Protein Expression and Purification 18, 182–192 (2000)
`doi:10.1006/prep.1999.1179, available online at http://www.idealibrary.com on
`
`Optimization of Inclusion Body Solubilization and
`Renaturation of Recombinant Human Growth
`Hormone from Escherichia coli
`
`Ashok K. Patra, R. Mukhopadhyay, R. Mukhija,* Anuja Krishnan,*
`L. C. Garg,* and Amulya K. Panda1
`Product Development Cell, *Gene Regulation Laboratory, National Institute of Immunology,
`Aruna Asaf Ali Marg, New Delhi 110067, India
`
`Received September 24, 1999, and in revised form November 10, 1999
`
`Recombinant human growth hormone (r-hGH) was
`expressed in Escherichia coli as inclusion bodies. In
`10 h of fed-batch fermentation, 1.6 g/L of r-hGH was
`produced at a cell concentration of 25 g dry cell
`weight/L. Inclusion bodies from the cells were isolated
`and purified to homogeneity. Various buffers with and
`without reducing agents were used to solubilize r-hGH
`from the inclusion bodies and the extent of solubility
`was compared with that of 8 M urea as well as 6 M
`Gdn-HCl. Hydrophobic interactions as well as ionic
`interactions were found to be the dominant forces re-
`sponsible for the formation of r-hGH inclusion bodies
`during its high-level expression in E. coli. Complete
`solubilization of r-hGH inclusion bodies was observed
`in 100 mM Tris buffer at pH 12.5 containing 2 M urea.
`Solubilization of r-hGH inclusion bodies in the pres-
`ence of low concentrations of urea helped in retaining
`the existing native-like secondary structures of r-hGH,
`thus improving the yield of bioactive protein during
`refolding. Solubilized r-hGH in Tris buffer containing
`2 M urea was found to be less susceptible to aggrega-
`tion during buffer exchange and thus was refolded by
`simple dilution. The r-hGH was purified by use of
`DEAE-Sepharose ion-exchange chromatography and
`the pure monomeric r-hGH was finally obtained by
`using size-exclusion chromatography. The overall
`yield of the purified monomeric r-hGH was ;50% of the
`initial inclusion body proteins and was found to be
`biologically active in promoting growth of rat Nb2
`lymphoma cell lines.
`© 2000 Academic Press
`Key Words: recombinant human growth hormone;
`Escherichia coli;
`inclusion bodies; purification;
`bioactivity.
`
`1 To whom correspondence should be addressed at Product Devel-
`opment Cell, National Institute of Immunology, Aruna Asaf Ali
`Road, New Delhi 110067, India. E-mail: amulya@nii.res.in.
`
`182
`
`High-level expression of recombinant proteins in
`Escherichia coli often accumulates as insoluble aggre-
`gates in vivo as inclusion bodies (1). The formation of
`inclusion bodies is mainly attributed to the overexpres-
`sion of proteins in the cell lacking the required acces-
`sories for its folding to the native form (2). Endogenous
`proteins when overexpressed in E. coli also accumulate
`as inclusion bodies (3). There is no direct correlation
`between the propensity of the inclusion body formation
`of a certain protein and its intrinsic properties, such as
`molecular weight, hydrophobicity, and folding path-
`ways (4). In the case of proteins having disulfide bonds,
`formation of protein aggregation as inclusion bodies is
`anticipated since the reducing environment of bacte-
`rial cytosol inhibits the formation of disulfide bonds.
`Significant features of protein aggregates in inclusion
`bodies are the existence of native-like secondary struc-
`tures of the expressed protein (5) and their resistance
`to proteolytic degradation (6). The aggregation leading
`to inclusion body formation has also been reported to
`be due to specific intermolecular interactions among a
`single type of protein molecule (7). The formation of
`inclusion bodies thus facilitates the easy isolation and
`recovery of the expressed proteins in the denatured
`form.
`In general, proteins expressed as inclusion bodies
`are solubilized by the use of high concentrations of
`chaotropic solvents. Chaotropic agents such as urea,
`guanidine hydrochloride (Gdn-HCl)2, and thiocyanate
`salts (8,9), detergents such as sodium dodecyl sulfate
`
`2 Abbreviations used: r-hGH, recombinant human growth hor-
`mone; Gdn-HCl, guanidine hydrochloride; CTAB, n-cetyltrimethyl-
`ammonium bromide; NLS, sodium N-lauroyl sarcosine; BSA, bovine
`serum albumin; HS, horse serum; FBS, fetal bovine serum; BCA,
`bicinchoninic acid; IPTG, isopropyl-b-D-thiogalactopyranoside; CD,
`circular dichroism; SDS-PAGE, sodium dodecyl sulfate-polyacrylam-
`ide gel electrophoresis; DMSO, dimethyl sulfoxide; PEG, polyethyl-
`ene gycol; PMSF, phenylmethylsulfonyl fluoride.
