Efficient expression of cyclodextrin glycosyltransferase from Geobacillus stearothermophilus in Escherichia coli by promoter engineering and downstream box evolution

Accepted Manuscript Title: Efficient expression of cyclodextrin glycosyltransferase from Geobacillus stearothermophilus in Escherichia coli by promoter engineering and downstream box evolution Authors: Chen Deng, Jianghua Li, Hyun-dong Shin, Guocheng Du, Jian Chen, Long Liu PII: DOI: Reference:

S0168-1656(17)31766-2 https://doi.org/10.1016/j.jbiotec.2017.12.009 BIOTEC 8071

To appear in:

Journal of Biotechnology

Received date: Revised date: Accepted date:

3-9-2017 14-10-2017 9-12-2017

Please cite this article as: Deng, Chen, Li, Jianghua, Shin, Hyun-dong, Du, Guocheng, Chen, Jian, Liu, Long, Efficient expression of cyclodextrin glycosyltransferase from Geobacillus stearothermophilus in Escherichia coli by promoter engineering and downstream box evolution.Journal of Biotechnology https://doi.org/10.1016/j.jbiotec.2017.12.009 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Efficient

expression

of

cyclodextrin

glycosyltransferase

from

Geobacillus

stearothermophilus in Escherichia coli by promoter engineering and downstream box evolution

IP T

Chen Deng1, 2, Jianghua Li1, 2, Hyun-dong Shin3, Guocheng Du1, 2, Jian Chen2, Long Liu1, 2*

SC R

1. Key Laboratory of Carbohydrate Chemistry and Biotechnology, Ministry of Education, Jiangnan University, Wuxi 214122, China

U

2. Key Laboratory of Industrial Biotechnology, Ministry of Education, Jiangnan University, Wuxi

N

214122, China

A

3. School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta

ED

M

30332, USA

*Corresponding author: Long Liu, Tel.: +86-510-85918312, Fax: +86-510-85918309, E-mail:

CC E

PT

[email protected]

Highlights

The CGTase gene from Geobacillus stearothermophilus was successfully expressed in E. coli.



The activity of CGTase reached 110.2 U/mL with the promoter PtacI.

A





The activity of CGTase reached 170.56 U/mL by optimizing the promoter spacer sequence.



The activity of CGTase reached 213.96 U/mL by optimizing the downstream box sequence.

1

Abstract: Cyclodextrin glycosyltransferase (CGTase) catalyzes hydrolysis, cyclization, coupling, and disproportionation reactions and is widely used in the starch processing industry. In this work, the expression of CGTase from Geobacillus stearothermophilus in Escherichia coli BL21 (DE3) was significantly improved by promoter engineering and downstream box evolution. Firstly, the

IP T

effects of the promoter type (PT7, Ptrp, PlacUV5, and the hybrid promoters PtacI and PtacII) and spacer sequence on the expression of CGTase were examined. PtacI demonstrated the highest rate of

SC R

transcriptional activity, which was 4.4-, 7.1-, 3.3-, and 1.5-fold greater than that of PT7, Ptrp, PlacUV5, and PtacII, respectively. The spacer sequence of the promoter was optimized using a degenerate

U

base library, and the GC content of the spacer was found to be inversely proportional to CGTase

N

expression. In addition, CGTase expression was higher when TG:CA and TA:TA dimers were

A

present in the spacer sequence. Under the control of the PtacI promoter with an optimized spacer

M

sequence, extracellular CGTase activity reached 170.6 U/mL, which was seven times higher than

ED

that of the original strain (25.2 U/mL). Directed evolution of the downstream box sequence was then performed by randomization of the sequence using degenerate codons, similarly as for the

PT

optimization of the spacer sequence. After optimizing the downstream box sequence, CGTase

CC E

activity increased from 170.6 to 214 U/mL. The results obtained here indicate that in addition to promoter type, the spacer sequence of the promoter and the downstream box are important

A

target elements for improved gene expression.

Key words: Cyclodextrin glycosyltransferase, Geobacillus stearothermophilus, hybrid promoter, spacer sequence, downstream box

2

1. Introduction Cyclodextrin glycosyltransferase (CGTase) belongs to glycosyl hydrolase family 13 and catalyzes transglycosylation reactions, including hydrolysis,

cyclization, coupling,

and

IP T

disproportionation (van der Veen et al., 2000). Because of its extensive applications, demand for CGTase is increasing. Escherichia coli (Lee et al., 2013), Saccharomyces cerevisiae (Wang et al.,

SC R

2006), Bacillus subtilis (Paloheimo et al., 1992), and Bacillus megaterium (Zhou et al., 2012) have been used for heterologous expression of CGTase.

U

PT7 is a strong constitutive promoter commonly used in E. coli. This promoter can be used to

N

drive expression of foreign genes in E. coli, but this process is energy-intensive, affecting the

A

necessary functions of the host cell. To overcome this drawback, expression of cloned genes may

M

be controlled by an inducible promoter such as PlacUV5 or Ptrp from E. coli, or the hybrid promoters

ED

PtacI or PtacII, which include regions of both the PlacUV5 and Ptrp promoters. There is evidence that E. coli RNA polymerase (RNAP) binds to three sites on the promoter: the -10 sequence, the -35

PT

sequence, and the sites between them (Shimada et al., 2014). The relative positions of these DNA

CC E

sequences are illustrated in Fig. 1A. PtacI consists of the -35 sequence of the Ptrp promoter fused to the -10 sequence of the PlacUV5 promoter. PtacII consists of the -11 sequence of the Ptrp promoter fused to part of the PlacUV5 promoter and a synthetic ribosome binding site (RBS) and

