Expression system for recombinant human growth hormone production from Bacillus subtilis

Authors


Abstract

We demonstrate for the first time, an expression system mimicking serine alkaline protease synthesis and secretion, producing native form of human growth hormone (hGH) from Bacillus subtilis. A hybrid-gene of two DNA fragments, i.e., signal (pre-) DNA sequence of B. licheniformis serine alkaline protease gene (subC) and cDNA encoding hGH, were cloned into pMK4 and expressed under deg-promoter in B. subtilis. Recombinant-hGH (rhGH) produced by B. subtilis carrying pMK4::pre(subC)::hGH was secreted. N-terminal sequence and mass spectrometry analyses of rhGH confirm the mature hGH sequence, and indicate that the signal peptide was properly processed by B. subtilis signal-peptidase. The highest rhGH concentration was obtained at t = 32 h as CrhGH = 70 mg L−1 with a product yield on substrate YrhGH/S = 9 g kg−1, in a glucose based defined medium. Fermentation characteristics and influence of hGH gene on the rhGH production were investigated by comparing B. subtilis carrying pMK4::pre(subC)::hGH with that of carrying merely pMK4. Excreted organic-acid concentrations were higher by B. subtilis carrying pMK4::pre(subC)::hGH, whereas excreted amino-acid concentrations were higher by B. subtilis carrying pMK4. The approach developed is expected to be applicable to the design of expression systems for heterologous protein production from Bacillus species. © 2009 American Institute of Chemical Engineers Biotechnol. Prog., 2009

Introduction

Human growth hormone (hGH) is anionic, nonglycosylated four helix-bundle protein known as somatotropin, having a molar mass of 22 kDa and 191 amino acid residues. It has been used to treat hypopituitary dwarfism, injuries, bone fractures, bleeding ulcers, and burns.1 Recently, it appears to be of considerable benefit to girls with Turner's syndrome, children with chronic renal failure, and adults with growth hormone deficiency or human immunodeficiency virus (HIV) syndrome.2

Bacillus species, producers of several industrial enzymes, are potential hosts for production of heterologous nonglycosylated proteins of commercial interest. The advantages of using the Gram positive bacteria, besides the ability to secrete functional extracellular proteins directly into the bioreactor culture medium, are the lack of pathogenicity and the absence of lipopolysaccharides (endotoxins) from the cell wall.3 Amongst, Bacillus subtilis has become a model system for the study of many aspects of the biochemistry, genetics, and physiology of Gram-positive bacteria, and particularly of sporulation and associated metabolism4; whereupon, information concerning its secretion mechanism has been gathered as the genome sequence was resolved.5 Nevertheless, the secretion of heterologous recombinant proteins from the bacilli might be inefficient. On the basis of the growing availability of information on genomics and proteomics of B. subtilis, difficulties can now be systematically addressed and overcome.6 For the secretion of a recombinant protein produced, either protease deficient Bacillus cells7 or protease inhibitors are used. Westers et al.,8 in their review article, summarised the efforts employed to improve B. subtilis as a host for protein secretion. Expression and secretion of nonglycosylated proteins in the genus Bacillus require the assistance of the N-terminal signal-sequence of precursors. Brockmeier et al.9 and Fu et al.10 reported the use of various promoters and signal DNA sequences for recombinant protein production by B. subtilis.

Extracellular production of a recombinant foreign protein from a B. subtilis host requires a neat design and engineering of an expression and secretion system; wherein, the choice of the promoter and signal DNA sequence in combination with the DNA vector, in particular, is important. For recombinant protein production using Bacillus species, there is no work in the literature reporting on the use of the promoter and signal sequence of the DNA encoding the industrial enzyme serine alkaline protease (SAP).

The idea in this work is based on the construction of a recombinant plasmid for the synthesis and secretion of rhGH that mimics the synthesis of SAP in bacilli. Thus, the hybrid-gene of two DNA fragments, i.e., signal (pre-) DNA sequence of B. licheniformis serine alkaline protease (SAP) gene (subC) and cDNA encoding hGH, were cloned into pMK4 plasmid and expressed under the deg-promoter in a B. subtilis host. Production of rhGH from B. subtilis and the fermentation characteristics in a defined medium were investigated, using the designed hybrid-gene system.