`
`1046-5928/00 $35.00
`Copyright © 2000 by Academic Press
`All rights of reproduction in any form reserved.
`
`APOTEX EX1032
`
`Page 1
`
`

`
`RECOMBINANT HUMAN GROWTH HORMONE
`
`183
`
`(SDS) (10), N-cetyltrimethylammonium chloride (11)
`and sarkosyl (sodium N-lauroyl sarcosine; NLS) (12)
`along with reducing agents like b-mercaptoethanol,
`dithiothreitol, or cysteine have been extensively used
`for solubilizing the inclusion body proteins. The soluble
`proteins are then refolded to their native state after the
`chaotropic agents or other salts are removed by dialyz-
`ing the proteins in buffers containing reducing and
`oxidizing agents (8,13). Often, additives such as ace-
`tone, acetoamide, urea, DMSO, and PEG (14) are used
`to enhance the yield of folded bioactive protein. Rena-
`turation of proteins into the native conformation has
`also been reported by using immobilized minichaper-
`ones (15) and size-exclusion chromatography (16). De-
`spite several protocols available for protein solubiliza-
`tion and refolding, the overall recovery of bioactive
`proteins from inclusion body is often very low. It is
`expected that the overall yield of purified bioactive
`proteins from the inclusion bodies can be improved if
`the existing secondary structure of the proteins is pro-
`tected during solubilization without the use of high
`concentrations of chaotropic agents (17).
`Human growth hormone (hGH), a single chain
`polypeptide containing 191 amino acid residues, apart
`from stimulating cell growth, plays an important role
`in a variety of metabolic, physiologic, and anatomic
`processes (18). The protein folds into a four-helix bun-
`dle structure with two disulfide bridges, one connecting
`distant parts of the molecule involving amino acid res-
`idues 53 and 165 (large loop) and another between
`residues 182 and 189 (small loop) (19). The large-scale
`requirement of r-hGH necessitates its high-level ex-
`pression in E. coli as inclusion bodies (20). However,
`expression of the protein along with fusion tag and
`subsequent use of high concentrations of chaotropic
`reagents for solubilization and purification makes the
`overall process more complex and expensive as the
`yield of bioactive r-hGH is lowered (21). In this report,
`we have described a simple and efficient process for the
`production of bioactive r-hGH from the inclusion bodies
`of E. coli. The solubilization behavior of r-hGH inclu-
`sion bodies in different buffers was analyzed for an
`understanding of the nature of protein aggregation in
`inclusion bodies. Solubilization of r-hGH from purified
`inclusion bodies was carried out without disturbing the
`existing native-like secondary structure and the solu-
`bilized r-hGH was subsequently purified and refolded
`into the bioactive form.
`
`MATERIALS AND METHODS
`Chemicals
`Urea, acrylamide and bis-acrylamide, sodium dode-
`cyl sulfate, deoxycholate, and CTAB were of analytical
`grade, obtained from Amresco (U.S.A.). Sephacryl
`S-200 and DEAE-Sepharose were from Pharmacia Bio-
`tech (Sweden). NLS, BSA, prolactin, and RPMI 1640
`
`were from Sigma Chemicals (U.S.A.). Fetal bovine se-
`rum (FBS) and horse serum (HS) were from Gibco-BRL
`(U.S.A.). Commercially available recombinant human
`growth hormone was from Boehringer Mannheim
`(Germany). All other chemicals were of analytical
`grade. The spectral measurements and analytical
`HPLC were performed in degassed and filtered buffers
`prepared in Milli-Q water.
`
`Cloning and Expression of r-hGH
`To clone hGH without any tag and its signal se-
`quence, the cDNA fragment coding for hGH was ex-
`cised with HinfI-HindIII from pRMhGH (20). This ex-
`cised cDNA fragment was lacking the first 18 bp. The
`18 bp were chemically synthesized with HinfI over-
`hang at the 39-end and NcoI overhang at the 59-end
`[59-CATG TTC CCA ACT ATT CCA CTG-39; 39-AAG
`GGT TGA TAA GGT GAC TCA-59].