A

Shine-Dalgarno (SD) sequence (Fig. 1B). The inclusion of a lactose operon downstream of the PtacI and PtacII promoters enables their

induction by isopropyl-β-D-thiogalactoside

(IPTG)

(Aghaabdollahian et al., 2014). The expression of high levels of exogenous protein in E. coli depends not only on the 3

promoter, terminator, and RBS sequences, but also on another regulatory region: the downstream box (DB) sequence (Farran et al., 2010; Gray et al., 2009). The DB sequence has a functional role in many prokaryotic mRNAs and was initially thought to be a translational enhancer element. The sequence consists of approximately 8-13 nucleotides downstream of the

IP T

initiation codon, and is often present downstream of highly expressed E. coli and phage mRNAs. Previous studies have revealed that the DB sequence can influence both translation efficiency

SC R

(Gray et al., 2011) and RNA levels (Gray et al., 2009), and that the DB sequence and SD sequence, both of which are sites that must be bound for translation initiation, can enhance translation

U

synergistically (Sprengart et al., 1996). Thus, optimization of DB sequences and characterization

N

of their distribution is critical for the expression of exogenous genes in E. coli.

A

In previous work, the PT7 promoter was used to express CGTase (Han et al., 2013). However,

M

the rate of protein expression was too high, and aggregation of the protein into inclusion bodies

ED

was observed. In this work, we aimed to find a more suitable promoter to express CGTase. Four promoters were assessed: Ptrp, PlacUV5, and the hybrid promoters PtacI and PtacII. The 16-18 bp

PT

nucleic acid sequence located between the -35 and -10 sequences of the promoter, referred to as

CC E

the spacer sequence, affects the efficiency of interaction between the bacterial RNAP and the conserved -10 and -35 promoter elements, which leads to the formation of an open complex (Sztiller-Sikorska et al., 2011). We utilized degenerate codon libraries to chemically synthesize an

A

optimized spacer sequence for PtacI, and optimized the GC content of the spacer sequence in the hybrid promoter to optimize DNA integration and enhance expression of the CGTase gene. Furthermore, we optimized DB sequences to improve the efficiency of translation. Through promoter engineering and directed evolution of the DB, the expression level of CGTase was 4

improved 8.5-fold, from 25.2 to 214 U/mL.

2. Materials and methods 2.1. Bacterial strains, plasmids, and materials.

IP T

E. coli JM109 and E. coli BL21 (DE3) were used for plasmid construction and expression, respectively, and the expression vector was pET-20b(+). These materials were stored in our

SC R

laboratory. PrimeSTAR HS DNA polymerase premix, T4 DNA ligase, PCR reagents, and the DNA

purification kit (TaKaRa MiniBEST Agarose Gel DNA Extraction Kit Ver. 4.0) were acquired from

U

TaKaRa (Otsu, Japan). Restriction enzymes were purchased from Thermo Fisher Scientific

N

(Shanghai, China). All primers were synthesized by Sangon Biological Engineering Technology &

A

Services Co., Ltd. (Shanghai, China). MES SDS running buffer and LDS sample buffer were

ED

M

purchased from Novex (San Diego, CA). All other chemicals and reagents were of analytical grade.

PT

2.2. Recombinant plasmid construction

The Ptrp and PlacUV5 promoters were amplified by PCR, using genomic DNA from E. coli strain

CC E

O177:H21 (NCBI accession number CP016546) and the pASK84 plasmid as templates, respectively (Skerra, 1994). The sequences of primers P1, P2, P3, and P4 are listed in Table 1. A

A

BglII restriction site was introduced at the 5ʹ end of primers P1 and P3, while an NdeI restriction site was introduced at the 5ʹ end of primers P2 and P4. The pET-20b(+) plasmid was digested with BglII and NdeI, allowing replacement of the PT7 promoter with either PlacUV5 or Ptrp from the

5

similarly digested PCR products. The resulting recombinant plasmids were named pET-20b(+)-lacUV5 and pET-20b(+)-trp, respectively. To construct PtacI, the -35 fragment of the Ptrp promoter was extracted from pET-20b(+)-trp-CGT through digestion with BglII and TaqI, and the -10 fragment of the PlacUV5

IP T

promoter was isolated from pET-20b(+)-lacUV5 using HpaII and NdeI. To construct PtacII, the DNA

downstream of positions -35 and -20 was derived from the Ptrp promoter through digestion of

SC R

pET-20b(+)-trp with BglII and HpaII, and the sequence at position -10 was synthesized based on the sequence of the PlacUV5 promoter. The promoter sequences are shown in Fig. 1B. The RBS

U

sequence of each promoter was fused to the 5ʹ end of the CGTase gene (Fig. 1B) (Coste et al.,

N

2009; Kim et al., 2011; Studholme and Dixon, 2003).

