| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
(Received for publication, January 13, 1995) From the
Mouse Interleukin 4 is a 20-kDa glycoprotein, synthesized by
activated T lymphocytes and mast cells, which regulates the growth
and/or differentiation of a broad spectrum of target cells of the
immune system, including B and T lymphocytes, macrophages, and
hematopoietic progenitor cells. Using an inducible recA promoter and the g10-L ribosome-binding site, recombinant
non-glycosylated interleukin 4 (IL-4) was expressed as 17% of total
cellular protein in Escherichia coli inclusion bodies, as a
reduced, inactive 14.5-kDa polypeptide. The protein was refolded and
aggregates dissociated when three disulfide bonds were reformed by
slowly decreasing the concentration of guanidine hydrochloride and
cysteine. The oxidized monomer was purified to homogeneity by
sequential ion-exchange and size exclusion chromatography. When
compared with native IL-4, E. coli-derived IL-4 displayed an
identical specific activity of 4-7 Murine Interleukin 4 (IL-4) ( We have
previously described an inducible, efficient expression vector, which
was designed to optimize the transcriptional and translational
efficiency of foreign gene expression in E. coli (Olins et
al., 1988). Transcriptional regulation of recombinant IL-4
expression is mediated by the DNA repair/recombination promoter rec A which can be induced with nalidixic acid. Translational
efficiency is increased by using the T7 phage g10-L
translational enhancer which has been shown to be one of the most
effective ribosome-binding sites for the initiation of
translation in E. coli (Olins and Rangwala, 1989, 1990). In
this report we describe the construction of an inducible expression
vector for the synthesis of recombinant mouse IL-4 in E. coli.
The IL-4 molecule was expressed as a 14-kDa protein which,
after disruption of aggregates, refolding, and purification,
retained all of the biological activities tested that have been
ascribed in vitro to both natural and recombinant glycosylated
mouse IL-4. The ability to purify milligram quantities of this protein
has allowed us to examine the immunoregulatory properties of IL-4 in vivo (Racke et al., 1994) and has aided the effort
to characterize the biochemical properties of this molecule (Carr et al., 1991).
Figure 1:
Plasmid vector for expression of mIL-4
in E. coli. The plasmid is based on pBR 327 (Soberon et
al., 1980) (from the SalI to EcoRI sites).
Inducible transcription is from the E. coli recA promoter
(Feinstein et al., 1983; Olins et al., 1988), and
efficient translation is provided by the g10-L
ribosome-binding site (Olins et al., 1988; Olins and Rangwala,
1989). The NcoI and EcoRI sites which were filled-in
with DNA polymerase are denoted by NcoI-X and EcoRI-X, respectively. The plasmid includes the pBR 327 origin
of replication (ori-327) and the F1 bacteriophage origin of
single-stranded replication (ori-f1). The selectable
marker is
Figure 2:
SDS-polyacrylamide showing the induction
and expression of E. coli-derived IL-4. E. coli cells
bearing plasmid pMON 5738 were grown to an OD
Figure 3:
Summary of
the refold protocol and purification scheme of E. coli-derived
IL-4. Inclusion bodies were isolated as described under
``Experimental Procedures.'' Recombinant mouse IL-4 was
refolded and purified as depicted by the cartoon and described
in detail under ``Experimental
Procedures.''
Figure 4:
HPLC-TSK analytic chromatography detailing
the time course of oxidation and refolding of IL-4. Oxidation and
refolding efficiency of IL-4 was monitored by size exclusion
chromatography on a 25-cm TSK SW3000 column with a 7-cm precolumn. The
column was run at 0.5 ml/min with a mobile phase of 0.1 M Na
Figure 5:
Cation-exchange chromatography of refolded
IL-4. After the precipitation of the preparation by lowering the pH to
5.0, IL-4 was further purified on a S-Sepharose Fast Flow column. The
sterile filtered sample was applied to a 1
Figure 6:
Purification of IL-4 by TSK sizing matrix
HW55 fine. The pooled biologically active fractions from the
ion-exchange column were applied to a 4.4
Figure 7:
Refolded E. coli-derived
recombinant IL-4 shares an identical dose-response curve with natural
IL-4. Purified E. coli-derived (TSK HW55 monomeric fraction)
and natural IL-4 were diluted into cell culture media and added to
FDC-P1 cells at half-log dilutions for an overnight incubation. After a
4-h pulse with [
Figure 8:
Two- dimensional gel electrophoresis
analysis of purified, monomeric, biologically active IL-4. The IL-4 was
focused under non-equilibrium conditions for 1050 V-hrs. The gel was
oriented with the acidic end of the gel to the left. A,
Coomassie Blue R-250 stained, 1.0 µg of IL-4; B,
immunoblot of A, 2-h exposure; C, silver stained, 0.5
µg of IL-4. Molecular mass markers are phosphorylase b, 98
kDa; bovine serum albumin, 68 kDa; ovalbumin, 43 kDa, carbonic
anhydrase, 29 kDa; soybean trypsin inhibitor, 21 kDa; lysozyme, 14 kDa.