`This synthetic oligonucleotide and HinfI-HindIII-di-
`gested cDNA fragments were inserted into NcoI-Hin-
`dIII-digested pQE-60 expression vector
`(Qiagen,
`U.S.A.). The construct thus obtained has hGH with
`just one extra methionine at the N-terminus under the
`control of the phage T5 promoter. The NcoI site origi-
`nally present in pQE-60 was lost during construction of
`this plasmid. E. coli M15 cells containing the recombi-
`nant expression plasmid (pQE 60-hGH) were grown in
`LB or complex medium in the presence of kanamycin
`(25 mg/ml) and ampicillin (50 mg/ml). The cultures were
`induced with 1 mM IPTG and were further grown for
`4 h. Expression of r-hGH in the total cell extracts from
`both uninduced and induced cultures was checked by
`SDS-PAGE.
`
`Fermentation
`For large-scale production of r-hGH, recombinant E.
`coli cells were grown in a 3.5-L fermenter (2-L working
`volume) in complex medium. The composition of the
`complex medium was as described (22), except that the
`initial glucose and yeast extract concentrations were
`10 g/L. Fermentation was carried out at 37°C with
`vigorous aeration and agitation and the pH of the me-
`dium was maintained at 7 by use of 5 N NaOH. After
`3 h of batch growth, the cells were grown in a fed-batch
`mode with a continuous supply of glucose and yeast
`extract. Details of the fed-batch fermentation strategy
`are described elsewhere (23). The culture at OD600 (op-
`tical density) of 40 was induced with 1 mM IPTG,
`cultivated for another 4 h, and then harvested. The
`harvested cells were checked for expression and pro-
`cessed for purification of inclusion bodies. Samples
`were collected at regular intervals during the fermen-
`tation to check cell growth, glucose consumption, and
`r-hGH expression. Cell density was determined by
`measuring the OD of the culture at 600 nm with a
`Kontron (Kontron AG, Switzerland) UV-visible spec-
`
`Page 2
`
`

`
`184
`
`PATRA ET AL.
`
`trophotometer. Higher OD samples were diluted suit-
`ably to have an absorbance in the range of 0.2–0.6. Dry
`cell weight was determined by centrifuging the sample
`broth at 4000g for 20 min and drying the washed cell to
`constant weight at 105°C. One absorbance unit was
`equivalent to 0.35 g L21 of dry cell weight for uninduced
`culture and 0.4 g L21 dry cell weight for induced cul-
`ture. Residual glucose in the fermentation broth was
`measured by a glucose kit (Sigma), and the acetic acid
`concentration was monitored by an acetic acid kit
`(Boehringer Mannheim).
`
`Isolation, Purification, and Estimation of r-hGH from
`Inclusion Bodies
`Induced E. coli cells sampled at different time points
`during fed-batch fermentation were centrifuged at
`4000g for 30 min and the cell pellet was dissolved in 50
`mM Tris-HCl buffer (pH 8.0) containing 5 mM EDTA
`and 1 mM PMSF. Cells were lysed by sonication and
`centrifuged at 8000g for 30 min to isolate r-hGH inclu-
`sion bodies. For large-scale isolation and purification of
`r-hGH, induced E. coli cells were lysed by a French
`press at 18000 psi and the inclusion bodies were recov-
`ered by centrifugation at 8000g. The inclusion body
`pellets thus obtained were washed with 50 mM Tris-
`HCl buffer (pH 8.0) containing 5 mM EDTA and 2%
`deoxycholate (17). Finally, the inclusion bodies were
`washed with distilled water to remove contaminating
`salt and detergent and centrifuged at 8000g for 30 min
`and the pellet was used for estimation of r-hGH. At this
`stage,
`inclusion bodies
`contained mostly r-hGH
`(.90%), the majority in the form of monomer around
`21 kDa in SDS-PAGE along with some high molecular
`aggregates. The inclusion bodies were completely sol-
`uble in 50 mM Tris-HCl buffer (pH 8.5) containing 1%
`SDS and the solubilized r-hGH samples were diluted
`appropriately and estimated using BCA protein assay.
`SDS-PAGE was carried out using the method de-
`scribed by Laemmli (24).