A

Genomic DNA from G. stearothermophilus strain NO. 2 (ATCC number: 7953-MINI-PACK™)

M

was extracted using a Genomic Extraction Kit (Tiangen Biochemical Technology Co., Ltd, Beijing,

ED

China) and used as the template for PCR amplification of the cyclomaltodextrin glucanotransferase gene. The sequence of the CGTase gene from G. stearothermophilus strain

PT

NO. 2 (GenBank accession number X59043.1) was obtained from the GenBank database of the

CC E

National Center for Biotechnology Information, and the signal peptide of the protein was predicted using the SignalP server (http://www.cbs.dtu.dk/services/SignalP/), which indicated

A

that the first 16 amino acids of the protein constitute the signal peptide. The primers CGT F1 and CGT R1, whose sequences are listed in Table 1, were designed to amplify the CGTase gene without the original signal peptide. Restriction sites for BamHI and XhoI were introduced into the forward primer CGT F1 and the reverse primer CGT R1, respectively. The PCR product, which was gel-purified using the DNA purification kit, was digested with BamHI and XhoI and ligated with 6

the similarly digested expression vector pET-20b(+) to construct the recombinant plasmid pET-20b(+)-CGT1. In order to improve expression of CGTase in E. coli, the plasmid pET-20b(+)-CGT1, which contains the original CGTase gene, was sent to Sangon Biological Engineering Technology & Services Co., Ltd. (Shanghai, China) for codon optimization. The

IP T

plasmid containing codon-optimized CGTase was named pET-20b(+)-CGT. The primers CGT F and CGT R, which contain cleavage sites for the restriction endonucleases BamHI and XhoI,

SC R

respectively, and target each end of the codon-optimized CGTase gene, were designed. The pET-20b(+)-CGT plasmid was then used as the template to amplify the codon-optimized CGTase

U

gene (each end containing a cleavage site for BamHI or XhoI). The resulting PCR product was

N

digested with gel and then digested with BamHI and XhoI restriction enzymes for 2 h. Similarly,

A

the pET-20b(+)-trp, pET-20b(+)-lacUV5, pET-20b(+)-tacI, and pET-20b(+)-tacII plasmids were all

M

digested with BamHI and XhoI restriction enzymes for 2 h. The recombinant plasmids

ED

pET-20b(+)-trp-CGT, pET-20b(+)-lacUV5-CGT, pET-20b(+)-tacI-CGT, and pET-20b(+)-tacII-CGT were obtained by ligating pET-20b(+)-trp, pET-20b(+)-lacUV5, pET-20b(+)-tacI, and pET-20b(+)-tacII,

PT

respectively, with the digested CGTase gene overnight at 16 °C. The recombinant plasmids were

CC E

used to transform chemically competent E. coli JM109. The plasmid sequences were confirmed by DNA sequencing (Sangon Biological Engineering Technology & Services Co., Ltd., Shanghai, China). These plasmids were then transformed into E. coli BL21(DE3). The preservation strains

A

were used for fermentation experiments.

2.3. Spacer region engineering To construct plasmids with different spacers (the spacer sequences are shown in Table 2), the 7

pET-20b (+)-tacI-CGT plasmid developed in this study was used as a template for PCR using the primers F1 and R1, which were designed to randomize the spacer sequence (ATTAATCATCGGCTCG) of the promoter. PCR products were validated using electrophoresis, and the template was removed from each reaction by digestion with DpnI for 2 h. The resulting constructs were cloned

IP T

into competent E. coli BL21 (DE3), and colonies were inoculated into 96-well plates with each well filled with 600 μL Luria-Bertani (LB) medium. One well in each plate was used for a control

SC R

sample (the original strain). After 8 hours of seed culture, 200 μL of the fermentation broth was

removed from each well, and 400 μL of Terrific Broth (TB: 18 g/L peptone, 20 g/L yeast extract,

U

16.4 g/L K2HPO4, 2.3 g/L KH2PO4, 10 g/L glycerol, 0.1 g/L ampicillin, pH 7.0) was added to each

N

well to carry out fermentation experiments. Fermentation was performed at 25 °C for 16 h, and

A

the temperature was then adjusted to 30 °C for 29 h. After 45 h of fermentation, the cultures

M

were centrifuged, and the supernatants were assayed to measure CGTase activity. The mutant

ED

strain with the highest CGTase activity was designated P1-1, containing the plasmid CGT-P1-1 and spacer sequence 1. In order to study the effect of the GC content of the spacer sequence of the

PT

promoter, on the basis of plasmid CGT-P1-1, a series of recombinant plasmids containing spacers

CC E

with different GC contents were constructed: plasmids CGT-P1-2, CGT-P1-3, CGT-P1-4, CGT-P1-5, and CGT-P1-6, containing spacers 2, 3, 4, 5, and 6 (Table 2), respectively, were constructed using the primer pairs F2/R2, F3/R3, F4/R4, F5/R5, and F6/R6, respectively. The sequences of these

A

primers used to mutate the original spacer sequence are shown in Table 1. The strains containing these plasmids were named P1-2, P1-3, P1-4, P1-5, and P1-6, respectively. The gel-purified PCR products without template were transformed into chemically competent E. coli JM109. The sequences of the plasmids isolated from these transformants were confirmed by DNA sequencing. 8

After the sequence of each plasmid was validated, the plasmid was extracted from E. coli JM109 and transformed into E. coli BL21(DE3).

2.4. Directed evolution of the downstream box sequence

IP T

To rearrange the DB sequence, the plasmid CGT-P1-1 was used as a template and the primers dbF and dbR, which contain 13 degenerate bases between positions +9 and +21, were used to

SC R

replace the original 13-base DB sequence (CCTGCTGCCGACC) with randomized sequences. The resulting constructs were cloned into competent E. coli BL21 (DE3), and colonies were inoculated

U

into 96-well plates. Mutant strains were screened using the same method as for optimization of

N

the spacer sequence, and those with high CGTase activity were selected for flask fermentation

ED

2.5. Production of CGTase

M

A

screening.