Experimental Methods

Bacterial strains, plasmids, and growth media for genetic manipulation

The strains, plasmids, and primers used in this study are described in Table 1. Bacterial strains, plasmids, and growth media were prepared using standard techniques.15B. licheniformis (DSM 1969), B. subtilis, and Escherichia coli XL1-Blue16 were maintained and grown on LB-agar that contained (g L−1): tryptone, 10; NaCl, 5; yeast extract, 5; agar, 15 and in LB broth (without agar) at 37°C. Ampicillin (100 μg/mL) was used for the plasmid maintenance in E. coli strains; 7 μg/mL chloramphenicol was used for plasmid maintenance in the recombinant B. subtilis.

Table 1. Strains, Plasmids and Primers used in this Study
NameDescriptionReference or Source
Strain
 Bacillus licheniformisWild type carrying subC geneDSM 1969 (11)
 B. subtilisnpraprBGSC- 1A751
 B. subtilisspoBGSC- 1A179
 Escherichia coli XL1Blue  
Plasmids
 pHGH107 (12)
 pUC19 (13)
 pMK4 (14)
 pUC19::pre(subC)::hGH This work
 pMK4::pre(subC)::hGH This work
Primers for pre(subC)::hGH
 pre(subC) forward primer5′_GCT CTA GAG CGC AAT CTC CTG TCA TTC G_3′ 
 Complimentary strand to hGH + pre(subC) reverse5′_GGT ATA GTT GGG AAA GCA GAA GCG GAA TCG_3′ 
 Complimentary strand to pre(subC) + hGH forward primer5′_GCT TCT GCT TTC CCA ACT ATA CCA CTA TCT C_3′ 
  hGH reverse primer5′_GCG GAT CCG CAC TGG GGA GGG GTC AC_3′ 

Manipulation of DNA, PCR, cloning, and DNA sequencing

B. licheniformis chromosomal DNA was isolated as described by Posprech and Neumann.17subC gene (GenBank Acc. No. X03341)18, 19 that encodes for extracellular serine alkaline protease (SAP) enzyme of B. licheniformis was used as the template for amplification of signal (pre-) sequence. HGH cDNA16 was amplified from E. coli host strain carrying the plasmid pHGH107 (ATCC 31538; US patent no. 4,342,832), featuring the growth hormone ORF (NCBI accession number A00501) from Homo sapiens and antibiotic resistance genes to ampicillin and tetracycline.

The primers used for the amplification are given in Table 1. Signal DNA sequence of subC was fused in front of the hGH gene using gene splicing by overlap extension method.20XbaI restriction site was incorporated to the forward primer of pre(subC) sequence, whereas BamHI restriction site was incorporated to the reverse primer of hGH gene. To verify the cloning, nucleotide sequencing analyses were performed at Microsynth GmbH (Switzerland) using the designed primers.

Culture maintenance and media for fermentation

For the bioprocess experiments, B. subtilis BGSC-1A751 (nprapr−) and B. subtilis BGSC-1A197 (spo) stock cultures were maintained on agar slants that contained (g L−1): peptone, 5; beef extract, 3; agar, 15; and initial pH = 7.25. The cells on the newly prepared slants were inoculated into the preculture medium for preparation of inocula that contained (g L−1): soytryptone, 15; peptone, 5; MnSO4.2H2O, 0.010; Na2HPO4, 0.25; CaCl2, 0.100 and grown at 37°C for 6 h. The defined reference production medium for batch-bioreactor was as follows (g L−1): glucose, 6.0; (NH4)2HPO4, 4.7; KH2PO4, 2.0; 0.04 M Na2HPO4 and NaH2PO4; the initial pH = 7.25.11, 21 Chloramphenicol (7 μg/mL) was used in all bioprocess experiments of plasmid-bearing B. subtilis. Complete EDTA-free protease inhibitor (Roche) was used to prevent proteolytic hydrolysis of the produced rhGH.

Laboratory-scale batch fermentations

Batch laboratory-scale fermentation experiments were conducted in orbital shakers under agitation and heating rate control, using air-filtered 500 mL Erlenmeyer-flasks having 220 mL working volume capacities. Batch-bioreactor experiments were conducted in 1.0 L bioreactor systems (BBraun, Germany) consisted of temperature, pH, foam, air inlet, and stirring rate controls with 0.5 L working volume. Each experiment was conducted in two bioreactors operating in parallel, to check reproducibility.