`
`Solubilization of r-hGH from Inclusion Bodies
`Buffers of different pH, denaturants, combination of
`denaturant and salt, and ionic detergent as well as
`oxidizing and reducing agents were used to solubilize
`the r-hGH inclusion bodies. Purified r-hGH inclusion
`bodies were solubilized in 100 mM Tris buffer at dif-
`ferent pH (3–13) both in the presence and in the ab-
`sence of urea. Other solubilizing buffers such as 2 M
`Tris buffer at pH 12, 2 M Tris buffer at pH 12 contain-
`ing 2 M Urea, 2 M Tris buffer at pH 12 with reduced:
`oxidized glutathione (5:1 and 10:1 mM), 1% N-lauroyl
`sarcosine in Tris buffer at pH 8.5, 1% N-cetyltrimeth-
`ylammonium bromide in 50 mM Tris at pH 8.5, 8 M
`urea, 6 M guanidine hydrochloride, 1% SDS in 50 mM
`Tris at pH 8.5, and 100 mM Tris buffer at pH 12.5 with
`2 M urea along with reduced:oxidized glutathione
`
`(10:1) were used to soubilize r-hGH from the inclusion
`bodies. One hundred microliters of purified inclusion
`bodies (8 mg/ml) in 50 mM Tris buffer pH 8.5 was
`taken in different microcentrifuge tubes and centri-
`fuged. The supernatant was discarded and 1 ml of each
`of the above solubilizing buffers was added to the pel-
`lets. The suspension was vortexed and left for 30 min
`at room temperature and the turbidity of the solution
`was measured at 450 nm. The samples were again
`centrifuged at 12000 rpm for 10 min. The supernatant
`after filtering through a 0.45-mm Millipore filter was
`measured at 280 nm for an estimation of protein con-
`tent. The same protocol was followed to measure the
`effect of pH and urea on inclusion body solubilization.
`To dissociate the r-hGH oligomers present in inclu-
`sion bodies into monomers, different concentrations of
`b-mercaptoethanol (2 to 20 mM) in 100 mM Tris buffer
`at pH 12.5 containing 2 M urea were used. Solubiliza-
`tion of inclusion bodies in 8 M urea with different
`concentrations of b-mercaptoethanol (100–200 mM)
`was also tried to dissociate the oligomers into the
`monomeric form during solubilization. The extent of
`solubilization in these buffers was determined as de-
`scribed above.
`
`Purification of r-hGH
`For large-scale purification of r-hGH, pure inclusion
`bodies were isolated (;104 mg protein isolated from 65
`ml of high cell density fermentation broth) and solubi-
`lized in 16 ml of 100 mM Tris buffer, pH 12.5, contain-
`ing 2 M urea. The solubilized r-hGH was diluted five
`times with Milli-Q water and the pH was brought down
`to 8.5 by adding 1 N HCl. A DEAE-Sepharose ion-
`exchange column (5 3 5 cm) equilibrated with 20 mM
`Tris buffer containing 5 mM EDTA, 0.4 M urea, and
`0.02% sodium azide, pH 8.5, was used for the purifica-
`tion of r-hGH. The solution was loaded to an ion-ex-
`change column at a flow rate of 0.5 ml/min. The column
`was washed with 5 column volumes of equilibration
`buffer followed by 3 column volumes of equilibration
`buffer containing 0.1 M NaCl. Recombinant hGH was
`eluted with a gradient of 0.1–0.25 M NaCl in the equil-
`ibration buffer at a flow rate of 1.5 ml/min. The absor-
`bance was measured on line with a Pharmacia UV
`monitor. Peak fractions showing r-hGH were pooled,
`checked in SDS-PAGE, and lyophilized. The lyophi-
`lized r-hGH solubilized in 5 ml of 20 mM Tris buffer
`containing 5 mM EDTA, 0.02% (w/v) sodium azide, and
`1 mM PMSF was loaded onto a Sephacryl S-200 col-
`umn (bed volume 5 90 3 1.6 cm, flow rate of 20 ml/h)
`for further purification. The fractions which showed a
`single r-hGH protein band on SDS-PAGE were pooled
`together, dialyzed against 20 mM Tris buffer at pH 8.5,
`and lyophilized. Lyophilized r-hGH was used for phys-
`icochemical and bioactivity assays and were stored at
`220°C for further use. HPLC analysis of the pure
`
`Page 3
`
`

`
`RECOMBINANT HUMAN GROWTH HORMONE
`
`185
`
`refolded r-hGH was carried out through a gel-filtration
`Shodex (Protein KW-804, Waters) column. The flow
`rate of the solvent was 1 ml/min. The absorbance was
`monitored at 280 nm, using a Waters HPLC-UV Model
`490 detector.
`
`Spectroscopic Analysis
`A circular dichroism spectrum was obtained at 25°C
`in the wavelength range of 190–250 nm using a
`JASCO-Spectropolarimeter in 20 mM Tris buffer. The
`sample was scanned 10 times for data accumulation
`and the average spectrum was plotted. Similarly, the
`UV spectrum of r-hGH was scanned within the wave-
`length range 240 to 350 nm (Contron UV-Vis spectro-
`photometer). Fluorescence emission spectrum was
`taken by exciting the protein molecules at 280 nm and
`measuring the emission in the wavelength region from
`320 to 350 nm using a Schimadzu spectrofluoropho-
`tometer (Model 1501).