Plasmids were transformed into E. coli BL21 (DE3), and the resulting transformants were

PT

inoculated into 25 mL of LB medium containing 100 μg/mL ampicillin. Cultures were grown for 8

CC E

h at 37 °C, with shaking at 220 rpm. The resulting seed culture was then used to inoculate TB at a ratio of 4% (v/v) to give a final volume of 50 mL. Cultures were incubated in shake flasks at 37 °C, with shaking at 220 rpm, until the optical density at 600 nm (OD600) reached 0.7. IPTG was added

A

to each culture to a final concentration of 10 µmol/L in order to induce the PtacI and PtacII promoters, both of which carry the lac operator. The shake flasks were then rapidly transferred to 25 °C and incubated for 32 h, and the temperature was adjusted to 30 °C for a further 58 h incubation before centrifugation at 10,000 × g and 4 °C for 10 min (Li et al., 2010). Unless 9

otherwise noted, all experiments were performed independently at least three times, and the corresponding figures show the average value and the standard error for each experiment.

2.6. Analysis of CGTase activity

IP T

Out of the four reactions that CGTase can perform, the disproportion reaction is the main reaction. This reaction consists of cleavage of a linear polysaccharide and transfer of one of the

SC R

cleavage products onto another linear acceptor polysaccharide. If the substrate is starch, disproportionation occurs mainly in the initial stages of the CGTase-catalyzed reaction, resulting

N

assessed by determination of disproportionation activity.

U

in a rapid decline in the viscosity of starch after gelatinization. In this paper, CGTase activity was

A

Disproportionation activity was determined as described previously (van der Veen et al.,

M

2000), using 1 mM 4-nitrophenyl-α-D-maltoheptaoside-4-6-O-ethylidene (EPS: Megazyme,

ED

County Wicklow, Ireland) and 10 mM maltose as the donor and acceptor substrates, respectively. One unit of activity was defined as the amount of enzyme required to convert 1 μmol EPS/min.

PT

Protein concentrations were determined using the Bradford method and the Bradford Protein

CC E

Assay Kit (Beyotime, Jiangsu, China), with bovine serum albumin as a standard.

A

2.7. Protein expression analysis The fermentation supernatant was diluted five-fold, mixed with the appropriate volume of

denaturing buffer (4× LDS sample buffer), and incubated for 10 min at 70 ° C. Twenty microliters of sample were resolved by 10% SDS-PAGE (Bio-Rad Laboratories, Hercules, CA, USA). The prepared samples and marker were loaded into wells and separated by electrophoresis in MES 10

SDS running buffer, and protein bands were visualized by staining with Coomassie Brilliant Blue R-250.

3. Results

IP T

3.1. Construction of recombinant plasmids and production of CGTase by fermentation The CGTase gene from G. stearothermophilus was successfully amplified by PCR. The gene

SC R

shared 66% sequence identity with the corresponding gene from Bacillus circulans 251. All mutant plasmids used in this study were successfully constructed, as verified by sequencing, and

U

transferred into the host E. coli BL21 (DE3) strain. The resulting strains exhibited differences in the

N

enzyme activity of the fermentation supernatant after fermentation. SDS-PAGE showed that the

A

molecular weight of CGTase produced by each strain was approximately 70 kDa, which is similar

ED

M

to the previously reported molecular weight of CGTase.

3.2. CGTase expression under the control of various promoters

PT

The expression level of the CGTase gene after fermentation varied depending on the

CC E

promoter. The levels of CGTase expression obtained using different promoters were compared by SDS-PAGE (Fig.2A). The strain containing the heterozygous promoter PtacI produced the highest CGTase expression after fermentation, followed by the strain containing PtacII. The enzyme activity

A

of CGTase produced under the control of different promoters (PT7, Ptrp, PlacUV5, PtacI, or PtacII) after 90 h of fermentation in shake flasks is shown in Fig. 2B. The activity of CGTase produced by the PtacI promoter increased significantly with increasing fermentation time, and enzyme activity mediated by the PtacII promoter increased to a greater extent with increasing fermentation time 11

than enzyme activity mediated by the promoters from which PtacII is derived. The extracellular enzyme activity resulting from expression of the plasmid containing the PtacI promoter reached 110.2 U/mL after 90 h, which was 3.4 times higher than that obtained using the original PT7 promoter. The activities obtained using the PtacII, PlacUV5, PT7, and Ptrp promoters were 70.4 U/mL,

IP T

33.8 U/mL, 25.2 U/mL, and 10 U/mL, respectively.

SC R

3.3. Optimization of the spacer sequence of PtacI for improved CGTase gene expression

After the 96-well plates were screened as described in Section 2.3, one of the strains with

U

the highest CGTase activity out of all plates was selected, and the mutant plasmid isolated from

N

this strain was named CGT-P1-1. The plasmid was sequenced, and the sequence of the spacer

A

sequence was determined to be ATTAATCATGCATGTA. A series of recombinant plasmids with

M

different GC contents in the spacer sequence (CGT-P1-2 / CGT-P1-3 / CGT-P1-4 / CGT-P1-5 /

ED

CGT-P1-6) were then constructed using plasmid CGT-P1-1 as a template. Determination of enzyme activity in fermentation supernatants showed that the strain containing the CGT-P1-1

PT

plasmid produced the most CGTase, followed by CGT-P1-2, CGT-P1-3, CGT-P1-4, CGT-P1-5, and

CC E

CGT-P1-6. The expression level of CGTase showed a trend based on the GC content of the promoter spacer sequence: the highest expression of CGTase was obtained in the fermentation medium of P1-1, with the lowest GC content in the spacer sequence, while CGTase expression

A

tended to decrease with increasing GC content (Fig. 3A). The strain containing the CGT-P1-1 plasmid produced 170.6 U/mL CGTase, which is 1.7 times greater than that produced by the PtacI strain (110.2 U/mL) (Fig. 3B). The CGT-P1-2, CGT-P1-3, CGT-P1-4, CGT-P1-5, and CGT-P1-6 plasmids produced 145 U/mL, 111 U/mL, 140 U/mL, 120.4 U/mL, and 105.6 U/mL CGTase, 12

respectively. The GC contents of the P1-1 and P1-2 promoter spacers were 25%, whereas those of the P1-3, P1-4, P1-5, P1-6, and PtacI promoter spacers were 31.3%, 31.3%, 37.5%, 43.8%, and 43.8%, respectively (Fig. 4). These data support the hypothesis that the GC content of a promoter’s spacer sequence may influence the expression level of a target gene. In this case,

IP T

high GC content in the spacer sequence of the promoter was inversely correlated with CGTase

SC R

activity.