Further details are described in the text and under ``Experimental
Procedures.''
Purified, biologically active IL-4 consisted of at least
three isoforms, each with an apparent molecular mass of approximately
14 kDa (Fig. 8A). All three isoforms are immunoreactive
with the anti-IL-4 peptide anti-sera (Fig. 8C). A minor
amount of 20-kDa material is detected by silver staining. However, this
material was not detected by either Coomassie Blue staining nor by the
anti-IL-4 peptide anti-sera (Fig. 8, B and C),
and probably represents a minor contaminant, which does not contain the
epitopes recognized by the anti-sera mixture. The purity of the rIL-4
was also assessed by chromatographic separation on a microbore reverse
phase C-8 column. While two-dimensional gel electrophoresis revealed
three isoforms of rIL-4, the reverse-phase column was unable to resolve
these structures (Fig. 9). Experiments to determine if these
three isoforms exist in deglycosylated native IL-4 have not been
performed.
Figure 9:
Reverse-phase microbore HPLC of IL-4.
Purified E. coli-derived IL-4 was applied to an Applied
Biosystems C8-microbore column run at 1 ml/min at room temperature. The
protein was eluted with a gradient from 0 to 60% (v/v) acetonitrile in
0.05% (v/v) trifluoroacetic acid. The protein elution was monitored
with a UV detector at 280 nm.
E. coli contains a methionine aminopeptidase
whose activity is dependent on the second encoded amino acid
(Ben-Bassat et al., 1987). As shown in Table 2,
NH IL-4 regulates the growth and differentiation of a variety of
cells of hematopoietic lineage, when it is assayed in in vitro using cell culture models. While studies to examine the effects of
this regulatory protein in vivo are difficult to undertake
directly, three indirect approaches have been used. Two series of
transgenic mice have been constructed. One set expresses IL-4
constitutively by both B and T cells under the regulation of the
immunoglobulin enhancer (Tepper et al., 1990). The other line
of transgenic mice expresses IL-4 locally within thymocytes using the lck promoter (Lewis et al., 1990). The data revealed
thus far from these transgenic studies suggest an important in vivo role for IL-4 in IgE regulation and thymocyte differentiation. A
second approach to define the function of IL-4 in regulating an immune
response in vivo has been to neutralize its activity with
blocking monoclonal antibodies (Finkelman et al., 1990). These
studies demonstrated that IL-4 is necessary for the synthesis of IgE in
a variety of models (Finkelman et al., 1986, 1988, 1989), but
is not necessary to stimulate an IgG1 response, as is predicted from
the in vitro data (Noma et al., 1986; Snapper et
al., 1987; Finkelman et al., 1987). In addition,
transgenic mice homozygous for a mutation that inactivates the IL-4
gene were generated. Studies with these mice confirmed the critical
role played by IL-4 in regulating an IgE response and the Th2 phenotype
(Kuhn et al., 1991; Kopf et al., 1993). Each of these
approaches have provided important clues toward unraveling the complex
regulatory network associated with the biology of IL-4, but each has
its limitation. A complementary approach to this problem is to
administer IL-4 protein directly to an animal in controlled doses and
at specific times. We describe in this report the synthesis,
purification, and characterization of recombinant IL-4 from a
prokaryotic host. A scale up of the procedure described here generates
100 mg of purified rIL-4/liter of fermentation, which has already lead
to the evaluation of IL-4 in vivo in murine models of immune
dysfunction (Racke et al., 1994). This correctly folded,
purified protein was also used for studying its biochemical properties
(Carr et al., 1991) and in preparing monoconal and polyclonal
antibodies. The majority of foreign proteins produced in E. coli accumulate in inclusion bodies (Olins and Rangwala, 1990). Since
inclusion bodies can be rapidly removed from the total cell lysate by
centrifugation, IL-4 can be readily purified as the major protein
component in the sample (Fig. 1, Pellet). However,
since IL-4 can potentially form three disulfide bonds and at least one
of these is required for biological activity (Ohara et al.,
1987; Carr et al., 1991), the use of inclusion bodies for
purification is only attractive if conditions for efficient refolding
of the protein can be identified. Other immunomodulatory proteins, such
as macrophage colony-stimulating factor (Halenbeck et al.,
1989; Taylor et al., 1994) and interferon- It was not possible to determine the biological activity of
IL-4 in the inclusion body lysate due to the toxicity of the guanidine
on the target cells. Instead the efficiency of refolding was monitored
by following the conversion of the reduced monomer to oxidized monomer
on the size exclusion column (Fig. 4). Surprisingly, the
oxidized form eluted from the TSK column first. We do not have an
explanation, since oxidized proteins normally assume a globular
conformation which reduces the Stokes' radius (Freifelder, 1982).