Analyses

Cell concentrations based on dry weights were measured with a UV-vis spectrophotometer (Shimadzu UV-160A, Tokyo, Japan) using a calibration curve obtained at 600 nm. Glucose consumption was followed by the glucose oxidation method at 505 nm with UV-vis spectrophotometer.22 Excreted amino acid concentrations were measured with an amino acid analysis system (Waters HPLC, Milford, MA), using the Pico Tag method.23 Excreted organic acid concentrations were measured with an HPLC (Waters, HPLC, Alliance 2695).24 HGH concentrations were measured using a high-performance capillary electrophoresis (Waters HPCE, Quanta 4000 E, Milford, MA). The samples were analyzed at 12 kV and 15°C with a positive power supply using 60 cm × 75 μm silica capillary using modified 100 mM borate buffer (pH = 10) including zwitterions (Z1-Methyl, Waters) as the separation buffer. Proteins were detected by UV absorbance at 214 nm, as mentioned elsewhere.23, 25 Humatrope (Eli Lilly, France) was used as the standard. The Dynamic method26 was applied to find the oxygen uptake rate (OUR) and oxygen transfer coefficient (KLa) values.

The physiological data for each operation were from at least two independent experiments, and the average values were given.

Ultrafiltration and purification

Concentration and desalting of the production medium was achieved by ultrafiltration under nitrogen gas (55 psi, 3.8 bar) at 4°C using Amicon 400 mL stirred pressure cells (Millipore, Bedford, MA) with regenerated cellulose ultrafiltration membranes having MWCO of 10 kDa (Millipore, Bedford, MA). Purification of rhGH was achieved by aptamer-based affinity chromatography. Concentrated samples were mixed with hGH specific aptamer which was immobilized onto microparticles and hGH-aptamer binding was carried out at 25°C for 30 min, which has been developed and is being studied in our research group.

SDS-PAGE, Western Blotting, and N-terminal sequence analysis

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed as described by Laemmli27 by using 4% stacking and 12% separating polyacrylamide gel, run on a Mini Protean II DUAL SLAB cell (Bio-Rad) according to the manufacturer's instructions and silver stained. For Western blot analysis, polyclonal rabbit anti-human growth hormone (BioMeda, USA) was used as the primary antibody and horseradish peroxidase labeled goat-anti rabbit IgG (H+L) (BioMeda) was used as the secondary antibody. For the N-terminal analysis, rhGH was electrophoresed as described above and transferred onto a polyvinylidene difluoride membrane (Millipore, USA). After being stained with Coomassie blue, the rhGH band was excised, and automated Edman degradation was performed by PROCISE 494 gas-phase/liquid-pulse sequencer (Applied Biosystems, Foster City, CA).

MALDI-ToF mass spectrometry analysis

The molecular weight of rhGH was determined by the use of a MALDI-LR (Waters-Micromass, UK) instrument. Spectra were generated using a pulsed nitrogen gas laser (337 nm) in positive linear mode with a low mass gate of 1,000 Da.25 The accelerating voltage was 15 kV. Three microliters of a 10 mg mL−1 sinapinic acid matrix dissolved in 50% acetonitrile and 0.1% TFA solution, was mixed with 1 μL of ∼10 pmol μL−1 sample and 1 μL of this mixture was spotted on the target plate and air-dried by the “dried droplet” technique.28 Cytochrome c and Humatrope (standard hGH) were used as molecular weight standards for purposes of mass correction. Spectra were generated from the sum of 100–200 laser pulses and mass determinations were made by finding the peak centroid of a smoothed signal (by Savitzky-Golay algorithm) after background subtraction.29

Results

Construction of the plasmid pMK4::pre(subC)::hGH

B.licheniformis (DSM 1969) chromosomal DNA and pHGH107 plasmid, containing the hGH cDNA, were isolated to be used as templates in PCR reactions. The two target genes, pre(subC) of subC gene (360 bp) from B. licheniformis chromosomal DNA and mature peptide sequence of hGH (639 bp) from pHGH107 plasmid, were amplified by PCRs (Figure 1). The primers (Table 1) used at the ends to be joined were designed as complementary to one another by including nucleotides at their 5′ ends that are complementary to the 3′ portion of the other primer. The two PCR products containing the overlapping fragments at the ends to be joined were purified and, by the third PCR reaction using the external primers carrying XbaI and BamHI restriction sites, extension of the overlap by DNA polymerase has yielded the hybrid-gene product, i.e., pre(subC)::hGH (999 bp), where pre(subC) DNA sequence, was fused in front of the hGH sequence (Figure 1). The hybrid-gene pre(subC)::hGH was then cloned into the XbaI and BamHI sites of pUC19 E. coli plasmid, and transformed into E. coli XLI-Blue cells by CaCl2 method. Thereafter, pre(subC)::hGH was sub-cloned to pMK4 SalI and BamHI sites and expressed in the hosts B. subtilis BGSC-1A751 (nprapr−) and B. subtilis BGSC-1A197 (spo).