3.4. Directed evolution of the downstream box for improved CGTase expression

U

Degenerate primers (dbF and dbR) were designed to randomize the DB sequence, resulting in a

N

series of recombinant plasmids with different DB sequences. CGTase activity was measured in the

A

fermentation supernatants of 384 strains (including four original, unaltered strains) in four

M

96-well plates. Two strains with significantly improved enzyme activity and one strain with

ED

significantly decreased enzyme activity were selected, and the corresponding plasmids were sequenced. The strains with improved activity were named DB1 and DB2, and the strain with

PT

reduced activity was named DB3 (Fig. 5). Fig. 5A shows differences in the amount of enzyme

CC E

produced after fermentation using several strains with different DB sequences. Fig. 5B shows the sequence of the DB region for several selected strains. The sequence of the DB region is represented by uppercase letters. The protein concentrations in the supernatants of DB1 and DB2

A

were 6.3 mg/mL and 5.6 mg/mL, respectively, corresponding to CGTase activities of 214 U/mL and 189.6 U/mL (Fig. 5C). These activities represented increases of 25.5% and 11.2%, respectively, compared with the activity of the P1-1 strain containing the original DB sequence. The protein concentration in the supernatant of the DB3 strain was 3.9 mg/mL, and the activity of the 13

extracellular enzyme was 46.2 U/mL.

4. Discussion In the present study, we optimized the promoter of a CGTase-expressing vector, identifying

IP T

the hybrid promoter PtacI as the most efficient for protein production. We then optimized the spacer sequence between the -35 sequence and the -10 sequence of this construct and

SC R

optimized the DB sequence by randomizing the sequence using degenerate codons.

The promoter sequence is the basic determinant of gene expression, and mutations in its

U

-35 or - 10 sequences can affect the strength of the promoter. When the same sequence as the

N

existing promoter region, namely -10 sequence of 5ʹ-TATAAT-3ʹ and - 35 sequence of 5ʹ-TTGACA-3ʹ,

A

the intensity of the promoter enhanced. The E. coli -35 promoter sequence is generally

M

5ʹ-TTGACA-3ʹ, and the sequence of the Pribnow box is generally 5ʹ-TATAAT-3ʹ. The Ptrp promoter

ED

has the -35 sequence 5ʹ-TTGACA-3ʹ but does not have the -10 sequence 5ʹ-TATAAT-3ʹ. The PlacUV5 promoter has the -10 sequence 5ʹ-TATAAT-3ʹ but does not have the -35 sequence 5ʹ-TTGACA-3ʹ.

PT

The PtacI promoter has not only the -35 sequence 5ʹ-TTGACA-3ʹ, but also the -10 sequence

CC E

5ʹ-TATAAT-3ʹ. To understand the relationship between promoter architecture and function, in a previous study (Cox et al., 2007), a combinatorial library of random promoter architectures was constructed to explore the effects of the -35 and -10 sequences on the transcriptional activity of

A

the promoter. The results showed that a promoter containing a 5ʹ-TTGACA-3ʹ sequence at the -35 position and a 5ʹ-TATAAT- 3ʹ sequence at the -10 position resulted in higher transcript levels. This result may be explained by the fact that when the -10 sequence is 5ʹ-TATAAT-3ʹ and the -35 sequence is 5ʹ-TTGACA-3ʹ, the DNA structure in the promoter region is more favorable for RNAP 14

recognition and binding to these sites. Structural differences in this region may affect the identification of the promoter by RNAP and regulation of the transcription rate. The distance between the SD sequence at the ribosome binding site and the transcription initiation site (ATG) also has a certain effect on the level of transcription (Kawaguchi et al., 1986),

PtacI and PtacII promoters are therefore superior to that of the Ptrp promoter.

IP T

with a shorter distance being more favorable for transcription. The transcriptional levels of the

SC R

The spacer sequence also affects the strength of the promoter. We optimized the spacer sequence of the hybrid promoter PtacI and greatly enhanced the extracellular activity of CGTase

U

compared with the original construct. These results show that differences in the promoter

N

sequence, including its spacer sequence, significantly affect gene expression levels. In a previous

A

study, Dietrich et al. used a fork junction DNA and observed an open complex of DNA structure

M

mutations found in the -10 promoter element whose RNAP-binding capacity showed an alarming

ED

loss of dependence on RNAP binding to the -35 promoter element (Dietrich et al., 1993). This suggests that the sequence of the spacer region is essential for the integration of DNA and that

PT

many naturally occurring promoter spacers are inefficient or of inappropriate length. The spacer

CC E

sequences between and around the -35 and -10 positions of E. coli promoters have little homology but conserved length. Our results suggest that the optimal length of the spacer sequence between the -35 and -10 sequences is 16 bp, which may be favorable for gene

A

transcription. PlacUV5 contains an 18-bp spacer, and Ptrp and PtacII both contain a 17-bp spacer. Compared with the above three promoters, the promoter PtacI, which contains a spacer sequence of 16 bp, is more ideal for gene transcription. By artificially changing the nucleotide combination and length of the spacer sequence, the expression level of a gene can be effectively increased. 15