We propose that the reduced monomer may have interacted more directly
with the surface of the column and been retarded by the matrix. An
important observation that led to the first successful isolation of
active IL-4 was made during analytical size exclusion chromatography of
the reduced refractile bodies dissolved in 4 M guanidine. The
apparent molecular mass of the crude refractile body preparation was
predominantly greater than 100 kDa, and on-line diode array scanning of
these high molecular mass peaks revealed a high 260-280-nm
absorbance ratio. When the pH was changed to 9.0 and 0.5 M NaCl added to the buffer, 90% of the protein previously observed
in several high molecular weight peaks coalesced into a single peak
with a retention time consistent with monomeric IL-4. Scanning of this
peak showed a much lower 260-280-nm ratio. We suspect that
aggregation with nucleic acids may have inhibited previous refolding
efforts. The negative precipitation step documented in Fig. 4E selectively removed only inactive monomeric
proteins, as expected, since we have observed that mammalian cell
culture-derived mouse IL-4 is soluble at pH 5. While the loss of
biologically active material during this acidic precipitation was not
monitored directly, the amount of oxidized, monomeric IL-4 as followed
by analytic HPLC was not reduced. The three isoforms of IL-4
revealed by two-dimensional gel electrophoresis are confirmed in the
immunoblot to be the same protein species with different net charges.
Further confirmation that the final preparation of IL-4 is a mixture of
three isoforms is shown by the NH The observation that modification
of the amino terminus of mouse IL-4 does not effect its biological
activity is supported by the data presented in this report and by the
research of others. As shown in Fig. 7and Table 1,
recombinant mouse IL-4 containing either a four- or three-amino-acid
[(M)TRS] extension has an identical specific activity
(4-7 Regulation of antibody isotype selection, thymocyte
differentiation, and T cell maturation is a complex process, dependent
on the timing between immunization and challenge, route of antigen
exposure, age, and immunologic history of the animal. The availability
of milligram quantities of recombinant, murine IL-4 will permit an
exploration of the role of this cytokine in vivo. The data
presented in this report demonstrate how bacterially expressed IL-4 can
be refolded to full biological activity and provide useful techniques
to aid in the expression and refolding of other important biological
mediators.
Volume 270,
Number 13,
Issue of March 31, 1995 pp. 7445-7452
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
10
units/mg. This recombinant IL-4 contained a threeamino-acid
NH
-terminal extension, which did not affect its biological
activity. Purified biologically active protein consisted of three
isoforms as shown by two-dimensional gel electrophoresis, with a pI
greater than 9.0. These data suggest that neither glycosylation nor the
NH terminus of mouse IL-4 play a critical role in
contributing to its in vitro biological activity.
)is a potent mediator of
an immune response, affecting both the growth and differentiation of a
wide variety of cells in the hematopoietic lineage (Ohara and Paul,
1987). This cytokine is expressed by activated T lymphocytes (Howard et al., 1982) and mast cells (Brown et al., 1987;
Burd et al., 1989; Plaut et al., 1989) as a 20-kDa
glycoprotein (Ohara et al., 1987). The cDNA for IL-4 was
initially isolated by two laboratories, using expression vectors and
screening for either a IgG-inducing factor (Noma et al., 1986)
or a mast cell growth factor (Lee et al., 1986). The derived
amino acid sequence from the cDNA clones was used to predict a protein
backbone for IL-4 of 14 kDa. This is consistent with the observation
that N-glycanase treatment of natural IL-4, to remove N-linked carbohydrates, yields a protein core of 14 kDa (Ohara et al., 1987). Initial experiments with deglycosylated native
IL-4 and with deglycosylated recombinant IL-4, expressed initially in
yeast as a heterogeneous, hyperglycosylated molecule, suggested that
the carbohydrate modifications of IL-4 do not affect its ability to
bind to receptor and to stimulate T and B cell growth (Grabstein et
al., 1986; Park et al., 1987; Paul and Ohara, 1987).
Characterization of the role of the carbohydrate moieties on IL-4 using in vivo assays has not been rigorously examined. The Escherichia coli-derived recombinant, non-glycosylated IL-4
described in this report retains biological activity in vitro and is being used in vivo to characterize the role of
IL-4 in the synthesis of IgE, stimulating erythropoiesis, and
regulating autoimmune T cells (Racke et al., 1994).
Bacteria and Cell Lines
E. coli strain JM 101 (Messing, 1979) was used for
the subcloning and bench top expression of the cDNA coding for mouse
IL-4. The hematopoietic precursor IL-3-dependent cell line
FDC-P1 was kindly supplied by Dr. J. McKearn, Searle, St. Louis, MO,
and was maintained in Iscove's modified Dulbecco's medium,
supplemented with 10% fetal calf serum, 1 mM pyruvate, 2
mM glutamate, and 10% Wehi 3B conditioned media as a source of
IL-3 (London and McKearn, 1988).Monoclonal Antibodies and Polyclonal Sera
11B11 is a rat-mouse hybridoma which secretes an IgG2b
neutralizing antibody for mouse IL-4 (Ohara and Paul, 1985) that was
grown as an ascites tumor in pristine-primed nude mice (Charles River).
