Solubility studies, rational amino acid replacements and structural analyses of streptomyces jumonjinensis isopenicillin n synthase 2

And often time, the recombinant protein may end up being insoluble when it is highly expressed, just like in the case of sjIPNS.. coli BLStar were expressed at 37°C and 25°C and the res

Chapter 4 Results

4.2 Investigation of factors that may influence expression of sjIPNS in E coli

The ability to rapidly produce high levels of recombinant proteins is a critical link between the discovery of new genes and the identification of targets for drug development Presently, the expression and purification of protein targets greatly lag behind the enormous new gene sequences generated Understandably, the bottleneck is due to the absence of a universal expression system that can efficiently express vast number of proteins with different characteristics And often time, the recombinant protein may end up being insoluble when it is highly expressed, just like in the case of sjIPNS As such, sjIPNS becomes a suitable model protein for us to study ways to overcome insolubility problem during protein expression For a start, this section focuses on modification of various expression conditions to possibly achieve high-level production of soluble sjIPNS These factors include the use of an alternative host strain, fusion partners and wider range of induction temperatures To ensure that the expression

results were comparable, non-recombinant control cultures and recombinant E coli cultures

were grown to O.D.600 of ~1.2-1.3, induced with 1mM IPTG and incubated for 15 hours With the exception of experiments for studying effects of induction temperature, all other experiments were carried out at 37°C and 25°C, the two reference temperatures used for grouping IPNS isozymes based on their solubility at these temperatures

4.2.1 Host strain

E coli harbors proteases that are liable to degrade foreign proteins and cause aggregate formation (Gottesman, 1990; Enfors, 1992) Consequently, different strains of E coli

possessing unique sets of proteases can influence the foreign protein expressed to different

extent (Kenealy et al., 1987; Obukowicz et al., 1992) To circumvent this limitation, a common approach is to use E coli strains deficient in certain proteases to allow accumulation

of recombinant proteins at high rates by decreasing the chances of protease degradation during

Chapter 4 Results

expression The E coli strain, BL21(DE3), used in the expression of different IPNS isozymes

in Section 4.1.4 does not produce the lon and OmpT proteases Although large amounts of

soluble enzymes of bacterial scIPNS, sfIPNS, nIPNS and sIIPNS and fungal cIPNS have been

produced in E coli BL21(DE3) at the respective optimal temperatures (Fig 4.9), sjIPNS

protein overexpressed in this strain remains insoluble

The wide array of commercially available E coli deficient strains for protein expression has allowed us to use an alternative strain, E coli BLStar, to examine whether this choice of host strain can increase the production of soluble sjIPNS In addition to lon and OmpT mutations, E coli BLStar also carries mutation in gene coding for RNaseE which is

responsible for degradation of mRNAs The lack of RNaseE would reduce the susceptibility of heterologous mRNAs to degradation, resulting in more heterologous mRNAs being available for protein translation and hence has been proposed to enhance protein production The genotype of the strain is described in Table 3.2

To facilitate this study, recombinant construct pET-SJ was transformed into E coli BLStar The recombinant and non-recombinant E coli BLStar were expressed at 37°C and

25°C and the respective protein fractions were analyzed by SDS-PAGE (Fig 4.11a) No overexpressed protein band was observed in both the soluble (S) and insoluble (I) fractions of

the non-recombinant E coli BLStar control culture and hence only the soluble fraction was

shown in lane marked C1 of Fig 4.11aI Majority of the 37kDa sjIPNS overexpressed band

was associated with the insoluble fractions of E coli BLStar (Fig 4.11aI) The expression results of sjIPNS in E coli BL21(DE3) have already been discussed in Section 4.1.4 Here, the results are presented alongside with the E coli BLStar expression results for comparison (Fig 4.11aII) The percentages of insoluble sjIPNS expressed in E coli BLStar were measured by

