Based on suggestions from other workers (Pongs, 1994), an expression time o f 5 hours and an [IPTG] o f 0.1 mM were used in the expression o f G S T / K \ 1 . 1 Nt fusion protein.
Subsequent variations on these parameters did not improve yield. Initial expressions o f K v L I N t were performed in order to establish whether the fusion protein would be expressed in large enough quantities to be easily visualised by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) (Section 3.2 4.2), and what type o f extraction method would be needed to recover the protein.
It is common for proteins over-expressed in bacterial cells to be incorporated into insoluble cytoplasmic granules called inclusion bodies which, once the cells are lysed, can be solubilised using agents such as detergents or 8 M urea (Marston et a i, 1984; Schoner et a i, 1985). A lysis and extraction method was used which would, on examination o f the end products, provide information on levels o f protein expression and on whether or not the protein was being incorporated into inclusion bodies. This was a method based on urea-solubilisation o f inclusion bodies, and was adapted from existing urea-based protocols (Sambrook et a l,
E X P R E S S I O N A N D S I X O N D A R Y S T R U C T U R E Ol R E C O M B I N A N T Kv l . I N - T I :R M I N A L F R A G M E N T 3.2.4.1 Initial expression and extraction of GS 17K\ 1.1 N-terniinal fragment fusion protein
JM109 cells transformed with pKyl. IN DNA, stabbed either from bacterial plates or glycerol stocks, were used to inoculate 100 ml o f autoclaved LB medium, pre-warmed to 37 “C, in a 2 litre flask. The medium was supplemented with ampicillin to a concentration o f 100 pg/ml. This culture was allowed to grow overnight in a 220 rpm orbital shaking incubator {Gallenkamp) at 37 °C.
After overnight incubation the culture was diluted ten-fold with autoclaved, pre-warmed LB medium containing 100 pg/ml ampicillin, and distributed into several 2 litre flasks, 400 ml per flask. To allow adequate oxygen for the optimum growth o f the bacteria, the culture was not permitted to take up more than 20 % o f the volume o f the flask. Cells were incubated until exponential growth phase was reached. At this point, fiision protein expression was induced by addition o f IPTG to a final concentration o f 0.5 mM. After a ftnther 5 hours incubation at 37 °C, the cells were pelleted by centrifugation at 7000 x g for 7 minutes in a Sorvall RC 28S centrifuge at 4 °C. The supernatant was discarded and the pellet consisting o f bacterial cells was resuspended in 20 ml ice-cold PBS (150 mM NaCl, 16 mM NazHPO;, 4 mM N aH2P0 4; Smith & Johnson, 1988). The cells were then lysed by mild sonication on ice. Following lysis the cells were centrifuged at 10,000 x g for 10 minutes at 4 X . This produced a supernatant containing PBS-soluble proteins which was kept for subsequent SDS-PAGE analysis o f protein content.
The pellet from this centrifugation step contained PBS-insoluble proteins, cell membranes and any inclusion bodies. This pellet was thoroughly resuspended in 16 ml 8 M urea containing 1 mM dithiothreitol (DTT). Urea was used at a concentration o f 8 M, as explained above, to solubilise and break open inclusion bodies. DTT was included as a reducing agent to break up disulphide bonds which can fonn between cysteine residues in oxidising conditions and which can lead to protein aggregation. The resulting mixture was sonicated on ice for 1 minute then homogenised in a glass vessel using a motorised homogeniser, before being left on ice for 30 minutes to facilitate maximal solubilisation o f inclusion bodies and proteins therein.
E X P R E S S I O N A N D S E C O N D A R Y SI'R UC I'URI: Ol R H C O M H I N A N T Kv l . 1 N -T E R M IN A L F R A G M E N T The homogenate was centrifuged at 10,000 x g for 5 minutes at 4 "C in order to sediment urea-insoluble proteins and cell debris. The resulting supernatant, containing urea- soluble proteins, was analysed by SDS-PAGE for protein content. The expected molecular mass o f the fusion protein is 45 kD. This is composed o f 27.5 kD for the GST and 17.5 kO for the N-terminal fragment (aa 14-162).
3.2.4.2 SDS P A G E w ith C oom assie blue s ta in in g
SOS PA G E was performed essentially according to the method o f Laemmli (Laemmli, 1970), using 1.5 mm thick 10, 12 and 15 % slab gels. All materials were “Electran” grade {BDH) or electrophoresis purity reagents from Bio-Rad.
Protein samples were solubilised in a sample buffer (50 mM Tris-HCl pH 6.8, 100 mM DTT, 2 % (w/v) SDS, 0.1 % (w/v) bromophenol blue, 10 % (v/v) glycerol) and heated to 100 °C for 5 minutes. Aliquots o f the protein samples (5-50 pi) were separated in each lane o f a 10 cm separating gel (12, 15 or 18 % acrydamide / 0.27, 0.32, 0.37 % bisacrylamide in a buffer o f 375 mM Tris-HCl pH 8.8 plus 0.1 % SDS). The gel was polymerised by the addition o f 0.1 % (w/v) ammonium persulphate and 0.016 % (v/v) N N N ’N ’ tetramethylethylenediamine (TEMED). In order to achieve such separation, protein samples were loaded into individual wells o f a 1.5 cm stacking gel (3 % acrylamide / 0.08 % bisacrylamide in a buffer o f 125 mM Tris-HCl pH 6.8 containing 0.1 % (w/v) SDS), polymerised by 0.1 % ammonium persulphate and 0.005 % (v/v) TE M E D ) which had been overlaid on the separating gel. The electrode buffer comprised 25 mM Tris, 190 mM glycine and 0.1 % (w/v) SDS, pH 8.3. Electrophoresis was performed at 25 mA (constant current) over 2 hours using a Bio-Rad mini-gel system. To enable the molecular weights o f the samples to be estimated molecular weight markers {Pharmacia; Novagen) were electrophoresed in parallel to the samples.
For Coomassie blue staining o f the proteins, the gels were soaked for 1 hour in Coomassie blue stain (0.25 % (w/v) Coomassie Brilliant Blue R250, 45 % (v/v) methanol, 10 % (v/v) glacial acetic acid) immediately followed by destaining in 10% (v/v) acetic acid 10 % methanol (v/v).
E X P R E S S I O N A N D S E C O N D A R Y S T R U C T U R E OF R E C O M B I N A N T K v l . I N - T E R M I N A L F R A G M E N T 3.2.4.3 Results of initial expression and extraction of G S T /K \ 1.1 N-terminal fragm ent fusion protein
SDS-PAGE analysis o f the PBS-soluble and 8 M urea-soluble fractions from the initial fusion protein extraction (Figure 3.3) clearly show over-expression o f the fusion protein, which appears as a band at 45 kD. This band was densest in the urea-soluble fraction. Also apparent are bands at 17 kD and 27 kD which probably correspond to K v l.I N-terminal fragment and GST respectively. These most likely arise from partial proteolytic cleavage o f the target fusion protein by bacterial proteases which recognise the engineered thrombin cleavage site.
46 kD
21.5 kD
Figure 3.3: SD S-PAG E analysis o f expression and extraction o f G S T /K y l.l N-terminal fragment
SD S-PA G E gel o f products o f large-scale G S T /K y l.l N -ten n in al fragm ent expression and urea extraction. L ane 1: m olecular w eight m arkers: lane 2: PBS w ash o f intact cells: lane 3: PB S-soluble fraction follow ing cell lysis: lane 4: 8 M urea soluble fraction follow ing cell lysis: lane 5: 8 M urea insoluble fraction follow ing cell lysis.
T he band at 45 kD contains Cooniassic-bluc stained G S T / K y l.l N -len n in al fragm ent fusion protein. Tlie bands at 17 kD and 27 kD rcspecti\ely probably contain K y i. 1 N term inal fragm ent and G ST respectively.
E X P R E S S I O N A N D S E C O N D A R Y S T R U C I URI: Ol- R E C O M l il N A N T Kv l . I N - T E R M l N A L i- R A G M E N T 3.2.5 Purification of recombinant G ST /K \ 1.1 N-terminal fragment fusion protein
Purification o f fusion protein away from the bacterial proteins involved binding the fusion protein to glutathione affinity beads (Glutathione Sepharose 4B, Si^niia). The GST portion of the fusion protein interacts with and binds to glutathione immobilised on the beads. Bacterial proteins can then be washed away leaving only fusion protein attached to the beads. The fusion protein can then be eluted whole by washing the beads with free glutathione. Alternatively the non-GST portion o f the fusion protein, in this case the K y l . l N-terminal fragment, can be cleaved from the GST using thrombin
Before binding to the beads, the urea was dialysed from the urea-solubilised lysate containing urea-denatured proteins. This step was included because the fusion protein will not bind to the glutathione beads in a denatured state, as the GST portion must be in its native folded state to recognise the glutathione substrate It was also important because the structure o f the
K y l . l N-terminal fragment was to be investigated and so had to be renatured in order to adopt its native folded state.
3.2.5.1 Refolding of recombinant G ST/K \ 1.1 N-terminal fragment fusion protein
Once the urea-soluble lysates from several large-scale expressions o f G S T/K yl.l N-terminal fragment had been collected and stored at -20 ”C, they were thawed, pooled and then dialysed to remove the urea. In order to return the urea-denatured protein back to a native folded state, dialysis was perfonned gradually with a number o f steps. Bringing the protein straight from 8 M urea to PBS was likely to result in the protein coming out o f solution. Dialysis tubing with a molecular mass cut-off o f 14 kD was used. Tubing was left in boiling water for 15 minutes and then rinsed with ice-cold PBS before the lysate was added and the tubing sealed with dialysis clips.
The lysate was dialysed successively at 4 "C against 2 litres o f PBS containing; 1. 4 M urea + 1 mM DTT
2. 2 M urea + 1 mM DTT 3 . 1 M urea + 1 mM DTT
4. 0.5 M urea + 1 mM DTT 5. 1 mM DTT.
E X P R E S S I O N A N D S E C O N D A R Y S T R U C T U R E Ol R E C O M B I N A N T K v l . I N - T E R M I N A L F R A G M E N T 3.2.5.2 Affinity purification of G ST/K\ 1.1 N-terminal fragment fusion protein using glutathione sepharose
Glutathione sepharose 4B {Si^nui) (0.5 ml per litre ot' bacterial culture) was swollen overnight at 4 “C in 50 ml o f sterile deionised water. The swollen sepharose was transferred to a 20 ml liquid chromatography column (Bio-Rac/) and washed with 10 column volumes o f PBS buffer. The sepharose was resuspended with the dialysed lysate, transferred to a 50 ml Greiner centrifuge tube and incubated at 4 "C overnight. To allow maximum binding o f fusion protein, the sepharose was kept in suspension by rotating the tube on a Denley Spiramix 5.
After overnight incubation the suspension was poured into a 20 ml chromatography column and the unbound material drained through the column membrane thereby allowing the sepharose to accumulate at the bottom o f the column. The sepharose was washed thoroughly with 1 litre o f PBS to remove non-specifically bound proteins and cell debris. The sepharose was adjudged “clean” when the O D2S0 o f the flow through dropped below 0.001 indicating that the non-specifically bound proteins, which absorb strongly in this region, were at an insignificantly low concentration.
Initially, to assess the purity o f the fusion protein at this stage, the bound fusion protein was eluted from the sepharose by competition with 20 mM reduced free glutathione. SD S-P A G E analysis o f the eluate at this stage established that the large majority o f bacterial proteins had been removed from the beads at earlier purification washes, leaving virtually pure fusion protein (see Figure 3.4). In subsequent experiments, pure K v l.I N- terminal fragment was cleaved directly off the GST portion o f the fusion protein whilst the G ST was still attached to the affinity beads.
3.2.5.3 Throm bin cleavage o f G ST/K \ 1.1 N-terminal fragment fusion protein
10 units o f bovine thrombin (Sigma) per 500 ml original culture volume were added to the glutathione beads with bound fusion protein in the affinity column in 1 ml PBS per 1 ml bead volume. The cleavage reaction was performed in the affinity column (sealed at both ends) incubated at room temperature with gentle agitation for 8-16 hours. The extent to which the fusion protein was digested was determined by S D S-P A G E (see
E X P R H S S I O N A N D S J iC O N D A R Y S T R U C T U R E Ol' RE C OM U lN A N 'l' K v i -i N- I'ERMINAE F R A G M E N T Figure 3.4). This revealed that in addition to the expected K\ 1.1 N-terminal fragment band at 17.5 kD molecular mass, there was a contaminant band at 27.5 kD molecular mass. This band was assumed to be GST because o f its apparent molecular mass, but subsequent reincubation o f the protein mixture with fresh glutathione affinity beads did not remove this contaminant which suggests that if the protein was G ST it had become modified at some stage resulting in altered substrate binding properties. In order to remove the contaminant to produce pure K\ l. 1 N-terminal fragment, the protein mixture was purified using reverse phase High Pressure Liquid Chrom atography (HPLC).
3.2.5.4 HPLC purification o f recombinant K\ 1.1 N-terminal fragment
Reverse phase HPLC separates proteins primarily on the basis o f their hydrophobicity. Hydrophobic proteins interact with the solid phase and are thus retained more readily relative to hydrophilic proteins. Trifuoroacetic acid (TFA) was included in the eluant. This solvent facilitates the separation o f non-covalently bound proteins, and was included in case the N-terminal fragment was interacting with the contaminant under native conditions. Protein sample (500 pi) was injected into a Vydac Cs (208TP54) column equilibrated in 70 % (v/v) buffer A (0.1 % (v/v) aqueous TFA) and 30 % (v/v) buffer B (90 % (v/v) acetonitrile, 10 % (v/v) buffer A). The proteins were separated on a linear gradient o f buffer B, from 30 to 60 % (v/v) in 120 minutes, at a f o w rate o f 1 ml/minute. Elutions were monitored for protein content by measuring absorbance o f light at 280 nm. The eluted protein fractions were freeze dried and analysed by SD S-PA G E (Section 3.2.4.2). For samples intended for structural analysis, eluted N-terminal fragment protein fraction was freeze dried in the presence o f 0.1 M HCl in order to remove residual traces o f TFA which can interfere with spectroscopic analysis o f proteins.
The FIPLC 280 nm absorbance trace (Figure 3.5 a) showed one main peak which was identified using SD S-PA G E to be pure Kvl . I N-terminal fragment. A much smaller peak, corresponding to protein eluted at more hydrophobic conditions, was identified using SD S-PA G E as the 27 kD contaminant (probably GST).
SD S-PA G E analysis o f the pooled HPLC-purified N-terminal fragment peak fractions from several HPLC runs (Figure 3.5 b) showed a single band at an apparent molecular
E X P R E S S I O N A N D S E C O N D A R Y ST R U C TURE Ol K lX 'O M D lN A N i' Kv l . I N - T E R M I N A L F R A G M E N T mass o f 17 kD indicating that the N-terminal fragment was sufllciently pure at this stage for structural analysis purposes.