The ascites fluid was filter sterilized and used directly. Polyclonal
antisera for immunoblot assays consisted of a mixture at optimal titers
of five rabbit sera directed against synthetic mouse IL-4 peptides.
Rabbit 101 (anti-peptide 47-66), 110 (anti-peptide 21-39),
and 113 (anti-peptide 38-54) sera were diluted 1/100,000; rabbit
109 (anti-peptide 114-140) to 1/5,400; and rabbit 112
(anti-peptide 79-95) to 1/25,000. Peptides were synthesized by
the Merrifield solid-phase method on an Applied Biosystems model 430A
peptide synthesizer at an 0.5 mmol scale. A p-methylbenzhydrylamine
resin was employed for peptide amides and a phenylacetamidomethyl resin
for peptide acids. Coupling of appropriate Boc-amino acids was
performed using dicyclohexyl-carbodiimide-hydroxybenzotriazole coupling
cycles as recommended by the manufacturer. Peptides were removed from
the resin and treated with hydrogen fluoride-anisole-dimethyl sulfide.
Purification to >90% purity was accomplished by high pressure liquid
chromatography, using either a Vydac C-18 reverse-phase column
(Separations Group) or a µBondapak column (Waters) with gradients
of 0-30% acetronitrile containing 0.05% trifluoroacetic acid.
Peptide-protein conjugates were prepared by coupling cysteinyl-peptides
to thyroglobulin (Sigma) with N-succinimidyl
3-(2-pyridyldithio)propionate (Pharmacia LKB Biotechnol.) according to
the manufacturer's instructions. Amino acid numbering follows the
convention of Noma et al.(1986).Plasmid Constructions
A cDNA clone (pCB1) isolated from the murine thymoma EL-4,
subcloned into the HincII site of pGEM 3, was kindly provided
by Dr. W. Paul (Brown et al., 1987). Standard methods for DNA
manipulation were employed (Maniatis et al., 1982). Plasmid
vector pMON 5743 was used for expression of mIL-4 in E. coli (Olins and Rangwala, 1990). The coding region for the mature mIL-4
protein was obtained as an RsaI fragment and was cloned into
the NcoI site of pMON 5743, which had previously been
filled-in using dNTPs and DNA polymerase. The resulting plasmid, pMON
5738, is shown in Fig. 1. The coding region comprises the amino
acid sequence Met-Thr-Arg-Ser, followed by the mature mIL-4 coding
sequence.
-lactamase (bla). The diagram is not to
scale.
Bacterial Culture Conditions
Plasmid pMON 5738 was transformed into E. coli strain JM 101 (Messing, 1979) and selected in the presence of 200
µg/ml ampicillin. Cultures were grown at 37 °C in M9 medium
(Maniatis et al., 1982), supplemented as described in
Obukowicz et al.(1988). When the cultures reached OD
= 0.5, transcription from the recA promoter was
induced by the addition of 50 µg/ml nalidixic acid, and growth was
continued for a further 4 h. Hourly aliquots of cells were taken, and
cells were harvested by centrifugation.Cell Lysis
The total cell pellet was resuspended in 10 mM Tris-HCl, pH 8.0, 50 mM NaCl and sonicated on ice for
three 20-s bursts. The soluble supernatant fraction and insoluble
pellet were separated by centrifugation at 10,000 rpm (16, 900 g) for 10 min.Fermentation Protocol for IL-4 Production
Fermentations were run in 15-liter LSL Biolafitte fermentors
containing 10 liters of M9 minimal salts medium supplemented with 20
g/liter casamino acids, as described previously (Padgette et
al., 1987). The fermentors were inoculated with a 1-liter
overnight shake flask culture. The fermentations were run at 37 °C
with an air sparge rate of 15 liters/min and 5 pounds/square inch
backpressure. The pH was controlled at 7.0 with ammonium hydroxide, and
dissolved oxygen was maintained at 30% saturation by increasing
agitation rate from 500 to 1000 rpm. Glucose was added to the fermentor
at an initial concentration of 10 g/liter and allowed to decrease to
5.0 g/liter after which the glucose concentration was maintained at 5.0
± 1.0 g/liter by controlled feed of a 50% (w/v) glucose
solution. A Gilson Stasar II spectrophotometer was used to monitor
optical density of the culture at 550 nm. At OD =
20, induction was initiated by the addition of nalidixic acid to a
final concentration of 50 µg/ml. Cells were harvested at 3-h
post-induction by concentration in an Amicon DC10L using an HP100
hollow fiber cartridge. The cell slurry was then harvested by
centrifugation in a Beckman J2-21 centrifuge. The cell paste was
frozen at -80 °C.Purification of E. coli-derived Mouse IL-4
Cell Disruption and Isolation of Inclusion Bodies
10 g of frozen cell paste were resuspended in 1.0 liter of
0.05 M Tris-HCl, pH 7.5 (buffer A). The suspension was
homogenized in one pass at 8000 pounds/square inch with a Gaulin 15 M homogenizer equipped with a knife-edged valve. The
homogenized suspension was separated by centrifugation in a Sorvall
RC5C centrifuge with a GS-3 rotor at 6500 rpm (7140 g)
for 30 min. The supernatant was discarded. The pellet was fully
resuspended in 1.0 liter of 0.05 M Tris-HCl and precipitated
again by centrifugation at 6500 rpm in a Sorvall RC5C centrifuge with a
GS-3 rotor. This washing step was repeated twice.Formation of Intramolecular Disulfide Bonds
Denaturation, Reduction, and Oxidation
Conditions
All steps were performed at room temperature. The
washed inclusion bodies were dissolved (15-20 g pellet/liter) in
buffer B (4 M guanidine, 0.5 M NaCl, 0.05 M DTT, and 0.05 M CHES, pH 9.5, for 4-6 h. This
solution was loaded into Spectra/Por dialysis membrane tubing
(molecular weight cutoff, 3500) previously washed in Milli-Q
HO and dialyzed against 15 volumes of 4 M guanidine, 0.5 M NaCl, 0.02 M cysteine, 0.05 M CHES, pH 9.5 (buffer C) to begin slow disulfide formation.