`
`Estimation of Extinction Coefficient
`The purified and lyophilized r-hGH was used for
`measurement of extinction coefficient. Prior to the
`measurement, r-hGH was dialyzed extensively against
`Milli-Q water. Absorbance of water-dialyzed r-hGH
`was recorded and 1 ml of the protein was dried at
`105°C. The dried fraction was weighed and the molar
`extinction coefficient of r-hGH was derived from the
`formula
`
`FIG. 1. Fed-batch fermentation kinetics of E. coli for the produc-
`tion of r-hGH. Cells were grown in batch mode until the initial
`glucose was consumed. After 3 h, glucose feeding was started at a
`predetermined rate for fed-batch growth of cells. At OD600 of 40, the
`cells were induced with 1 mM IPTG. Samples were taken at intervals
`to monitor glucose (E), cell OD at 600 nm ((cid:130)), acetic acid (h), and
`r-hGH concentration (F).
`
`A 5 ebC,
`
`carried out in triplicate under atmospheric conditions
`with 5% CO2 at 37°C.
`
`where A is the absorbance at 280 nm, b is the path
`length of incident light in cm, C is the protein concen-
`tration in moles per liter, and e is the molar extinction
`coefficient.
`
`Bioactivity Assay
`The biological activity of r-hGH was determined by
`its growth-promoting action on rat Nb2 lymphoma cell
`lines. Commercially available recombinant human
`growth hormone form Boehringer Mannheim was used
`as standard. The Nb2 cell lines were maintained in
`RPMI medium supplemented with 10% FBS and 10%
`HS. Quiescent Nb2 cells arrested at the G0/G1 phases
`were prepared by incubating cells in RPMI supple-
`mented with 1% FBS and 10% HS. To initiate cellular
`proliferation, different concentrations (1–25 ng/ml)
`each of BSA, commercial hGH, or r-hGH were added to
`the culture medium. The assay was set in a 96-well flat
`bottom culture plates using RPMI as control. Growth-
`promoting activity was evaluated by counting the num-
`ber of cells every 24 h for 5 days. Experiments were
`
`RESULTS
`Fed-Batch Fermentation and Isolation of r-hGH
`Inclusion Bodies
`E. coli cells expressing r-hGH were grown in a fed-
`batch fermentation process to produce large quantities
`of r-hGH. The cultures at cell OD of 40 (20 g/L dry cell
`weight) were induced with 1 mM IPTG (optimum IPTG
`for induction was 0.01 mM/L/ OD culture) and grown
`for another 4 h, and the batch was terminated at a cell
`OD of 60 (Fig. 1). A maximum of 1.6 g/L of r-hGH was
`expressed as inclusion bodies in 10 h of fed-batch fer-
`mentation. The expression of r-hGH plateaued after
`3 h of IPTG induction (Fig. 2) and the level of r-hGH
`expression was around 13% of the total cellular pro-
`tein. The specific cellular r-hGH yield was 26.6 mg/L/
`OD. The residual glucose concentration was main-
`tained around 0.5 to 1 g/L throughout the fed-batch
`operation and the acetic acid accumulation was also
`very low (,2.5 g/L). Inclusion bodies of r-hGH from E.
`coli cells were isolated and purified as described earlier
`for ovine growth hormone (17). Extensive washing with
`
`Page 4
`
`

`
`186
`
`PATRA ET AL.
`
`carried out to enhance the solubility of r-hGH from the
`inclusion bodies. Higher solubilization of r-hGH from
`inclusion bodies was observed by incorporating 2 M
`urea in 100 mM Tris buffer at pH 12.5 (Fig. 3B). Fur-
`ther addition of urea in 100 mM Tris buffer at pH 12.5
`did not help in solubilizing higher amounts of r-hGH
`
`FIG. 2. SDS-PAGE analysis of r-hGH expression during fed-batch
`fermentation. Lanes 1 and 2, uninduced cells. Lanes 3–6, induced
`cells after 0.5, 1, 2, and 3 h ofIPTG induction, respectively. Lane 7,
`MW markers in kDa. Lane 8, pure r-hGH inclusion bodies isolated
`from cells after extensive washing.
