Current Pharmaceutical Biotechnology, 2010, 11, 309-312 309

Non-Denaturing Solubilization of Inclusion Bodies

Kouhei Tsumoto1, Ryota Abe1, Daisuke Ejima2 and Tsutomu Arakawa3,*

1
Department of Medical Genome Sciences, Graduate School of Frontier Sciences, University of Tokyo, Kashiwa 277-
8562, Japan; 2Applied Research Department, Amino Science Laboratories, Ajinomoto, Inc., Kawasaki, Kanagawa 210-
8681, Japan; 3Alliance Protein Laboratories, 3957 Corte Cancion, Thousand Oaks, CA 91360, USA

Abstract: It has been a conventional notion that cytoplasmic recombinant expression leads to either soluble protein or in-
clusion bodies. In the latter case, it was always assumed that proteins in inclusion bodies (IBs) are more or less unfolded
and hence require complete denaturing condition for solubilization, which uses strong detergents, urea or guanidine hy-
drochloride. However, we often observe distribution of expressed proteins in both soluble and insoluble fractions. In such
expression, IBs are often loose and of flocculate morphology. We believe that such distribution is due to association of
near native structures of the expressed proteins, which cause either aggregation into insoluble fractions or unstable soluble
proteins. In our experience, although not reported by others, interleukin-1
, interferon-, tumor necrosis factors, fibroblast
growth factors, His-tagged fyn kinase and many other proteins showed such behavior. If this occurs, we have experienced
problems of instability, low yield and insolubility whether purification is done from the soluble fraction or by refolding of
IBs. Arginine has shown great promise in non-denaturaing solubilization of some of these proteins we have tested.
Keywords: Non-denaturing solubilization, inclusion bodies, arginine, GFP, FGF-20.

INTRODUCTION The pH of arginine can be adjusted to usually neutral value

Cytoplasmic expression, e.g. in Escherichia coli (E. coli),
of recombinant proteins occasionally results in distribution
of soluble and insoluble proteins trapped as inclusion bodies
(IBs) [1]. Although may not be universal, the soluble portion
of the proteins from such expression is unstable, requiring
solubilizing or stabilizing agents for purification. The in-
soluble portion is loose and of flocculate morphology [2],
making it difficult to separate from the supernatant by low
speed centrifugation. Although not certainly universal, pro-
teins trapped in such loose IBs are often difficult to refold
once fully unfolded by denaturing agents. When expressed
protein distributes as above, the ratio of soluble vs. insoluble
is also variable and difficult to control. It has been difficult
to treat either portion of the protein. Here we show three
cases that have been successfully managed to produce pro-
teins using aqueous arginine solution.
PROTOCOL
Distribution of expressed protein in E. coli is frequently
observed as described above and is schematically depicted in
Fig. (1). In such case, an approach is to separate the soluble
and insoluble fractions and process separately (see right col-
umn of Fig. 1). It is possible that the whole cell can be sus-
pended and lysed in aqueous arginine solution. In either case,
solubilization increases with arginine concentration. As ar-
ginine does not denature proteins [3], even 2 M solution can
be used for solubilization. Salting-out buffers (e.g., citrate
and phosphate) may offset salting-in effects of arginine and
hence high concentration of such buffers should be avoided.
*Address correspondence to this author at the Alliance Protein Laboratories,
3957 Corte Cancion, Thousand Oaks, CA 91360, USA; Tel: 805-388-1074;
Fax: 805-388-7252; E-mail: tarakawa2@aol.com
with HCl, as the commonly used form of arginine is arginine
hydrochloride. Solubilization may be temperature dependent,
higher temperature leading to greater solubilization. As ar-
ginine does not stabilize (nor destabilize) the proteins, exces-
sively high temperatures may denature the solubilized pro-
teins [3].
FGF-20
Purification of fibroblast growth factor-20 (FGF-20) was
carried out in the presence of arginine, although the reason is
not explicitly provided [4], although recent publication
shows higher solubility of FGF-20 in the presence of argin-
ine up to 0.5 M [5], in particular at 4oC. With arginine sul-
fate, however, solubility decrease was observed at 1.5 M,
perhaps due to salting-out nature arising from the counter ion
of sulfate [5]. E. coli cells, expressing His-tagged FGF-20,
were suspended in phosphate-buffered saline (PBS) contain-
ing 0.5 M NaCl and 1 M arginine (counter ion used was not
explicitly stated) and disrupted by pressure homogeneization.
This leads to suspension of whole cells containing a majority
of proteins, both E. coli proteins and FGF-20, in the super-
natant, which requires extensive purification fold. The su-
pernatant was loaded on to a Ni-Sepharose. The column was
washed with PBS containing 0.5 M NaCl and 1 M arginine.
As arginine weakly competes with histidine for binding to
Ni-Sepharose, it is possible that this washing step may elute
a majority of non-specific protein binding and a portion of
His-tagged FGF-20 [6]. The bound protein was eluted with
imidazole and further purified in the presence of 1 M argin-
ine. It appears that 1 M arginine is required for FGF-20 to be
stable or soluble in aqueous solution. As FGF-20 is an en-
dogenous protein, its structure and solubility in circulation
may be maintained only at low concentration or by binding
to other proteins. To our knowledge, this seems to be an only