After 12-15 h of dialysis in buffer C, the dialysis buffer was
changed to 4 M guanidine, 0.5 M NaCl, 0.05 M CHES, pH 9.5 (buffer D) to initiate fast disulfide formation.
Buffer D was changed after 8-12 h. Oxidation was complete within
24 h after dialysis against buffer D.HPLC Analytical
Oxidation of IL-4 to monomeric
isoforms was monitored by size exclusion HPLC on a TSK SW3000 column
with a 7-cm precolumn. The mobile phase was 0.1 M NaHPO
, pH 7.6. The flowrate was 0.5
ml/min, and the run time was 60 min. DTT-treated monomeric IL-4 had a
retention time of approximately 30.5 min while oxidized monomeric IL-4
eluted at approximately 28 min. Oxidation was judged complete when no
additional IL-4 had accumulated in the 28-min peak.Refolding and Partial Purification of IL-4 by
Negative Precipitation
Oxidized IL-4 was refolded, partially purified, and prepared
for ion-exchange chromatography by two negative precipitation steps.
First, the guanidine was removed by dialysis against 15 volumes of
buffer D with two buffer changes (8-12 h apart). The salt was
then removed by dialysis against 15 volumes of 0.05 M CHES, pH
9.5 (buffer E). A heavy cream-colored precipitate formed in the
dialysis bag. The refold solution was clarified by centrifugation in a
Sorvall RC5C centrifuge with a GS-3 rotor at 10,000 rpm (16,900 g) for 30 min in preparation for the second precipitation
step. The supernatant was loaded into Spectra/Por dialysis membrane
tubing (molecular weight cutoff, 3500) previously washed in Milli-Q
HO and dialyzed against 15 volumes buffer F (0.05 M acetic acid adjusted to pH 5.0 with NaOH) at 4-10 °C
with two exchanges (8-12 h apart). A white precipitate formed in
the dialysis bag. The acidified refold solution was clarified by
centrifugation in a Sorvall RC5C centrifuge with a GS-3 rotor at 10,000
rpm (16,900 g) for 30 min. The supernatant was
filtered through a 0.22-µm cellulose acetate membrane and stored at
4-10 °C.Chromatographic Purification of IL-4
Analytical
Purification beyond refold and
precipitation of IL-4 was monitored by reverse-phase HPLC on a 4.6 mm
15 cm Vydac 214TP column. At the initiation of the run, the
column was equilibrated with 30% acetonitrile in 0.05% trifluoroacetic
acid at a flow rate of 1 ml/min. For the first 3 min, the column was in
the isocratic mode. From minute 3 to minute 35 the column was developed
with a linear gradient of 30% acetonitrile to 55% acetonitrile in 0.05%
trifluoroacetic acid. The column remained in 55% acetonitrile, 0.05%
trifluoroacetic acid for 5 min.S-Sepharose Fast Flow Ion-exchange
Chromatography
Partially purified refolded mIL-4 was further
purified on Pharmacia S-Sepharose Fast Flow. The sterile filtered
sample was applied to a 1 5-cm column packed with S-Sepharose
Fast Flow and equilibrated in buffer F with a flow rate of 0.5 ml/min.
Following loading, the column was washed with 0.2 M acetic
acid, pH 5.0, until the absorbance at 280 nm returned to base line. The
column was developed with a linear gradient from 0.2 M acetic
acid to 0.2 M acetic acid, 1 M NaCl, pH 5.0.