The GC content (Calistri et al., 2011) and presence of TG:CA and TA:TA dimers in the spacer region also affect the strength of the entire promoter (Liu et al., 2004). It has been found that the sequence of the spacer affects the ability of RNAP to bind to a promoter and form an open complex, ultimately affecting transcriptional strength. Promoters containing spacer sequences

IP T

with low GC content enhance the ability of RNAP to bind to the promoter and enhance gene expression. A higher GC content would create a greater energy barrier for structural distortion of

SC R

the spacer region, which would in turn decrease gene expression. Alternatively, high GC content

can directly or indirectly suppress the distortion of neighboring AT-rich sequences. The influence

U

of GC-rich sequences on the melting of contiguous AT-rich sequences has been reported (Calistri

N

et al., 2011).

A

Our results also confirm that if the spacer sequence contains TG:CA or TA:TA dimers, the

M

transcriptional level of the promoter is enhanced. The P1-1 strain exhibited higher CGTase

spacer region.

ED

expression than the P1-2 strain due to the presence of TG:CA and TA:TA dimers in its promoter

PT

The role of the DB sequence is to regulate the translation level of a gene by affecting

CC E

translation initiation and the secondary structure of the mRNA. Similar to the results obtained from spacer sequence optimization, an increase in the AT content of the DB nucleotide sequence can promote mRNA stability and gene expression. It has been demonstrated that the DB

A

sequence is an efficient and independent translation initiation signal and that protein accumulation can vary over several orders of magnitude depending on the DB region (Sprengart et al., 1996). A recent report on protein stability determinants in chloroplasts additionally suggested that the N-terminal part of the protein encoded by the DB sequence plays an 16

important role in protein stability (Apel et al., 2010). Additionally, as an original translation enhancer in E. coli, the DB sequence can promote translation independently of the SD sequence. Optimization of the DB at various positions downstream of the initiation codon has been shown to promote highly efficient protein synthesis despite the lack of an SD sequence (Sprengart et al.,

IP T

1996). In addition, by comparing the protein contents of fermentation supernatants produced by various strains, we found that optimization of the DB sequence yielded a strain that could

SC R

produce a greater amount of CGTase in a given fermentation time. In order to further enhance

the intensity of the promoter, the sequence of the RBS, the sequences upstream of the -35

U

sequence and downstream of the -10 sequence, and other positions near the DB sequence (e.g.,

N

from the start codon to the spacer region) could be optimized. In this paper, we conducted

A

rational optimization of promoter and DB sequences. Rational selection of these two gene

M

regulatory elements plays an important role in the efficiency of foreign gene expression. The

ED

methods described in this study are expected to be valuable tools for the optimization of CGTase

PT

production.

CC E

Conflict of Interest

The authors declare that they have no competing financial interests.

A

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

17

Acknowledgements This work was financially supported by the National Natural Science Foundation (31622001,31671845, 21676119), and the Fundamental Research Funds for the Central

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Universities (JUSRP51307A and JUSRP51307A).

18

References Aghaabdollahian, S., Rabbani, M., Ghaedi, K., Sadeghi, H.M., 2014. Molecular cloning of Reteplase and its expression in E. coli using tac promoter. Adv Biomed Res. 3, 190-190. Apel, W., Schulze, W.X., Bock, R., 2010. Identification of protein stability determinants in

IP T

chloroplasts. Plant J. 63, 636-650. Bart A. van der Veen, J.C.M.U., Bauke W. Dijkstra, Lubbert Dijkhuizen, 2000. The role of arginine

SC R

47 in the cyclization and coupling reactions of cyclodextrin glycosyltransferase from Bacillus circulans strain 251 Implications for product inhibition and product specificity. Eur. J. Biochem.

U

267, 3432-3441.

A

promoters. Mol Phylogenet Evol. 60, 228-235.

N

Calistri, E., Livi, R., Buiatti, M., 2011. Evolutionary trends of GC/AT distribution patterns in

M

Coste, A.T., Crittin, J., Bauser, C., Rohde, B., Sanglard, D., 2009. Functional analysis of cis- and

ED

trans-acting elements of the Candida albicans CDR2 promoter with a novel promoter reporter system. Eukaryot. Cell. 8, 1250-1267.

PT

Cox, R.S., 3rd, Surette, M.G., Elowitz, M.B., 2007. Programming gene expression with

CC E

combinatorial promoters. Mol. Syst. Biol. 3, 145. Farran, I., McCarthy-Suarez, I., Rio-Manterola, F., Mansilla, C., Lasarte, J.J., Mingo-Castel, A.M., 2010. The vaccine adjuvant extra domain A from fibronectin retains its proinflammatory

A

properties when expressed in tobacco chloroplasts. Planta. 231, 977-990. Gray, B.N., Ahner, B.A., Hanson, M.R., 2009. High-level bacterial cellulase accumulation in chloroplast-transformed tobacco mediated by downstream box fusions. Biotechnol. Bioeng. 102, 1045-1054. 19

Gray, B.N., Yang, H., Ahner, B.A., Hanson, M.R., 2011. An efficient downstream box fusion allows high-level accumulation of active bacterial beta-glucosidase in tobacco chloroplasts. Plant Mol. Biol. 76, 345-355. Han, R., Liu, L., Shin, H.D., Chen, R.R., Du, G., Chen, J., 2013. Site-saturation engineering of lysine

IP T

47 in cyclodextrin glycosyltransferase from Paenibacillus macerans to enhance substrate specificity towards maltodextrin for enzymatic synthesis of 2-O-D-glucopyranosyl-L-ascorbic acid

SC R

(AA-2G). Appl. Microbiol. Biotechnol. 97, 5851-5860.