`
`detergents resulted in pure inclusion bodies containing
`.90% monomeric r-hGH (Fig. 2). Washing of the inclu-
`sion body preparation in deoxycholate-containing
`buffer helped in removing the majority of the contam-
`inants. Dimers as well as multimers of r-hGH were
`also present in pure inclusion body preparations. For-
`mation of dimers was also observed at the early stages
`of recombinant protein synthesis. Most of the purified
`inclusion body proteins visualized in SDS-PAGE re-
`acted with polyclonal hGH antibody (data not shown),
`indicating the purity of the preparation. As this low
`level of contaminants has little effect on the refolding
`of proteins, the purified inclusion bodies containing
`mostly r-hGH were used for subsequent solubilization
`and refolding.
`
`Solubilization of r-hGH from Inclusion Bodies
`It
`Solubilization of inclusion bodies at different pH.
`has been widely reported that growth hormone inclu-
`sion bodies of different species expressed in E. coli can
`be solubilized by alkaline pH (17,21,25). Thus, the pu-
`rified r-hGH inclusion bodies were solubilized at differ-
`ent pH in 100 mM Tris buffer (pH 3–13) and percent-
`age solubilization of r-hGH was monitored (Fig. 3A).
`Solubilization of r-hGH from inclusion bodies was ob-
`served by increasing the pH from 8 to 12.5. High alka-
`line pH (.12.5), even though it helped in solubilizing
`r-hGH from inclusion bodies, resulted in extensive deg-
`radation of r-hGH (SDS-PAGE data not shown). A
`maximum of 2 mg/ml of r-hGH was solubilized in 100
`mM Tris buffer at pH 12.5 without the addition of urea
`or guanidine hydrochloride.
`Effect of urea and b-mercaptoethanol. Solubiliza-
`tion experiments in 100 mM Tris buffer at pH 12.5
`containing different molar concentrations of urea were
`
`FIG. 3.
`(A) Effect of pH on the solubility of r-hGH inclusion bodies.
`A constant amount of r-hGH inclusion bodies was solubilized at
`different pH in 100 mM Tris buffer and the turbidity was measured
`at 450 nm. The sample solution after centrifugation and filtration
`was used for estimation of protein by taking the absorbance at 280
`nm. (B) Effect of urea concentration on the solubility of r-hGH
`inclusion bodies. Different concentrations of urea were added to 100
`mM Tris buffer, pH 12.5, and solubilization behavior was monitored
`by estimating the amount of protein at 280 nm.
`
`Page 5
`
`

`
`RECOMBINANT HUMAN GROWTH HORMONE
`
`187
`
`from the inclusion bodies. In 100 mM Tris buffer at pH
`12.5 containing 2 M urea, a maximum of 6 mg/ml of
`r-hGH was solubilized from the inclusion bodies. Sol-
`ubility of r-hGH was comparable to that of 8 M urea in
`Tris buffer at pH 8. Solubilized r-hGH in 100 mM Tris
`buffer at pH 12.5 containing 2 M urea was analyzed in
`CD and fluorescence spectra and were found to have
`native-like secondary structures (data not shown).
`Isolation of r-hGH and their subsequent solubiliza-
`tion at alkaline pH were always associated with the
`presence of dimer (44 kDa, in SDS-PAGE), which con-
`stituted around 5–8% of the total inclusion body pro-
`tein. These r-hGH dimers were also observed in the
`SDS-PAGE despite the high reducing and denaturing
`environment. Addition of increasing amounts of b-mer-
`captoethanol (2–20 mM ) in 100 mM Tris buffer at pH
`12.5 containing 2 M urea had very little effect on dis-
`sociating oligomers into monomeric r-hGH. Dissocia-
`tion of oligomers and dimers to monomeric r-hGH was
`also not observed in 8 M urea solution containing 100
`to 200 mM b-mercaptoethanol (SDS-PAGE data not
`shown).
`Effect of different solvents. The solubilizing effect of
`100 mM Tris, pH 12.5, buffer containing 2 M urea on
`r-hGH inclusion bodies was compared with different
`solubilizing buffers as described under Materials and
`Methods. It was observed that the solubility achieved
`in 8 M urea, 6 M Gdn-HCl, 1% SDS, and 1% CTAB in
`50 mM Tris buffer at pH 8.5 was comparable to that of
`100 mM Tris buffer containing 2 M urea at pH 12.5
`(Fig. 4). Other buffers used for solubilization were not
`as effective as 100 mM Tris buffer at pH 12.5 contain-
`ing 2 M urea. The solubility of the r-hGH inclusion
`bodies in 1% NLS in 50 mM Tris at pH 8.5 was also
`very low. The solubility of r-hGH from inclusion bodies
`in 2 M Tris containing 2 M urea was lower than 100
`mM Tris buffer containing 2 M urea at pH 12.5. The
`presence of reduced/oxidized glutathione helped to in-
`crease the solubility of r-hGH from inclusion bodies in
`2 M Tris buffer containing 2 M urea at pH 12 but had
`little effect when used in 100 mM Tris buffer at pH 12.5
`with 2 M urea. In 100 mM Tris buffer at pH 12.5
`containing 2 M urea, 6.5 mg/ml of r-hGH was solubi-
`lized from the inclusion bodies. As observed earlier,
`none of these buffers could dissociate r-hGH dimers/
`oligomers present in the inclusion body preparation
`into monomers.