Biologically active IL-4 eluted at a conductivity between 57 and 62
milliSiemens.Toyopearle HW-55 Size Exclusion Chromatography
The
pooled biologically active fractions from the ion-exchange column were
applied to a 4.4 80-cm column packed with Toyopearle HW-55
(fine grade) equilibrated in 0.05 M NaHPO, 0.15 M NaCl, 0.01% Tween
80 (w/v), pH 7.5. The flow rate was 1.0 ml/min, and the sample was
injected into the bottom of the column. Monomeric biologically active
IL-4 eluted in the 125-135-ml fractions.Polyacrylamide Gel Electrophoresis
Cell lysates and purified proteins were analyzed by
SDS-polyacrylamide gel electrophoresis (Laemmli, 1970; Olins et
al., 1988). Two-dimensional gel electrophoresis was performed
using non-equilibrium pH gradient electrophoresis for the first
dimension essentially as described by O'Farrell(1975). Western
blots were performed as described by Renart et al.(1979).
Silver staining was performed using a Rapid Ag Stain Kit obtained from
ICN. Development of the stain was terminated with 7% acetic acid.Biological Assays for Mouse Interleukin 4
A detailed description of the biological assays which were
used to characterize the in vitro properties of recombinant
IL-4 is described elsewhere (Conrad et al., 1989; Carr et
al., 1991; Snapper et al., 1991). Refolding efficiency
and purification of biologically active mouse rIL-4 was monitored in a
hematopoietic precursor cell proliferation assay, using the
IL-3/IL-4/granulocyte macrophage colony-stimulating factor-dependent
cell line FDC-P1 (London and McKearn, 1988). Cells were maintained in
Iscove's modified Dulbecco's medium, 10% fetal calf serum,
1% glutamine and sodium pyruvate, and 10% supernatant from Wehi 3B
cells as a source of mouse IL-3. Briefly, cells were washed three times
to remove IL-3 and resuspended at 2 10
cells/ml in
the same media lacking Wehi supernatant. 50 µl of cells were mixed
in a 96-well flat-bottomed plate (CoStar) with 50 µl of 3-fold
serially diluted samples. The culture was incubated for 16 h at 5%
CO at 37 °C and was then pulsed with 1 µCi of
[
H]thymidine (DuPont New England Nuclear) per
well for 4 h. Cells are harvested using a Skatron cell harvester, and
[H]thymidine incorporation is measured with a
liquid scintillation beta counter (Micromedics). One unit of IL-4 is
defined as the mass of IL-4 required to half-maximally stimulate
[H]thymidine uptake in the 100-µl culture.
Natural murine IL-4 was obtained from phorbol myristic acid-induced
EL-4 thymona cells and was purified as described by Ohara et
al. (1987).
Expression of Mouse IL-4 in E. coli
The coding
region for the mature mouse IL-4 protein was cloned downstream of the
ATG initiator codon of plasmid pMON 5743, as described under
``Experimental Procedures.'' The resulting expression
plasmid, pMON 5738, is shown in Fig. 1. It contains the recA promoter of E. coli and the g10-L
ribosome-binding site (Olins and Rangwala, 1990). JM 101 cells carrying
plasmid pMON 5738 were grown as described under ``Experimental
Procedures.'' When a mid-logarithmic phase of growth was reached,
an aliquot of cells was taken, and the culture was induced by the
addition of nalidixic acid. Total cellular protein was analyzed by
SDS-polyacrylamide gel electrophoresis, as illustrated in Fig. 2. As can be seen, nalidixic acid induction resulted in the
high level accumulation of a new protein, with a molecular mass of
approximately 14 kDa. This corresponds to the predicted molecular
weight for full-length mouse IL-4. Further cell fractionation,
described in Fig. 2, showed that the 14-kDa protein was
insoluble and accumulated in inclusion bodies. This property aided
considerably in subsequent purification of the protein. The expression
level of IL-4 was quantitated by densitometric scanning of a Coomassie
Blue-stained SDS-polyacrylamide gel and enzyme-linked immunoassay for
mouse IL-4. IL-4 represented 17% of total cellular protein.
Verification that the major band denoted as IL-4 in Fig. 2was
obtained by immunoblotting using a mixture of five rabbit anti-sera
raised against five different peptides synthesized from the mouse IL-4
sequence, as described under ``Experimental Procedures''
(data not shown). No biological activity was found associated with the
large mass of IL-4 in the insoluble fraction. Although a substantial
level of biological activity was detected in the soluble fraction, a
reliable estimate of specific activity was not possible due to the high
level of contamination from bacterial proteins (data not shown). While
the initial transformants selected for the expression of recombinant
IL-4 grew slowly on solid medium, there was no evidence of toxicity to
the host cell in the uninduced state in larger shake flasks or
fermentations.