Figure 1.

Agarose gel electrophoresis view for PCR amplification of hGH, pre(subC), and pre(subC)::hGH.

M, low range marker (Fermantas); Lane 1, hGH; Lane 2, pre (subC)::hGH; and Lane 3, pre(subC).

SDS-PAGE, Western Blot, N-terminal, and mass spectrometry analyses

RhGH production potential of the recombinant cells, B. subtilis BGSC-1A751 (nprapr) carrying pMK4::pre(subC)::hGH and B. subtilis BGSC-1A197 (spo) carrying pMK4::pre (subC)::hGH were determined on a glucose (CG = 6 g L−1) based defined medium. The supernatant obtained by centrifugation at t = 27 h of the fermentation was partially purified by dead-end ultrafiltration and nearly 10-fold concentration was achieved. Western blot analysis showed that (Figure 2), the molecular mass of rhGH produced by B. subtilis BGSC-1A751 carrying pMK4::pre(subC)::hGH and B. subtilis BGSC-1A197 carrying pMK4::pre(subC)::hGH was 22 kDa being the same as the standard hGH (Humatrope, Eli Lilly, France).

Figure 2.

Western Blott analysis results of BGSC-1A751 (npr, apr) carrying pre(subC)::hGH and BGSC-1A197 (spo) carrying pre(subC)::hGH.

Lane 1, commercial (standard) hGH; Lane 2, hGH produced by r-B. 'subtilis BGSC-1A751 (npr, apr) carrying pMK4::pre (subC)::hGH; Lane 3, hGH produced by r-B. subtilis BGSC-1A197 (spo) carrying pMK4::pre(subC)::hGH; and Lane 4, marker (Sigma M 0671).

For further characterization, rhGH was purified, from 10-fold concentrated and partially purified fermentation broth (Figure 3, Lane 1), by aptamer-based affinity chromatography, which has been developed and is being studied in our research group. Concentrated samples were mixed with hGH specific aptamer and hGH-aptamer binding was carried out at 25°C for 30 min (Figure 3, Lane 2). To obtain higher purification, the aptamer-affinity separation step was applied second time and after the elution step, rhGH was found to be separated from the fermentation broth with 99.8% purity and 41% overall yield (Figure 3, Lane 3), the molecular mass of purified rhGH was determined by MALDI-ToF mass spectrometry (MS), to verify the structure of the secreted recombinant hormone. A commercial preparation (standard) of rhGH was analyzed first, and showed a spectral peak at m/z 22,126 (Figure 4a). The stated molecular mass of the standard hormone is 22,125 Da. Thus, the detected ion was [M+H]+, with the exact molecular mass of m = 22,126 Da, where the charge on the ion is z = +1. The native length of the protein was obtained as the peak at m/z 22,133 (Figure 4b), corresponding to the [M+H]+ ion of rhGH detected with just a 0.03% error, which is a reasonable error at high molecular weights in MALDI-ToF MS analysis. This result indicates that rhGH molecule was synthesised by the recombinant construct pMK4::pre(subC)::hGH, and then secreted into the fermentation medium properly.

Figure 3.

SDS-PAGE analysis of rhGH, produced by r-B.subtilis BGSC-1A751 (npr, apr) carrying pMK4:: pre(subC)::hGH.

M, protein marker (Fermentas); Lane 1, product mixture of r-B.subtilis containing rhGH; Lane 2, 1st rhGH separation with hGH specific aptamer; Lane 3, 2nd rhGH separation with hGH specific aptamer; and Lane 4, standard hGH.

Figure 4.

MALDI-ToF MS analysis.

(a) Standard hGH (b) B. subtilis produced and purified rhGH.

The amino acid sequence of the signal peptide fused in front of the rhGH was: “MMRKKSFWLGMLTA FMLVFTMAFSDSASA” and, the N-terminal and mass spectrometry analyses indicated that the SAP signal peptide was properly processed by the B. subtilis signal peptidase (Spase I), because the results of N-terminal sequencing of the first six amino acid residues of the putative rhGH product were Phe-Pro-Thr-Ile-Pro-Leu, identical to the true hGH sequence. In support of this, nucleotide sequencing results were also 100% matching.