`
`Purification of r-hGH
`Pure r-hGH inclusion bodies solubilized in 100 mM
`Tris buffer at pH 12.5 containing 2 M urea were fur-
`ther purified using ion-exchange and gel-filtration
`chromatography. The solubilized r-hGH was diluted
`five times and the pH of the buffer was adjusted to 8.5.
`No aggregation of the solubilized r-hGH was observed
`during dilution and buffer exchange. Solubilized
`
`FIG. 4. Effect of different solvents on the solubility of r-hGH inclu-
`sion bodies. A fixed amount of r-hGH inclusion bodies was used and
`the turbidity and solubility were measured at 450 and 280 nm,
`respectively. (A) 2 M Tris buffer at pH 12. (B) 2 M Tris buffer at pH
`12 with 2 M urea. (C) 2 M Tris buffer at pH 12 with 2 M urea and R:O
`glutathione (5:1). (D) 2 M Tris buffer at pH 12 with 2 M urea and R:O
`glutathione (10:1). (E) 1% SDS in 50 mM Tris buffer at pH 8.5. (F) 8
`M urea at pH 8.5. (G) 100 mM Tris buffer at pH 12.5. (H) 100 mM
`Tris buffer at pH 12.5 with 2 M urea. (I) 100 mM Tris buffer at pH
`12.5 with 2 M urea and R:O glutathione (10:1). (J) 50 mM Tris buffer
`at pH 8.5. (K) 1% CTAB in Tris buffer at pH 8.5. (L) 1% NLS in Tris
`buffer at pH 8.5.
`
`r-hGH were passed through a DEAE-Sepharose col-
`umn for purification. Recombinant human growth hor-
`mone which eluted between the conductivity range of
`14 to 16 mS/cm (Fractions 20 to 28) was found to be
`homogeneous and was 40% of the total protein. How-
`ever, some amount of r-hGH was coeluted along with
`r-hGH dimer between conductivities of 22 to 25 mS/cm
`(Fractions 35 to 45) which constituted about 25–30% of
`the total protein. The overall recovery of r-hGH from
`ion-exchange matrix was around 65%. Pure r-hGH con-
`taining dimers/oligomers was passed through the size-
`exclusion chromatography column for further purifica-
`tion. The dimeric or higher forms of the proteins were
`removed through gel filtration (Fig. 5A). The overall
`yield of the purified refolded r-hGH from the inclusion
`bodies of E. coli was ;50% (Table 1). HPLC analysis of
`the purified r-hGH showed a single peak at 11.38 min
`(Fig. 5B) and SDS-PAGE showed a single band (Fig.
`5C), thus indicating the high purity of the r-hGH.
`
`Characterization and Bioactivity
`Authenticity of the purified r-hGH was further con-
`firmed from the N-terminal analysis of r-hGH and from
`
`Page 6
`
`

`
`188
`
`PATRA ET AL.
`
`FIG. 5.
`(A) Final purification of r-hGH through gel-filtration chromatography. Lyophilized powder of r-hGH was solubilized in 5 ml of buffer
`and loaded onto a gel-filtration column. Fractions of 3.5 ml were collected and the peaks were pooled separately. The majority of the r-hGH
`eluted as a monomeric form between Fractions 30 and 40. (B) Monomeric pure r-hGH analyzed on a Shodex Protein KW-804 HPLC column
`at 280 nm showing a single peak. (C) SDS-PAGE analysis of purified r-hGH eluted from S-200 gel-filtration column. Lane 1, pure r-hGH.
`Lane 2, molecular weight markers.
`
`spectroscopic analysis. UV spectrum of the purified
`r-hGH showed an absorbance maxima at 276.8 nm,
`and a shoulder at 283 nm, which was comparable to
`that of native human growth hormone reported by
`Bewley and Li, in 1984 (29) (Fig. 6A). The fluorescence
`spectrum of refolded r-hGH was found to be identical to
`
`the native hGH which gave a peak at 340 nm (Fig.
`6B).The molar extinction coefficient of pure r-hGH was
`found to be 18,800 M21 cm21 which is very close to the
`reported value of 18,890 M21 cm21 for native hGH (26).
