013 | Volume 8 | Issue 2 | e56168

 Solubilization and Purification of Recombinant hGH

significantly increased by proper optimization of protein expres- Supporting Information

sion, extraction and purification procedures. Previously, prior to
optimization most of the overexpressed hGH was insoluble, but Figure S1 Solubility comparison of recombinant hGH
under our optimized conditions we obtained up to 95% of hGH in expressed at different IPTG concentrations. The levels of
the soluble fraction. The soluble recombinant hGH was subse- soluble hGH induced at various IPTG concentrations were
quently purified by a series of chromatographic steps, and various quantified and graphed by a densitometry assay using Image-
analytical techniques confirmed the high purity, molecular mass, QuantTM TL 5.2 analysis software.
(TIF)
and structural integrity of the purified protein. Moreover, the high
growth-promoting activity of purified hGH was confirmed by cell Table S1 Purification of His-hGH from E. coli.
proliferation assays. Therefore, the combined effects of optimized (DOCX)
protein expression, extraction and purification conditions signif- Table S2 Purification of untagged hGH from E. coli.
icantly increased the overall recovery of soluble hGH, which is (DOCX)
structurally intact and biologically active.
We believe that the procedures established in this work can be
Author Contributions
used by the biopharmaceutical industry to efficiently purify
recombinant hGH from E. coli and are also worth exploring with Conceived and designed the experiments: MJK HP JHC. Performed the
experiments: MJK HP KHS HJY SKK JHC. Analyzed the data: MJK HP
other recombinant proteins in bacterial systems.
KHS JHC. Contributed reagents/materials/analysis tools: MJK HP KHS
HJY SKK JHC. Wrote the paper: MJK HP KHS JHC.

References
1. Isaksson OG, Eden S, Jansson JO (1985) Mode of action of pituitary growth 17. Greenfield NJ (2006) Using circular dichroism spectra to estimate protein
hormone on target cells. Annu Rev Physiol 47: 483–499. secondary structure. Nat Protoc 1: 2876–2890.
2. Garcia-Barros M, Costoya JA, Rios R, Arce V, Devesa J (2000) N-glycosylated 18. Shin NK, Kim DY, Shin CS, Hong MS, Lee J, et al. (1998) High-level
variants of growth hormone in human pituitary extracts. Horm Res 53: 40–45. production of human growth hormone in Escherichia coli by a simple
3. Goeddel DV, Heyneker HL, Hozumi T, Arentzen R, Itakura K, et al. (1979) recombinant process. J Biotechnol 62: 143–151.
Direct expression in Escherichia coli of a DNA sequence coding for human 19. Patra AK, Mukhopadhyay R, Mukhija R, Krishnan A, Garg LC, et al. (2000)
growth hormone. Nature 281: 544-548. Optimization of inclusion body solubilization and renaturation of recombinant
4. Ikehara M, Ohtsuka E, Tokunaga T, Taniyama Y, Iwai S, et al. (1984) Synthesis human growth hormone from Escherichia coli. Protein Expr Purif 18: 182–192.
of a gene for human growth hormone and its expression in Escherichia coli. Proc 20. Mosmann T (1983) Rapid colorimetric assay for cellular growth and survival:
Natl Acad Sci U S A 81: 5956–5960. application to proliferation and cytotoxicity assays. J Immunol Methods 65: 55–
5. Fahnert B, Lilie H, Neubauer P (2004) Inclusion bodies: formation and 63.
utilisation. Adv Biochem Eng Biotechnol 89: 93–142. 21. Baneyx F, Mujacic M (2004) Recombinant protein folding and misfolding in
6. Datar RV, Cartwright T, Rosen CG (1993) Process economics of animal cell Escherichia coli. Nat Biotechnol 22: 1399–1408.
and bacterial fermentations: a case study analysis of tissue plasminogen activator. 22. Francis DM, Page R (2010) Strategies to optimize protein expression in E. coli.
Biotechnology (N Y) 11: 349–357. Curr Protoc Protein Sci Chapter 5: Unit 5 24 21–29.
7. Khan RH, Rao KB, Eshwari AN, Totey SM, Panda AK (1998) Solubilization of 23. Makrides SC (1996) Strategies for achieving high-level expression of genes in
recombinant ovine growth hormone with retention of native-like secondary Escherichia coli. Microbiol Rev 60: 512–538.
structure and its refolding from the inclusion bodies of Escherichia coli. 24. Sorensen HP, Mortensen KK (2005) Advanced genetic strategies for
Biotechnol Prog 14: 722–728. recombinant protein expression in Escherichia coli. J Biotechnol 115: 113–128.
25. Bondos SE, Bicknell A (2003) Detection and prevention of protein aggregation
8. Singh SM, Panda AK (2005) Solubilization and refolding of bacterial inclusion
before, during, and after purification. Anal Biochem 316: 223–231.
body proteins. J Biosci Bioeng 99: 303–310.
26. Tanaka T, Shiu RP, Gout PW, Beer CT, Noble RL, et al. (1980) A new sensitive
9. Sonoda H, Sugimura A (2008) Improved solubilization of recombinant human
and specific bioassay for lactogenic hormones: measurement of prolactin and
growth hormone inclusion body produced in Escherichia coli. Biosci Biotechnol
growth hormone in human serum. J Clin Endocrinol Metab 51: 1058–1063.
Biochem 72: 2675–2680.
27. Baldwin RL (1996) How Hofmeister ion interactions affect protein stability.
10. St John RJ, Carpenter JF, Balny C, Randolph TW (2001) High pressure Biophys J 71: 2056–2063.
refolding of recombinant human growth hormone from insoluble aggregates. 28. Gekko K, Timasheff SN (1981) Mechanism of protein stabilization by glycerol:
Structural transformations, kinetic barriers, and energetics. J Biol Chem 276: preferential hydration in glycerol-water mixtures. Biochemistry 20: 4667–4676.
46856–46863. 29. de Vos AM, Ultsch M, Kossiakoff AA (1992) Human growth hormone and
11. Becker GW, Hsiung HM (1986) Expression, secretion and folding of human extracellular domain of its receptor: crystal structure of the complex. Science
growth hormone in Escherichia coli. Purification and characterization. FEBS 255: 306–312.
Lett 204: 145–150. 30. Campbell RM, Kostyo JL, Scanes CG (1990) Lipolytic and antilipolytic effects of
12. Gray GL, Baldridge JS, McKeown KS, Heyneker HL, Chang CN (1985) human growth hormone, its 20-kilodalton variant, a reduced and carboxy-
Periplasmic production of correctly processed human growth hormone in methylated derivative, and human placental lactogen on chicken adipose tissue
Escherichia coli: natural and bacterial signal sequences are interchangeable. in vitro. Proc Soc Exp Biol Med 193: 269–273.
Gene 39: 247–254. 31. Necessary PC, Andersen TT, Ebner KE (1985) Activity of alkylated prolactin
13. Chang CN, Rey M, Bochner B, Heyneker H, Gray G (1987) High-level and human growth hormone in receptor and cell assays. Mol Cell Endocrinol
secretion of human growth hormone by Escherichia coli. Gene 55: 189–196. 39: 247–254.
14. Soares CR, Gomide FI, Ueda EK, Bartolini P (2003) Periplasmic expression of 32. Dixon JS, Li CH (1966) Retention of the biological potency of human pituitary
human growth hormone via plasmid vectors containing the lambdaPL growth hormone after reduction and carbamidomethylation. Science 154: 785–
promoter: use of HPLC for product quantification. Protein Eng 16: 1131–1138. 786.
15. Bradford MM (1976) A rapid and sensitive method for the quantitation of 33. Bewley TA, Brovetto-Cruz J, Li CH (1969) Human pituitary growth hormone.
microgram quantities of protein utilizing the principle of protein-dye binding. Physicochemical investigations of the native and reduced-alkylated protein.
Anal Biochem 72: 248–254. Biochemistry 8: 4701–4708.
16. Jeong JS, Lim HM, Kim SK, Ku HK, Oh KH, et al. (2011) Quantification of 34. Studier FW (2005) Protein production by auto-induction in high density shaking
human growth hormone by amino acid composition analysis using isotope cultures. Protein Expr Purif 41: 207–234.
dilution liquid-chromatography tandem mass spectrometry. J Chromatogr A 35. Sadaf S, Khan MA, Akhtar MW (2007) Production of bubaline somatotropin by
1218: 6596–6602. auto-induction in Escherichia coli. Biotechnol Appl Biochem 47: 21–26.

PLOS ONE | www.plosone.org 8 February 2013 | Volume 8 | Issue 2 | e56168