Kim, J.H., Lee, B.R., Lee, Y.-P., 2011. Secretory overproduction of the aminopeptidase from

U

Bacillus licheniformis by a novel hybrid promoter in Bacillus subtilis. World J Microb Biot.27,

N

2747-2751.

A

Lee, Y.S., Zhou, Y., Park, D.J., Chang, J., Choi, Y.L., 2013. Beta-cyclodextrin production by the

M

cyclodextrin glucanotransferase from Paenibacillus illinoisensis ZY-08: cloning, purification, and

ED

properties. World J. Microbiol. Biotechnol. 29, 865-873. Li, Z., Li, B., Gu, Z., Du, G., Wu, J., Chen, J., 2010. Extracellular expression and biochemical

PT

characterization of alpha-cyclodextrin glycosyltransferase from Paenibacillus macerans.

CC E

Carbohydr Res. 345, 886-892.

Liu, M., Tolstorukov, M., Zhurkin, V., Garges, S., Adhya, S., 2004. A mutant spacer sequence between -35 and -10 elements makes the Plac promoter hyperactive and cAMP receptor

A

protein-independent.. Proc Natl Acad Sci USA 101, 6911-6916. M L Sprengart, E Fuchs, Porter, A.G., 1996. The downstream box—an efficient and independent translation initiation signal in Escherichia coli. Embo J. 15, 665-674. Maria Paloheimo, Dan Haglund, Sirpa Aho, Korhola, M., 1992. Production of cyclomaltodextrin 20

glucanotransferase of Bacillus circulans var. alkalophilus ATCC21783 in B. subtilis. Appl. Microbiol. Biotechnol. 36, 584-591. P. Dietrich, M.B.S., Affonso., M.H.T., Floeter-Winter., L.M., 1993. The Trypanosoma cruzi ribosomal RNA-encoding gene analysis of promoter and upstream intergenic spacer sequences. Gene. 125,

IP T

103-107. Shimada, T., Yamazaki, Y., Tanaka, K., Ishihama, A., 2014. The whole set of constitutive promoters

SC R

recognized by RNA polymerase RpoD holoenzyme of Escherichia coli. PLoS One. 9, e90447.

Skerra, A., 1994. A general vector, pASK84, for cloning, bacterial production, and single-step

U

purification of antibody Fab fragments. Gene. 141, 79-84.

A

Activators. J. Bacteriol. 185, 1757-1767.

N

Studholme, D.J., Dixon, R., 2003. Domain Architectures of 54-Dependent Transcriptional

M

Sztiller-Sikorska, M., Heyduk, E., Heyduk, T., 2011. Promoter spacer DNA plays an active role in

ED

integrating the functional consequences of RNA polymerase contacts with -10 and -35 promoter elements. Biophys Chem. 159, 73-81.

PT

van der Veen, B.A., van Alebeek, G.-J.W.M., Uitdehaag, J.C.M., Dijkstra, B.W., Dijkhuizen, L., 2000.

CC E

The three transglycosylation reactions catalyzed by cyclodextrin glycosyltransferase from Bacillus circulans (strain 251) proceed via different kinetic mechanisms. Eur. J. Biochem. 267, 658-665. Wang, Z., Qi, Q., Wang, P.G., 2006. Engineering of Cyclodextrin Glucanotransferase on the Cell

A

Surface of Saccharomyces cerevisiae for Improved Cyclodextrin Production. Appl. Environ. Microbiol. 72, 1873-1877. Y Kawaguchi, N Yanagida, T Uozumi, Beppu, T., 1986. Improved direct expression of prochymosin cDNA through changing the SD-ATG codon length. Agric. Biol. Chem. 50, 499-500. 21

Yi Zhou, Yong-Seok Lee, In-Hye Park, Zheng-xiang Sun, Ting-xian Yang, Pei Yang, Yong-Lark Choi, Sun, M., 2012. Cyclodextrin glycosyltransferase encoded by a gene of Paenibacillus azotofixans YUPP-5 exhibited a new function to hydrolyze polysaccharides with β-1,4 linkage. Enzyme Microb

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Tech. Technol. 50, 151-157.

22

Figure captions Fig. 1 The position of the promoter structure on the DNA strand and the nucleotide sequence of the promoters used in this paper. A: The positions of the promoters -35 sequence, -10 sequence, and the spacer sequence B：Nucleic acid sequences of Ptrp, PlacUV5, PtacI and PtacII as shown in the above figure.

IP T

-35 sequence and -10 sequence are underlined. The SD sequence and the start codon were drawn with an overline. The tryptophan repressor binding site and the lac operator were dashed under the

SC R

dashed line. The point at which the PtacI and PtacII promoters were disconnected is the point of recombination. The transcription start site was indicated by +1.

U

Fig.2. Effects of different promoters on fermentation of CGTase. A: SDS-polyacrylamide gel

N

electrophoresis (SDS-PAGE) in the supernatant of CGTase produced by four different promoters. The

A

bands 1/3/5/7 showed the protein components of the fermentation supernatants of different bacterial

M

strains containing different promoters. The bands 2 / 4 / 6 / 8 were the corresponding intracellular

ED

protein contents. The target band size is about 70 kDa. B: the CGTase activity produced by four different promoters with time. ▼:PtacI;◆:PtacII; ▲: PlacUV5;

■:PT7;●: Ptrp.

PT

Fig. 3. Effects of different spacer sequences on fermentation of CGTase. A: SDS-PAGEimage of the

CC E

supernatant after 90 hours of CGTase fermentation with promoters containing different spacer regions. The target protein band was about 70 kDa. B: the CGTase activity with the promoter containing the different spacer regions varies with time. The enzyme activity was measured every 10 hours during

A

◀

▶

■

the fermentation cycle. ●:P1-1;▲:P1-2;▼:P1-3;◆:P1-4; :P1-5; :P1-6; :PtacI. Fig. 4. The relationship between GC content and CGTase activity of different promoter with different spacer regions. Fig. 5. The DB sequence is a translation enhancer site on the 3' side of the initiation codon (ATG). A: 23

the SDS-PAGE image of several fermentation supernatants after 90 hours of fermentation of strains containing different DB sequences. B: the codon composition of several different DB sequences, the first line represents the DB sequence of the original strain; the second/third/fourth row represents DB sequence of the DB1 / DB2 / DB3 strains respectively; the initiation codon is marked by an underscore,

PT

ED

M

A

N

U

SC R

supernatant protein concentration of the original strain, DB1, DB2, DB3 strains.

IP T

and the DB sequence is shown in uppercase letters. Fig.(C) CGTase activity and fermentation

A

CC E

(Fig. 1)

24

A ED

PT

CC E (Fig. 2)

25

IP T

SC R

U

N

A

M

A ED

PT

CC E (Fig. 3)

26

IP T

SC R

U

N

A

M

A ED

PT

CC E (Fig. 4)

27

IP T

SC R

U

N

A

M

A

(Fig. 5)

28

ED

PT

CC E

IP T

SC R

U

N

A

M

Table 1. Primers used in this study

Sequence (5’→3’)

P1

GGAAGATCTGAATTCAAGGCGCACTCCC

P2

GGAATTCCATATGTCGATACCTCCTTTTTACG

P3

GGAAGATCTGCGCAACGCAATTAATGTGA

P4

GGAATTCCATATGCATAGAATTCTGTTTCCTGTGTGAAA CGCGGATCCGCAATTTTTATAGTATCTGATACGC

CGT R1

CCGCTCGAGTTAGTTCTGCCAATCCACTATAAT

CGT F

CGCGGATCCGCAATCTTCATCGTGTCCGACA

CGT F

CCGCTCGAGTTAATTCTGCCAATCCACGATAATT

ACANNNNNNNNNNNNNNNNTATAATGTGTGGAATT

R1

ATANNNNNNNNNNNNNNNNTGTCAACAGGGGGAT

F2

ACAATTAATCATTCATGTCTATAATGTGTGGAATT

R2

ATAGACATGAATGATTAATTGTCAACAGGGGGAT

F3

ACAATTAATCATCCATGCATATAATGTGTGGAATT

R3

ATATGCATGGATGATTAATTGTCAACAGGGGGAT

F4

ACAATTAATCATCGATGCATATAATGTGTGGAATT

R4

ATATGCATCGATGATTAATTGTCAACAGGGGGAT

F5

ACAATTAATCATCGGTGCATATAATGTGTGGAATT

N

A

M

ED

PT

CC E

F6

U

F1

R5

R6

A

SC R

CGT F1

IP T

Primer

ATATGCACCGATGATTAATTGTCAACAGGGGGAT ACAATTAATCATCGGCGCATATAATGTGTGGAATT ATATGCGCCGATGATTAATTGTCAACAGGGGGAT

dbF

GAAACAGAATTCTATGAAATANNNNNNNNNNNNNGCTGCTGCTGGTCTGC TGCTCCTCGCTGCCCA

dbR

AGACCAGCAGCAGCNNNNNNNNNNNNNTATTTCATAGAATTCTGTTTCCT GTGTGAAATTGTTAT

P1 and P2 were used to amplify the Ptrp promoter gene, P3 and P4 were used to amplify the PlacUV5 gene. CGT F1 and CGT R1 were used to amplify the CGTase gene form the genome of G.

stearothermophilus str.NO.2. The restriction sites used and its protective bases are underlined. CGT F and CGT R were used to amplify the Codon-optimized CGTase gene. Primers F1 and R1 were used to randomly arrange the promoter spacer sequence. F2 and R2 were used to amplify 29

plasmids containing the promoter of the mutant spacer sequence 1-2. The other primers amplify the corresponding plasmids by the same method. Underlined parts in F2to R6 are mutated sequences in spacer regions. dbF / dbR were used to amplify plasmids containing different DB sequences. 13 degenerate bases in the middle of the primers replace the original +9 to +21 codon sequences. (Base "N" indicates that the position may be A / T / C / G.)

Promoter

DNA Sequence (5’→3’) GTTGACAATTAATCATCGGCTCGTATAATGTGTGGAA

P1-1

GTTGACAATTAATCATGCATGTATATAATGTGTGGAA

P1-2

GTTGACAATTAATCATTCATGTCTATAATGTGTGGAA

P1-3

GTTGACAATTAATCATCCATGCATATAATGTGTGGAA

P1-4

GTTGACAATTAATCATCGATGCATATAATGTGTGGAA

P1-5

GTTGACAATTAATCATCGGTGCATATAATGTGTGGAA

P1-6

GTTGACAATTAATCATCGGCGCATATAATGTGTGGAA

N

U

SC R

PtacI

IP T

Table 2.DNA sequences of several mutant spacer regions

A

CC E

PT

ED

M

A

The DNA sequences containing the spacer regions of the mutated regions are shown in the Table in which the dotted line portions are the -35 sequence and the -10 sequence, and the solid line portions are mutated positions.

30