`Growth kinetics of the prolactin-dependent Nb2 lym-
`phoma cell line was monitored to evaluate the bioac-
`
`Page 7
`
`

`
`RECOMBINANT HUMAN GROWTH HORMONE
`
`189
`
`TABLE 1
`Purification of r-hGH from High Cell Density Culturea
`
`Steps
`
`Total protein
`(mg)
`
`Step yield
`(%)
`
`Overall yield
`(%)
`
`Purityb
`(%)
`
`Cell lysate
`Pure inclusion body
`Solubilization
`Ion-exchange chromatography
`Gel-filtration chromatography
`
`800
`104
`98
`67
`52
`
`—
`100
`94
`71
`77
`
`—
`100
`94
`64
`50
`
`13
`90
`92
`95
`99
`
`a Sixty-five milliliters of high density culture at OD600 nm 5 60 contained about 1.6 g dry cell weight. Step and overall yields were calculated
`starting from the pure inclusion body preparation.
`b Purity is defined as the percentage of monomeric r-hGH in the purified protein preparation.
`
`tivity of purified r-hGH. The addition of prolactin, com-
`mercial hGH, and r-hGH promoted growth of Nb2 cells
`arrested at G0/G1 phases by serum deprivation.
`Growth of Nb2 cells in the presence of different con-
`centrations of r-hGH was found to be comparable to
`that observed for the commercial hGH (Fig. 7). No
`growth stimulation was observed in the presence of
`BSA which was used as a negative control.
`
`DISCUSSION
`Fed-batch fermentation of E. coli resulted in high
`volumetric yield of r-hGH. In 10 h of fed-batch fermen-
`tation, 1.6 g/L of r-hGH was produced in comparison to
`30–40 mg/L expressed in shaker flask culture. The
`r-hGH accumulated as inclusion bodies in E. coli and
`was isolated to more than 90% purity by extensive
`washing in 2% deoxycholate in 50 mM Tris buffer, pH
`8.5. However, among the last 5–10%, the dimers and
`oligomers of r-hGH were the major contaminants. The
`purity and homogeneity of the r-hGH inclusion body
`preparation were in agreement with the proposed com-
`position of the inclusion bodies that they are formed
`due to specific aggregation of single protein intermedi-
`ates (4,7). As the inclusion bodies consisted of mostly
`r-hGH, solubilization and refolding were carried out
`before further purification.
`Inclusion body proteins have been reported to have
`extensive native-like secondary structure (5,27) and
`thus when solubilized while retaining the native-like
`secondary structure results in high recovery of the
`bioactive protein (17). In order to protect the native-
`like secondary structure, r-hGH inclusion bodies were
`solubilized at alkaline pH, containing mild concentra-
`tions of chaotropic agents (2 M urea). Charge distribu-
`tion provided by high alkaline pH along the protein
`chain was responsible for higher solubilization of
`r-hGH from inclusion bodies. This suggested that pH
`has a crucial role in destabilizing the inclusion body
`aggregation. Changing the charge distribution along
`the protein molecule by changing the pH generally
`affects the protein stability and may lead to an unfold-
`
`ing of the native protein (28). However, for human
`growth hormone, there did not seem to be any indica-
`tions of general unfolding of the secondary structure
`(29,30) and the pH-induced unfolding has been re-
`ported to be reversible for a similar protein [bovine
`growth hormone (31)]. Use of 2 M urea at alkaline pH
`improved r-hGH solubilization from inclusion bodies
`without disturbing the existing native-like secondary
`structure of the proteins. Although diminishing hydro-
`phobic interaction between water and protein molecule
`is a linear function of urea concentration (32), this
`effect also existed substantially at as low as 2 M urea
`concentration. Since proteins could not be denatured at
`such a low concentration of urea, urea was probably
`only serving the purpose of physical separation of the
`molecules by disrupting the hydrophobic interactions.
`Thus in conclusion, r-hGH was solubilized from the
`inclusion bodies without its existing native-like sec-
`ondary structure being disturbed. The extent of solu-
`bilization in 100 mM Tris buffer, pH 12.5, containing 2
`M urea was found to be comparable with that of 8 M
`urea and a maximum amount of 6.5 mg/ml of r-hGH
`could be solubilized from the inclusion bodies. In-
`creased solubility of r-hGH in the above buffer could be
`an effect of both urea and pH, indicating the existence
`both ionic and hydrophobic interactions in the inclu-
`sion bodies.
`Expression of r-hGH in E. coli as inclusion bodies
`was always associated with the formation of dimers
`and oligomers which were also observed at the early
`s