1389-2010/10 $55.00+.00 © 2010 Bentham Science Publishers Ltd.

310 Current Pharmaceutical Biotechnology, 2010, Vol. 11, No. 3
Fig. (1). Distribution of expressed protein and arginine solubilization.
example where a protein was processed throughout in the
presence of arginine and proves that such purification is fea-
sible.
GREEN FLUORESCENT PROTEIN
Green fluorescent protein (GFP) also showed distribution
into soluble and insoluble fractions [7]. We thus applied the
arginine solubilization method to GFP. Expression of GFP in
E. coli resulted in only small amount of soluble and fluores-
cent GFP protein and hence most of the protein in insoluble
particles. The expressed GFP in insoluble particles, however,
was fluorescent (see Fig. 2A), indicating that it is at least in
part folded with an intact chromophore. The precipitated
insoluble particles were suspended in aqueous solution at pH
8.0 containing gurandine hydrochloride (GdnHCl) or argin-
ine hydrochloride at various concentrations (all the solutions
contain 50 mM Tris–HCl and 200 mM NaCl). The suspen-
sions were gently rotated at 4 °C for 12 h prior to centrifuga-
tion. The GFP in insoluble particles could not be solubilized
by an aqueous, denaturant-free buffer (Fig. 2B). Solubiliza-
tion from insoluble particles by 6 M GdnHCl led to complete
denaturation of the GFP existing in insoluble particles, as is
evident with no fluorescence of the supernatant (see Fig.
2C). Coversely, GdnHCl solution at lower concentration,
e.g., 2 M, could solubilize intact fluorescently active GFP
(Fig. 2D). Solubilization of active GFP from insoluble parti-
cles was also achieved by arginine at 1 and 2 M (e.g., Fig.
2E). It is noteworthy that arginine was stronger in solubiliz-
ing insoluble GFP than GdnHCl below 2 M. The solubilized
GFP was subjected to dialysis against an arginine-free solu-
tion (50 mM Tris–HCl, pH 8.0 containing 200 mM NaCl) for
12 h at 4 °C. GdnHCl-solubilized GFP lost its specific fluo-
rescence after dialysis; on the other hand, fluorescent GFP
could be recovered from arginine-solubilized GFP. These
results demonstrate that GFP expressed in E. coli may form
insoluble particles containing native conformation and argin-
ine may be used to recover the proteins in the native form
from IB, as in this case.
Tsumoto et al.

2-MICROGLOBULIN
We then examined IB of
2-microglobulin (
2-m) ex-
pressed in E. coli [8]. Expression of
2-m in E. coli resulted
in formation of inclusion bodies. Surprisingly, attenuated
total reflectance Fourier transform infrared analysis shown in
Fig. (3) suggested that
2-m in the inclusion bodies included
a native-like secondary structure. Then, we tried to solubilize
the native-like
2-m from IBs using arginine solution. A
300-mg sample (wet weight) of the
2-m IBs was soaked in
10 ml of 2 M arginine, 2 M GdnHCl or 6 M GdnHCl solu-
tions (all the arginine solutions or GdnHCl solutions also
contained 50 mM Tris–HCl, pH 8.0, 200 mM NaCl, and
1 mM EDTA) for 20 h at various temperatures, and then
centrifuged. The soluble
2-m present in each supernatant
was then buffer-exchanged at 4 °C by dialysis against a 50-
fold volume (relative to the volume of the supernatant) of a
denaturant-free solution (50 mM Tris–HCl, pH 8.0, 200 mM
NaCl, and 1 mM EDTA). Removal of GdnHCl after solubi-
lization with 6 and 2 M GdnHCl solutions resulted in a low
yield of soluble
2-m (6 M GdnHCl; 68%, 2 M GdnHCl;
44%), due to the formation of insoluble aggregates. On the
other hand, no insoluble aggregates were formed in the case
of arginine. Free thiol groups were quantified for the
2-m
preparations, so that few free thiol groups were observed for
all the
2-m preparations.
The effect of temperature on solubilization of
2-m by
arginine was examined. Higher temperatures clearly led to
greater solubilization of
2-m. The amount of solubilized
2-
m did not increase linearly with respect to the temperature.
Solubilization at moderate temperature gave
2-m with an
apparently native structure.
After removal of arginine, the
2-m preparations were
examined further for evidence of aggregation, by size exclu-
sion chromatography (SEC) analysis. The material that had
been solubilized at 28–50 °C revealed that
2-m was present
mainly in a monomeric form and that the extracts contained
only a small amount of oligomers eluting in earlier fractions.
The amount of the monomeric form decreased with increas-
ing oligomers for extraction temperatures greater than 60 °C.
Non-Denaturing Solubilization of Inclusion Bodies
Fig. (2). Solubilization of GFP insoluble particles by GdnHCl or arginine.
The monomeric form of 2-m fractionated by SEC was ana-
lyzed by measuring tryptophan fluorescence emission and
circular dichroism spectroscopy; the profiles from the 2-m
monomer solubilized at 28–50 °C were identical to those of
the native protein. These results are consistent with the idea
that the effect of arginine is neither the destabilization of
incorrectly folded structure nor the facilitation of refolding,
but rather the suppression of insoluble-aggregate formation
[3, 9].
FUTURE PROSPECT OF ARGININE SOLUBILI-
ZATION
We have shown above three cases where arginine does
solubilize proteins from IBs. How applicable is this technol-
ogy? Before answering this question, a more appropriate
question may be how universal it is to see an unstable, na-
tive-like fold, a loose “flocculate-type” IBs, or distribution of
the expressed proteins to these states. We have seen quite
often these. For example, several extracellular proteins, in-
cluding interferon-, interleukin-1
, tumor necrosis factors,
FGF and immunoglobulin (Ig) domain, showed such behav-
ior when expressed in E. coli. One common property among
these proteins is that either they contain no disulfide bonds
or disulfide bonds are not essential for folding. For intracel-
lular proteins, histidine-tagged Fyn tyrosine kinase (His-Fyn)
showed an interesting behavior when expressed in E. coli
(Arakawa, T. and Yamamoto, T., unpublished observation).
Current Pharmaceutical Biotechnology, 2010, Vol. 11, No. 3 311
A large portion of expressed His-Fyn appeared in the soluble
fraction of E. coli cell lysate and could readily be bound to
and eluted from Ni-columns. However, the His-Fyn thus
isolated from the lysates was unstable, resulting in precipita-
tion. Further purification turned out to be impossible due to
this stability problem. This could be due to number of rea-
sons, e.g., intrinsic instability of Fyn kinase. This protein
may be stabilized in the cytoplasmic environments, e.g., mo-
lecular crowding [10, 11] or existence of stabilizing mole-
cules. Isolation might have removed such stabilizing factors.
In these occasions, finding and adding such factors would be
a solution.
Interleukin-1
gives another unique example. Cytoplas-
mic soluble expression of the protein was highly variable for
unknown reason, but once expressed soluble, it was rela-
tively easy to purify, i.e., no instability issue. Thus, in this
case the soluble portion appears to be fully folded or at least
stable for purification. However, once the IBs are solubilized
by denaturants, its refolding turned out to be rather difficult.
Although we do not know whether arginine would be effec-
tive for solubilizing the IBs, non-denaturing solubilization
would have a great advantage.
Another interesting observation is Ig domain. When Ig
domain was expressed in E. coli, it was often soluble, but is
non-native. We have observed that the expressed Ig domain
could be readily purified, but the purified protein contained
312 Current Pharmaceutical Biotechnology, 2010, Vol. 11, No. 3 Tsumoto et al.