densitometric scanning (Fig 4.11a) A slightly higher level of insoluble sjIPNS was expressed

in E coli BL21(DE3) compared to E coli BLStar, however, the difference is only marginal In

both hosts, only ~2-5% of soluble sjIPNS were detected at 37°C and 25°C

Chapter 4 Results

105

Fig 4.11 Investigation of factors that may influence expression of soluble sjIPNS in E

coli

The table below indicates the respective recombinant E coli cultures used for expression studies of sjIPNS (a) in different E coli host strains; (b) as NusA and GST fusion proteins and (c) at wider

range of induction temperatures The first lane (marked M) in all gels shows the protein standards

of various molecular sizes (kDa) The second lane marked C1 in (a)I shows the soluble fraction of

E coli BLStar The second lane marked C2 in all gels except (a)I shows the soluble protein fraction

of E coli BL21(DE3) The respective soluble and insoluble protein fractions were marked S and I

accordingly The arrows indicate the positions of sjIPNS expressed The percentages of sjIPNS expressed in various protein fractions were measured using densitometric scanning and the values obtained were plotted in the charts shown

(a) Host strain

pET-SJ/ BL21Star (a) Host strain

pGST-SJ/ BL21(DE3) (b) Fusion protein

I E coli BLStar

% of sjIPNS expressed in different E coli hosts at 37°C and 25°C

0

5

10

15

20

25

30

35

37°C 25°C 37°C 25°C

E coli BLStar E coli BL21(DE3)

M C1 S I S I kDa M C2 S I S I 25°C

kDa

75

75

25

25

37kDa

37kDa

Chapter 4 Results

106

Fig 4.11 (ctd.)

(b) Fusion protein

M C1 S I S I M C1 S I S I

250

150

150

100

100

75

50

75

50

37

37

25

25

92 kDa

55kDa

III GST-sjIPNS

M C1 S I S I kDa M C1 S I S I

kDa

150

100

75

50

115

37

25

83.0

29.0

49.4

34.6

IV GST

63 kDa

26kDa

0

5

10

15

20

25

30

35

37°C 25°C 37°C

37°C 37°C

Chapter 4 Results

107

Fig 4.11 (ctd.)

(c) Induction temperature

0

5

10

15

20

25

30

35

Induction temperature

20.4

kDa

115

83.0

49.4

34.6

29.0

M C2 37°C 30°C 28°C 25°C 22°C 20°C 18°C 15°C

Soluble protein fractions

37kDa

% of soluble sjIPNS expressed in E coli BL21(DE3) at different

induction temperatures

(°C)

Chapter 4 Results

4.2.2 Fusion protein

Another approach that has achieved success in producing soluble heterologous

proteins in E coli is the use of fusion partners (Section 2.8.2) This has been attributed to the high levels of expression and highly soluble natures of the fusion partner proteins used (Smith

et al., 1988; Davis et al., 1999) However, studies have reported that some fusion partners are much better solubilizing agents than others (Kapust and Waugh, 1999; Wang et al., 1999; Tatsuda et al., 2001) In this Section, studies were undertaken to investigate the efficiencies of

two fusion partners, GST and NusA, in overcoming the low solubility of sjIPNS Both GST

and NusA have been reported to improve the solubility of some foreign proteins that would

otherwise be expressed as insoluble aggregates (Ray et al., 1993; Nygren et al., 1994; Bill et al., 1995; Chang et al., 1997; Davis et al., 1999) To carry out this investigation, sjIPNS was subcloned into fusion vectors, pGK and pET43.1a, that carried genes coding for Schistosoma japonicum GST and E coli NusA respectively The detailed plasmid maps of these expression

vectors are shown in Appendix III

4.2.2.1 PCR amplification of sjIPNS to create suitable flanking restriction

enzyme sites for subcloning into pGK and pET43.1a fusion vectors

The creation of GST-sjIPNS and NusA-sjIPNS fusion constructs requires the cloning

of sjIPNS gene downstream to the 3’ ends of the coding genes of the respective fusion partners Detailed examination of pGK and pET43.1a vector sequences revealed that the ATG start codon of sjIPNS could be precisely inserted in frame with the GST and NusA coding

sequences via the use of a BamHI restriction site located near the 3’ends of both fusion protein genes (Fig 4.12) For the 3’ cloning site, the same XhoI restriction site was chosen for the

construction of both GST-sjIPNS and NusA-sjIPNS fusion constructs pET-SJ plasmid

carrying the sjIPNS gene (Section 4.1.2.2) was used for the subcloning experiment However,

Chapter 4 Results

the flanking regions of sjIPNS gene in pET-SJ do not harbor the selected restriction enzyme

sites intended for subcloning Therefore, PCR amplification was used to create the BamHI and XhoI restriction enzyme sites in the flanking regions of sjIPNS using specially designed

primers OL187 and OL188 The sequences of the primers were shown in Appendix II and the optimal reaction conditions used in PCR amplification were the same as that specified in Table

3.5 The ~1 kb amplified BamHI/XhoI flanked sjIPNS product obtained from the PCR reaction

was resolved by agarose gel electrophoresis as shown in Fig 4.13a The primers designed were very specific since only one distinct amplified product of the correct size was obtained The amplified product was purified from the agarose gel and sequencing was performed to ensure that the insert corresponds to sjIPNS gene

4.2.2.2 Subcloning of sjIPNS into pGK and pET43.1a fusion vectors

A schematic diagram depicting the cloning strategy of sjIPNS into the two fusion vectors is shown in Fig 4.12 The sjIPNS PCR product was subsequently subcloned into pGEM-T Easy vector (Promega) (Clark, 1988) BamHI and XhoI double digestion was

performed to release the cloned gene insert for ligation to the corresponding sites in pGK and pET43.1a vectors The resultant recombinant pGK and pET43.1a fusion constructs carrying

the sjIPNS gene were named pGST-SJ and pNUSA-SJ respectively Restriction enzyme

digestion was done to affirm that the recombinant constructs contained the cloned gene inserts (Fig 4.13b)

Sequencing of pGST-SJ and pNUSA-SJ using selected sets of primers (Appendix II) was carried out to further confirm the identity of the cloned gene Electropherograms showing the partial sequenced region (nucleotide 71 to 155) of sjIPNS in both constructs are presented

in Fig 4.14 Repeat sequencing of the forward and reverse strands of the cloned sjIPNS in pGST-SJ and pNUSA-SJ showed no random gene mutations have been incorporated during the amplification and subcloning processes

Chapter 4 Results

110

Fig 4.12 Construction of recombinant sjIPNS fusion vectors

Schematic diagrams showing the subcloning strategy to construct recombinant fusion vector (a) pGST-SJ and (b) pNUSA-SJ

PCR amplification of sjIPNS

sjIPNS

pET-SJ 6300bp

With primers OL187 and OL188

sjIPNS

BamHI Xhol

Cloned into pGEM®-T Easy vector

BamHI/ Xhol

tac GST

BamHI Xhol

BamH1 Xhol

sjIPNS

BamH1/ Xhol

5189bp

6179bp pGST-SJ

BamHI/ Xhol BamHI

Xhol

T7

7275bp

sjIPNS

BamH1 Xhol

8265bp pNUSA-SJ

Chapter 4 Results

111

Fig 4.13 Gel electrophoresis of PCR amplified sjIPNS containing BamH1/Xho1

flanking ends and restriction enzyme digestion of recombinant fusion vectors

(a) The amplified sjIPNS product (~1kb) with BamH1/Xho1 flanking ends was separated by gel electrophoresis in lane 2 Lane 3 is the PCR control reaction without the addition of DNA template Lane 1 shows the λHindIII DNA marker (b) Gel electrophoresis showing the results for the restriction enzyme digestion to confirm that the recombinant pGST-SJ and pNUSA-SJ contain the sjIPNS insert Lanes 1 and 4 show the λHindIII DNA marker Lanes 2 and 3 show the products of BamH1 digested

pGK and pGST-SJ whereas lanes 5 and 6 show the products of BamH1 digested pET43.1a and

pNUSA-SJ

(a)

(b)

1 2 3

~1kb

6.2kb 5.2kb

8.3kb 7.3kb

23.1 9.4 6.6 4.4 2.2 2.0 0.5

23.1

9.4

6.6

4.4

2.2

2.0

0.5

23.1 9.4 6.6 4.4 2.2 2.0 0.5

Fig 4.14 Sequence confirmation of cloned sjIPNS in recombinant pGK and pET43.1a fusion vectors

Detailed diagram showing the relative position of IPNS insert with respect to the promoter and ribosome binding site (RBS) in (a) pGK and (c) pET43.1a (b) The electropherograms showing the partial sequenced regions (nucleotide 71-155) of the cloned sjIPNS gene in recombinant pGST-SJ and pNUSA-SJ fusion vectors are presented

in (i) and (ii) respectively

(ii) sjIPNS in pNUSA-SJ

Nucleotide 71-155

tac promoter lac operator RBS

IPNS insert

GGATCC

BamHI

translation start site

XhoI

CCTAGG

GST

ATG

T7 promoter lac operator RBS

IPNS insert

GGATCC

BamHI

translation start site

XhoI

CCTAGG

NusA

ATG Nucleotide 71-155

Chapter 4 Results

4.2.2.3 Expression of GST-sjIPNS and NusA-sjIPNS fusion proteins in E coli

BL21(DE3)

The recombinant fusion plasmids, pGST-SJ and pNUSA-SJ, were transformed into E coli BL21(DE3), thereafter the recombinant cultures were induced at 37°C and 25°C to examine the solubilizing effects of both fusion partners on sjIPNS At the same time, pGK and pET43.1a vector carrying the GST and NusA coding sequences respectively, were also

transformed into E coli BL21(DE3) to be used as controls for the expression studies under the

same conditions The induction and harvesting of recombinant cultures for cell-free extract preparations were performed as described in Section 3.2.9

GST and NusA proteins are 26kDa and 55kDa respectively, hence the expected sizes

of GST-sjIPNS and NusA-sjIPNS fusion proteins are ~63kDa and ~92kDa In recombinant vector pNUSA-SJ, sjIPNS is subcloned downstream of the NusA coding sequencing and the transcription of the NusA-sjIPNS fused gene product is under the control of T7-promoter (Fig 4.14) In pET-SJ, sjIPNS is similarly cloned downstream of the T7-promoter but is expressed

as a non-fused protein (Fig 4.2) Hence, comparison of the expression of sjIPNS in pNUSA-SJ

and pET-SJ can reveal the effect of NusA-fusion on sjIPNS expression in E coli The expression results of non-fused sjIPNS under the control of T7-promoter in E coli BL21(DE3)

have already been discussed in Section 4.2.1 Here, the results were again presented alongside with the expression results of fused NusA-sjIPNS protein in the tabulated chart for comparison (Fig 4.11b) At 37°C and 25°C, majority of the non-fused sjIPNS was expressed in the insoluble fractions Densitometric scanning results showed that ~29-33% of non-fused sjIPNS was expressed in the insoluble fractions at both temperatures Interestingly, when sjIPNS was

fused to NusA, an estimated 2 to 3-fold reduction in the synthesis of sjIPNS in E coli was

observed Only about 10-16% of NusA-sjIPNS fusion protein was observed in the insoluble protein fractions (Fig 4.11b) Nonetheless, observable amounts of soluble NusA-sjIPNS fusion protein up to ~8% of the total soluble proteins was obtained at 25°C The expression of NusA