=
0.5 in M9 media supplemented with casamino acids and were induced with
50 µg/ml nalidixic acid. Cells from a 1-ml culture were harvested
at 1-h time points after induction, lysed in 100 µl of Laemmli
sample buffer (Laemmli, 1970), and the proteins were fractionated by
electrophoresis on a 15% polyacrylamide gel. At 4 h after induction,
the cells were also lysed by sonication in 1 ml of PBS and soluble (Supernatant) and insoluble (Pellet) fractions were
isolated by centrifugation at 10,000 rpm in an Eppendorf
microcentrifuge and analyzed by gel electrophoresis. The mIL-4 standard
is E. coli-derived IL-4 purified as described in Fig. 3. The Coomassie Blue-staining pattern is shown. Molecular
mass markers are ovalbumin, 43 kDa; carbonic anhydrase, 29 kDa;
-lactoglobulin, 18 kDa; lysozyme, 14 kDa; bovine trypsin
inhibitor, 6 kDa; and insulin (A and B chain), 3 kDa.
Refolding of Inclusion Body-associated IL-4
While
the biological activity of the rIL-4 was found associated with the
soluble fraction of the cellular lysate, the majority of IL-4 protein
mass remained in the inclusion bodies (Fig. 2, lanes
Supernatant and Pellet). The inclusion bodies were
purified from 10 g of cell paste as described under ``Experimental
Procedures'' and the protein solubilized in 4 M guanidine, 0.5 M NaCl, 0.05 M DTT, and 0.05 M CHES, pH 9.5. The protocol of refolding is described in Fig. 3. Excess DTT was added prior to refolding to eliminate any
interpolypeptide disulfide bonds which might have formed during lysis.
Identification of the elution pattern of monomeric, oxidized IL-4 on a
size exclusion TSK SW3000 HPLC column was initially determined by
biological assay. All of the fractions shown in panel A of Fig. 4were assayed for their ability to stimulate the
proliferation of an IL-4 responsive prehematopoietic cell line. As is
shown with the dashed line, IL-4 activity eluted from the TSK
SW3000 column with a retention time of 28 min and represented a very
small percentage of total protein present in the inclusion bodies.
Oxidation of IL-4 was accomplished by sequential slow and fast
intramolecular disulfide formation. Slow disulfide bond formation was
attained by replacing the DTT reducing agent with cysteine, to retard
the rate of oxidation. The oxidation process was accelerated by removal
of the cysteine by dialysis. As shown in Fig. 4, the appearance
of oxidized IL-4 increased with the time of dialysis and was complete
within 24 h. Oxidation efficiency was 15% as measured by the analytic
HPLC size elution profile (Fig. 4D). From 1 g of
inclusion bodies, containing 500 mg IL-4, the yield of oxidized monomer
is 50-75 mg. The refolding process was initiated by the slow
removal of the guanidine denaturant by dialysis.
HPO, pH 7.6, for 60 min. Absorbance at
220 nm was monitored continuously. The initiation of disulfide bond
formation coincided with the removal of the DTT during dialysis (see Fig. 3). The appearance of oxidized IL-4 can been seen at the
elution time of 28 min. The dashed line represents the
distribution of IL-4 proliferation units, as measured by biological
assay of fractions collected from the TSK column. Samples were diluted
into PBS, sterilized with a Co-Star 0.22-µm Spin-X centrifuge
filtration unit, and added to the proliferation assay. The solid
line is absorbance at 220 nm (A) at the start of the
dialysis, 1 h after removal of the cysteine (B), 4 h later (C), and after 24 h (D). At the end of the formation
of intramolecular disulfide bonds, renaturation and refolding were
initiated when the guanidine was removed by dialysis against 0.05 M CHES, pH 9.5. The refolded solution was clarified by
centrifugation at 16,900 g for 30 min. The supernatant
was dialyzed against 0.05 M acetic acid, pH 5.0. The solution
was clarified at 16,900 g for 30 min and the soluble,
refolded protein (E) analyzed by chromatography on the TSK
column.
Purification of Refolded, Biologically Active
IL-4
As shown schematically in Fig. 3, monomeric,
oxidized IL-4 required further purification. The preparation was
acidified with acetic acid to pH 5.0 and clarified by centrifugation at
10,000 rpm for 10 min in a GS-3 rotor (Fig. 4E). The
soluble fraction, which contained a mixture of biologically active and
inactive monomeric, oxidized IL-4, was then applied to an S-Sepharose
Fast Flow ion-exchange column to remove the majority of cationic
proteins that did not bind to the column. IL-4 was eluted with an NaCl
gradient from 0.1 to 0.55 M, as shown in Fig. 5.
Elution of biologically active, refolded IL-4 was followed with the
proliferation assay as denoted by Pool in Fig. 5.
Active IL-4 eluted late from this cationic exchange resin (Fig. 5) and represented only 15% of the total IL-4 applied to
the column. Since there were aggregates of IL-4 present in this pooled
fraction, final purification involved an HPLC TSK sizing column with a
mobile phase of buffer G, applied isocratically. The majority of IL-4
protein mass eluted at 14 kDa, as predicted (Fig. 6). In
addition, the minor high molecular weight peak also contained IL-4
protein as judged by gel electrophoresis, immunoblotting, and amino
acid sequence analysis (data not shown). This aggregated IL-4 material
retained a moderate level of biological activity.