Fig. (3). FT-IR spectrum of 2-m IBs

Attenuated total reflectance (ATR) Fourier transform infrared spectrum of 2-m IBs was determined after purification of the IBs with a FT-
IR-680plus spectrometer (Jasco, Tokyo, Japan) at a resolution of 2 cm
1. The FT-IR absorption spectrum (black curve) was deconvolved
using the nonlinear curve-fitting program GRAMS/32 version 5.0 (Galactic, Salem, USA), and the fitting was continued until
2 (the sum of
squares of the deviations normalized by the variance of the count) was <3.0. The deconvolved spectrum shows a main peak at 1623 cm-1, con-
sistent with the secondary structure of the native 2-m rich in anti-parallel -sheets.

no intra-chain disulfide bond formed. In other words, the Ig
domain folds as native-like structure, but with two cysteines
unfavorable to form a disulfide bond. Purification and stor-
age in buffer had no impact on disulfide bond formation.
Unfolding and refolding coupled with oxidation were re-
quired for disulfide formation. Arginine may be used to fa-
cilitate conformational fluctuation without fully denaturing
the protein.
Lastly and more importantly, it would be desirable if
arginine solubilization is compatible with the purification
process following the solubilization. As described above for
Fyn kinase, fusion technology, in particular His-tag, is one of
the most commonly used technologies for both expression
and purification. Recently, we have observed that a Staphy-
lococcus aureus IsdA can be solubilized by 0.2 M arginine,
pH 8.0 and His-IsdA thus solubilized binds to Ni-column in
the presence of 0.2 M arginine [6]. The purified His-IsdA
was biologically active, consistent with the ability of argin-
ine to solubilize proteins in the native structure. More impor-
tantly, this protein and a few other His-tag proteins showed
binding to Ni-columns, although higher arginine concentra-
tion, e.g., 0.5 M, clearly interfered with the binding.
REFERENCES
[1] Jung, G.; Denefle, P.; Becquart, J.; Mayaux, J.-F. High-cell density
fermentation studies of recombinant Escherichia coli strains ex-
pressing human interleukin-1 beta. Ann. Inst. Pasteur. Microbiol.,
1988, 139, 129-146.
[2] Hart, R.A.; Rinas, U.; Bailey, J.E. Protein composition of Vit-
reoscilla hemoglobin inclusion bodies produced in Escherichia
coli. J. Biol. Chem., 1990, 265, 12728-12733.
[3] Arakawa, T.; Tsumoto, K. The effects of arginine on refolding of
aggregated proteins: not facilitate refolding, but suppress aggrega-
tion. Biochem. Biophys. Res. Commun., 2003, 304, 148-152.
[4] Jeffers, M.; McDonald, W.F.; Chillakuru, R.A.; Yang, M.; Nakase,
H.; Deegler, L.L.; Sylander, E.D.; Rittman, B.; Bendele, A.; Sartor,
B.; Lichenstein, H.S. A novel human fibroblast growth factor treats
experimental intestinal inflammation. Gasroenterology, 2002,
1123, 1151-1162.
[5] Maity, H.; Karkaria, C.; Davagnino, J. Mapping of solution com-
ponents, pH changes, protein stability and the elimination of pro-
tein precipitation during freeze-thawing of fibroblast growth factor
20. Int. J. Pharm., 2009, 378, 122-135.
[6] Abe, R.; Kudou, M.; Tanaka, Y.; Arakawa, T.; Tsumoto, K.
Immobilized metal affinity chromatography in the presence of ar-
ginine. Biochem. Biophys. Res. Commun., 2009, 381, 306-310.
[7] Tsumoto, K.; Umetsu, M.; Kumagai, I.; Ejima, D.; Arakawa, T.
Solubilization of active green fluorescent protein from insoluble
particles by guanidine and arginine. Biochem. Biophys. Res. Com-
mun., 2003, 312, 1383-1386.
[8] Umetsu, M.; Tsumoto, K.; Nitta, S.; Adschiri, T.; Ejima, D.; Ara-
kawa, T.; Kumagai, I. Nondenaturing solubilization of beta2 mi-
croglobulin from inclusion bodies by L-arginine. Biochem. Bio-
phys. Res. Commun., 2005, 328, 189-197.
[9] Arakawa, T.; Kita, Y.; Tsumoto, K.; Ejima, D.; Fukada, H.
Aggregation suppression of proteins by arginine during thermal un-
folding. Protein Pept. Lett., 2006, 13, 921-927.
[10] Minton, A.P. Influence of macromolecular crowding upon the
stability and state of association of proteins: predictions and obser-
vations. J. Pharm. Sci., 2005, 94, 1668-1675.
[11] Sasahara, K.; McPhie, P.; Minton, A.P. Effect of dextran on protein
stability and conformation attributed to macromolecular crowding.
J. Mol. Biol., 2003, 326, 1227-1237.

Received: June 01, 2009 Accepted: July 09, 2009

Copyright of Current Pharmaceutical Biotechnology is the property of Bentham Science Publishers Ltd. and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.