Host selection and effect of glucose concentration on rhGH fermentation

Effects of initial glucose concentration on the recombinant cells were investigated in laboratory-scale experiments by B. subtilis BGSC-1A751 and B. subtilis BGSC-1A197 carrying pMK4::pre(subC)::hGH, at the initial concentrations of CGo= 6.0, 8.0, 10.0, and 15.0 g L−1. The variations in glucose, cell and rhGH concentrations with the cultivation time by B. subtilis BGSC-1A751 carrying pMK4::pre(subC)::hGH are presented in Figures 5a–c, respectively. The cell concentration was not affected from the initial glucose concentration at t = 0–6 h. Because of the addition of the protease inhibitor at t = 6 h, an interruption in the cell growth was observed until t = 15 h. However, after t = 15 h the second cell growth phase was started with the initiation of rhGH synthesis (Figure 5c). The highest cell concentration was obtained at CGo= 15 g L−1 at t = 36 h as CX= 2.8 g L−1 (Figure 5a). In the first 15 h, the glucose consumption rates were close to each other; however at t > 15 h, parallel to the cell growth profiles, with the increase in cell growth rate, the glucose consumption rate increased being the highest at CGo = 15 g L−1. The highest rhGH was produced at CGo = 8 g L−1 at t = 36 h as CrhGH = 30 mg L−1. On the other hand, when the protease inhibitor was not used rhGH was not detected in the fermentation broths of B. subtilis BGSC-1A751 and B. subtilis BGSC-1A197 carrying pMK4::pre (subC)::hGH.

Figure 5.

(a) Variation in cell concentration with the cultivation time for B. subtilis BGSC-1A751 (nprapr) carrying pMK4::pre (subC)::hGH with the initial glucose concentration. CGo(g L−1): (♦) 6.0; (□) 8.0; (▴) 10.0; (○) 15.0. (b) Variation in glucose concentration with the cultivation time for B. subtilis BGSC-1A751 (nprapr) carrying pMK4::pre(subC)::hGH with the initial glucose concentration. CGo(g L−1): (♦) 6.0; (□) 8.0; (▴) 10.0; (○)15.0. (c) Variation in rhGH concentration with the cultivation time for B. subtilis BGSC-1A751 (nprapr) carrying pMK4::pre(subC)::hGH with the initial glucose concentration. CGo(g L−1): (♦) 6.0; (□) 8.0; (▴) 10.0; (○)15.0 as in 5c.

A parallel set of experiments were conducted by B. subtilis BGSC-1A197 carrying pMK4::pre(subC)::hGH. Similar to B. subtilis BGSC-1A751 results, the highest rhGH production was obtained at CGo = 8 g L−1 but with a lower rhGH value (CrhGH = 26 mg L−1). On the basis of the results, B. subtilis BGSC-1A751 was selected as the host.

Influence of hGH gene on the physiology of r-Bacillus subtilis

To determine the influence of hGH gene on the physiology of the bacilli, bioreactor experiments were performed by B. subtilis BGSC-1A751 carrying merely pMK4, and B. subtilis BGSC-1A751 carrying pMK4::pre(subC)::hGH, at T = 37°C, pH0 = 7.25, CGo = 8 g L−1, air inlet rate of 0.5 vvm and agitation rate of 800 min−1. The concentrations of the glucose, cell, extracellular rhGH, the by-products amino and organic acids, together with the oxygen-uptake (OUR) rates, oxygen-transfer coefficients (KLa), and yield coefficients were determined throughout the fermentations.

The variations in dissolved oxygen concentration (CO) and pH with the cultivation time are presented in Figure 6; and the variations in glucose, cell, and rhGH concentrations are presented in Figure 7. The loci of the CO vs. t profiles obtained in the two fermentation processes were similar until t = 18 h, where considerable decrease in dissolved oxygen concentration was observed at t = 0–4 h. However, in the process by B. subtilis carrying pMK4::pre(subC)::hGH, a considerable decrease in CO between t = 18 h and t = 22 h was detected due to rhGH synthesis that induced the cell growth. By the termination of the cell formation, a breakthrough in dissolved oxygen concentration at t = 22 h is observed.

Figure 6.

Variations in dissolved oxygen concentration and pH with the cultivation time by B.subtilis BGSC-1A751 (nprapr) carrying pMK4::pre(subC)::hGH (r-pMK4) and BGSC-1A751 (nprapr) carrying pMK4. Co: continuous lines, pH: dashed lines.

Figure 7.

Variations in glucose, cell, and hGH concentrations with the cultivation time by B. subtilis BGSC-1A751 (nprapr) carrying pMK4::pre(subC)::hGH and BGSC-1A751 (nprapr) carrying pMK4. Glucose concentration: (--▪--) pMK4; (□) pMK4::pre (subC)::hGH, Cell concentration: (--•--) pMK4; (○) pMK4::pre(subC)::hGH, rhGH concentration: (▵) pMK4::pre(subC)::hGH.

The pH in both fermentation media decreased until t = 18 h. After t = 18h, pH continued to decrease by depicting a characteristic curve by B. subtilis carrying pMK4::pre (subC)::hGH; contrarily, pH was increased in the bioreactor by B. subtilis carrying merely pMK4, as can be seen in Figure 6.

In the process by B. subtilis carrying pMK4, the glucose consumption was higher until t = 18 h of the fermentation, but after t = 18 h, it was almost zero; where the highest cell concentration and the overall cell yield on substrate (YX/S) were obtained as CX = 1.6 g L−1 (t = 12 h) and YX/S = 0.23 g g−1, respectively. Contrarily, by B. subtilis carrying pMK4::pre(subC)::hGH, the glucose consumption was increased after t = 18 h and the highest cell concentration was obtained as CX = 2.0 g L−1 (t = 24 h); and the overall cell yield on substrate was YX/S = 0.25 g g−1.

As expected, rhGH production was achieved only by B. subtilis carrying pMK4::pre(subC)::hGH. RhGH synthesis started at t = 18 h of the batch-bioprocess and increased with the cultivation time reaching the value CrhGH = 70 mg L−1 at t = 32 h. The highest product yield on substrate was obtained at 24 < t < 32 h as YrhGH/S = 0.65 g g−1, while the overall rhGH yield on substrate was YrhGH/S = 9 g kg−1.

The excreted amino acids that were detected at considerable concentrations in both fermentations are leucine, isoleucine, and phenylalanine. The highest concentrations of leucine, isoleucine, and phenylalanine by B. subtilis carrying pMK4::pre (subC)::hGH were 0.191, 0.096, and 0.132 g L−1; whereas by B. subtilis carrying pMK4 were as 0.214, 0.107, and 0.281 g L−1, respectively. Nevertheless, alanine, arginine, asparagine, aspartic acid, glutamic acid, glycine, histidine, methionine, lysine, valine, treonine, and tryptophan were not detected in the broths of the two fermentation processes. Thus, the total excreted amino acid concentrations were higher in the fermentation by B. subtilis carrying merely pMK4.

Variations in excreted organic acid concentration produced by B. subtilis carrying pMK4::pre(subC)::hGH and B. subtilis carrying pMK4 are presented in Figures 8a,b, respectively. Oxaloacetic acid, which is known to be produced in cell regeneration, was not excreted in both fermentations. In the fermentation by B. subtilis carrying pMK4::pre (subC)::hGH, the main extracellular by-products were succinic, gluconic and formic acid (Figure 8a). Lactic, oxalic, citric acids are the organic acids having lower concentrations, i.e., ca. 0.1 g L−1; moreover, pyruvic, α-ketoglutaric, and acetic acids were detected at a concentration of ca. 0.01 g L−1. Indeed, the organic acid profiles obtained by B. subtilis carrying pMK4::pre(subC)::hGH are different than that of B. subtilis carrying pMK4; where in the latter, the main excreted by-products were malic and gluconic acids (Figure 8b), and the concentration of the other organic acids were lower than 0.01 g L−1. Thus, the amount of total organic acids excreted by B. subtilis carrying pMK4::pre (subC)::hGH were higher.

Figure 8.

(a) Variation in organic acid concentrations with the cultivation time by B. subtilis BGSC-1A751 (nprapr) carrying pMK4::pre(subC)::hGH. (b) Variation in organic acid concentrations with the cultivation time by B.subtilis BGSC-1A751 (nprapr) carrying pMK4.

Fermentation characteristics

The variations in KLa and the oxygen uptake rate (OUR) are presented in Table 2. Considering the characteristic cell and rhGH concentration profiles of B. subtilis carrying pMK4::pre(subC)::hGH, the bioprocess was divided into five periods. 0 < t < 4 h is the cell first-growth-phase; 4 < t < 12 h is the growth-interruption-phase; 12 < t < 16 h is the lag-phase where rhGH synthesis starts; 16 < t < 24 h is the second-cell-growth-phase where rhGH synthesis increases; and 24 < t < 32 h is the end of the growth-phase where rhGH synthesis was the highest.

Table 2. Variation in Oxygen Transfer Characteristics with the Cultivation Time
MicroorganismPeriodKLa(s−1)OUR*103 (mol m−3 s−1)
BGSC-1A751 (nprapr−) carrying pMK4::pre(subC)::hGHFirst growth phase, 0 < t < 4 h0.0173.5
Growth-interruption-phase, 4 < t < 12 h0.0183.0
Lag-phase and rhGH synthesis phase, 12 < t < 16 h0.0142.4
Second cell growth and rhGH synthesis phase, 16 < t < 24 h0.0153.5
End of the growth and rhGH synthesis phase, 24 < t < 32 h0.0100.4
BGSC-1A751 (nprapr−) carrying pMK40 < t < 4 h0.0284.6
4 < t < 12 h
12 < t < 16 h
16 < t < 24 h0.0130.4
24 < t < 32 h0.0100.3

In both fermentation processes, the oxygen transfer coefficient, KLa, increased with the increase in the cultivation time, and then decreased. At 0 < t < 4 h of the bioprocess, KLa and oxygen uptake rate of B. subtilis carrying pMK4 was higher than that of B. subtilis carrying hGH gene. Throughout the bioprocess, the highest KLa value was obtained by B. subtilis carrying merely pMK4 as KLa= 0.028 s−1 at 0 < t < 4 h; however, due to low oxygen concentrations within t = 4–16 h in the production medium, the dynamic method could not be applied. Related with B. subtilis carrying pMK4::pre(subC)::hGH, the highest OUR values were obtained in the first growth- (0 < t< 4 h) and second-cell-growth- phases (16 < t < 24 h) (Table 2).

Discussion and Conclusions

An expression system producing therapeutic protein human growth hormone that conceptually mimics the extracellular serine alkaline protease synthesis in the genus Bacillus was designed and implemented. For the extracellular production of human growth hormone by B. subtilis, a recombinant plasmid carrying the hybrid-gene of two DNA fragments, i.e., signal (pre-) DNA sequence of a Bacillus licheniformis extracellular SAP enzyme gene (subC) and the DNA encoding hGH, was constructed and transferred into two host Bacillus strains, namely B. subtilis BGSC-1A751 (nprapr−) and B. subtilis BGSC-1A197 (spo). These strains were selected for their deficiencies in two protease genes and sporulation gene, respectively. RhGH, expressed by the hybrid-gene pre(subC)::hGH cloned into pMK4 in both hosts, was secreted.

The approach developed is expected to be applicable to the design of expression systems for heterologous protein productions from Bacillus. As the rhGH concentration obtained from the protease-deficient host B. subtilis BGSC-1A751 was higher than that of the host B. subtilis BGSC-1A197, the first was selected as the host owing to two gene deletions targeting the decrease in protease activities.

Secreted proteins are generally synthesised as precursors with a cleavable signal peptide, and then the signal peptide is removed by signal peptidases, where preprotein processing by signal peptidases are essential for bacterial growth and viability.30, 31 The native length of rhGH was detected in SDS-PAGE (Figure 2, Lane 2), Western blot (Figure 3, Lane 3) and MALDI-ToF MS analysis as the peak at m/z 22,133 (Figure 4b) with just a 0.03% error, which is a reasonable error at high molecular weights. The N-terminal and mass spectrometry analyses indicate that the signal peptidase has cut at the site within Spase I group. Thus, the system designed functioned with its intended purpose effectively in expression and cleavage of the recombinant product.

The other peak in Figure 4b at m/z 25,854 is possibly an unspecifically bound impurity protein to the hGH-ligand during purification. From the facts that MALDI-ToF MS can not be used in quantitation of proteins because protein detection depends on ionisation efficiency and that even femto-mole amounts can be detected, and from SDS-PAGE (Figure 3, Lane 3) analysis where a 25.8 kDa band was not detected, it was concluded that the impurity was in negligible amounts. Furthermore, the peaks detected at m/z 20.4 kDa in Figures 4a,b, are possibly the cleaved 20 kDa forms of hGH.32

The constructed expression system produces extracellular rhGH from B. subtilis starting from the beginning of the fermentation process parallel to the cell growth, giving a breakthrough at t = 12 h. RhGH concentration was the highest at t = 32 h as CrhGH = 70 mg L−1 and overall specific-product yield on substrate was YrhGH/S = 9 g kg−1, in the defined medium with sole carbon source glucose. Nakayama et al.,33 reported rhGH secretion level of 40 mg L−1 which is 1.75-fold lower than that obtained in this study. Kajino et al.,34 modified the “middle wall protein (MWP) signal peptide” of B. brevis and constructed a hGH expression system and reported rhGH secretion from B. brevis with an overall specific-product yield on glucose+polypeptone YrhGH/S = 4 g kg−1 which is 2.25-fold lower than the value reported in this work. Related with a different expression system for human interleukin-3, Westers et al.,35 reported 0.1 g L−1 recombinant human interleukin-3 secretion by an eight-protease-deficient B. subtilis, using a semi-defined enriched medium; where the overall specific-product yield on substrate was lower then the YrhGH/S value reported in this work. Therefore, we conclude that the expression system designed, which is based on the idea of using the ribosomal binding site- promoter- and the signal peptide of serine alkaline protease enzyme gene subC, is successful for the extracellular recombinant protein production.

The transcription for rhGH synthesis by B. subtilis BGSC-1A751 (nprapr−) carrying the hybrid-gene pre (subC)::hGH is under the control of degQ promoter; therefore, the synthesis and secretion pattern of rhGH mimics the synthesis and secretion of SAP enzyme in B. licheniformis.11,21,36 Because the ribosomal binding site, promoter, and signal peptide of the SAP gene (subC) were used in the constructed expression system for the synthesis and secretion of rhGH, a similar rhGH concentration profile to that of the SAP productions from B. licheniformis carrying pHV1431::subC21 and B. subtilis carrying pHV1431::subC11 was obtained. The slight difference observed in the concentration profiles was likely due to an interruption caused by the addition of protease inhibitors. Therefore, we conclude that the expression and secretion system constructed for rhGH production from the genus Bacillus is dependent on the bioreactor operation conditions (which is being studied), similar to that of SAP production by Bacillus species.37

To investigate the influence and perturbation effect of hGH gene on the physiology of r-B. subtilis, comparative bioreactor experiments were performed by B. subtilis carrying merely pMK4 and B. subtilis carrying pMK4::pre (subC)::hGH, on glucose as sole carbon source. The results reveal that the expression of rhGH influences the physiology of the r-Bacillus cells, as expected. The cell concentration profile of B. subtilis carrying pMK4::pre(subC)::hGH shows a perturbed biphasic variation because of the introduction of a protease inhibitor at t = 4 h, where a drastic decrease in the growth rate occurs until t = 16 h, that proceeds with an increase in the growth rate until ca. t = 24 h.

RhGH synthesis and secretion proceeded until t = 32 h, giving the highest concentration as CrhGH = 70 mg L−1. As expected, there was no rhGH production by the microorganism that does not carry hGH gene.

Due to the introduction of new biochemical reactions into the intracellular reaction network producing the heterologous extracellular protein, the fermentation and oxygen transfer characteristics and by-product distributions of B. subtilis carrying pMK4::pre(subC)::hGH were different than that of the B. subtilis carrying merely pMK4. Moreover, higher concentrations of organic acids detected in the broth of B. subtilis carrying pMK4::pre(subC)::hGH, and higher concentrations of amino acids detected in the broth of B. subtilis carrying pMK4, reveal the impact of the structure of the plasmids on the synthesis capacity of the host throughout the fermentation.

The expression system carrying the foreign gene, hGH, in the hybrid-gene fused behind the signal (pre-) DNA sequence, i.e., pre(subC), synthesizing and secreting rhGH from B. subtilis carrying pMK4::pre(subC)::hGH, was influenced by the bioreactor operation conditions which, in turn effects the oxygen transfer characteristics of the fermentation process, in the defined medium based on sole carbon source glucose. In the first cell growth phase, the oxygen uptake and glucose uptake rates by B. subtilis carrying pMK4::pre (subC)::hGH cells were lower than that of B. subtilis carrying pMK4, which resulted in lower KLa values. Furthermore, after the second-cell-growth-phase where rhGH synthesis starts, the oxygen uptake rates were higher by B. subtilis carrying pMK4::pre(subC)::hGH.

The bioreactor operation conditions applied for the fermentation experiments were the favourable conditions that were found for the SAP production by B. licheniformis carrying pHV1431::subC21 and B. subtilis carrying pHV1431::subC.11 The fermentation and oxygen-transfer characteristics reveal that the extracellular rhGH production proceeds through the constructed expression system by forming unique intracellular reaction pathways with different intracellular reaction rates compared to that of SAP production. Thus, these results encourage metabolic flux analysis using the recombinant B. subtilis carrying the constructed expression system encoding extracellular rhGH, which is being studied.

Acknowledgements

This work was supported by the Scientific and Technical Research Council of Turkey (TUBITAK) through the projects 104M012 and 107M420. Ankara University Biotechnology Institute is gratefully acknowledged for providing the mass spectrometer. Humatrope was supplied kindly by Pharmacist Tülay Latifoğlu. E, Çelik's contribution was performing the MALDI-ToF analysis.

Ancillary