5-cm column
equilibrated in 0.05 M acetic acid, pH 5.0, with a flow rate
of 0.5 ml/min. The column was washed with 0.2 M acetic acid
until the absorbance at 280 nm returned to base line (NaCl
arrow). Elution was performed with a gradient of NaCl from
0-1 M in 0.02 M acetic acid, pH 5.0. The
biologically active fractions are indicated by the Pool notation.
80-cm Toyopearle
HW-55 size exclusion column equilibrated in 0.05 M Na HPO, 0.15 M NaCl, 0.01% Tween 80 (w/v), pH
7.5. Further details are described in the text. The major absorbance
peak of pooled fractions contains monomeric, active
IL-4.
Biological Activity of Purified rIL-4 Compared to Natural
EL-4-derived IL-4
The relative potency and specific activity of
purified recombinant IL-4 from the TSK column was compared to purified,
natural IL-4 secreted by phorbol myristic acid-stimulated EL-4 thymona.
As shown in Fig. 7, rIL-4 has an identical ED
(5
pM), a similarly shaped concentration response curve, and can
stimulate FDC-P1 cells to an identical level of proliferation as
natural IL-4. In both cases the protein concentration was determined by
amino acid composition and purity by NH-terminal sequencing
(see Table 2). The variation in quantitating protein mass by a
variety of techniques is summarized in Table 1. Purified, E.
coli-derived IL-4 has a specific activity of 4-7
10 units/mg.
H]thymidine, the cells were
harvested and counted. Protein concentrations were determined by amino
acid composition. 
, recombinant E. coli-derived IL-4;
, natural EL-4-derived IL-4.
Biochemical and Biophysical Characterization of Purified
IL-4
The biochemical properties of recombinant IL-4 were
characterized by two-dimensional gel electrophoresis. The
non-glycosylated, bacterially synthesized IL-4 exhibited insolubility
in an isoelectric focusing gel (first dimension) at sample loads of
greater than 1 µg. This form of IL-4 is also insoluble if a
non-ionic detergent such as Nonidet P-40 was present (data not shown).
Optimal results were obtained when 0.5 µg of IL-4 was focused in a
gel containing urea and the zwitterionic detergent CHAPS. The pH
gradient of the first dimension gel spanned the range of 3.6-9.0.
Since rIL-4 exhibited a pH greater than 9.0, non-equilibrium conditions
were necessary (Fig. 8), which precluded the determination of a
precise pI.
-terminal amino acid sequence analysis of purified IL-4
revealed that approximately half the protein lacked the
NH-terminal methionine. This NH-terminal
microheterogeneity may explain the two major bands of IL-4 revealed by
non-equilibrium pH gradient electrophoresis (Fig. 8). Since our
biological assay is not sensitive within a factor of 2, we are unable
to compare the relative potency of each of these species. A trace
amount (4%) of IL-4 lacks the first four amino acids (Table 2).
(Kleman et
al., 1990) have also been purified by refolding from inclusion
bodies.-terminal sequencing data
presented in Table 2. The preparation of IL-4 described here is a
mixture containing NH-terminal heterogeneity represented as
49% full-length protein, 47% des-Met IL-4, and 4% amino-terminal
proteolyzed protein. We propose that the three isoforms revealed by
two-dimensional gel electrophoresis (Fig. 8) are the three
amino-terminal variants, but have not matched each spot with its
corresponding sequence. It is formally possible that deglycosylated
natural IL-4 may also electrophorese as distinct isoforms and that the
variants identified in this report represent natural folding isoforms.
This possibility has not yet been addressed. One important technical
note of caution from this data is that the finding of a single,
symmetric, homogeneous peak on a reverse phase microbore column is no
guarantee that a single species of protein is present (contrast Fig. 8with Fig. 9). 10 units/mg) and dose-response curve as
native IL-4. When Park et al.(1987) derivatized the NH terminus of IL-4 synthesized in yeast to create an affinity
column for purification of the IL-4 receptor, the binding of receptor
to immobilized IL-4 was not altered. These data suggest that the amino
terminus of mouse IL-4 does not play a role in receptor binding or
signaling.
)
We thank Dan Conrad, William Paul, and David Tiemeier
for providing commentary and Pat Kallaos and Carla Campbell for
patiently typing this manuscript. Arnie Hershman provided encouragement
through this project. We thank Joseph W. Bulock who synthesized and
conjugated the peptides and Edwin Rowold for injecting and bleeding the
rabbits.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Himmelrich, P. Launois, I. Maillard, T. Biedermann, F. Tacchini-Cottier, R. M. Locksley, M. Rocken, and J. A. Louis In BALB/c Mice, IL-4 Production During the Initial Phase of Infection with Leishmania major Is Necessary and Sufficient to Instruct Th2 Cell Development Resulting in Progressive Disease J. Immunol., May 1, 2000; 164(9): 4819 - 4825. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |
