Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination

G Model

ARTICLE IN PRESS

JVAC 17312 1–7

Vaccine xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination

1

2

Q1

3

Yukihiro Tamaki, Tetsuya Harakuni, Rui Yamaguchi, Takeshi Miyata 1 , Takeshi Arakawa ∗ Laboratory of Vaccinology and Vaccine Immunology, Center of Molecular Biosciences, University of the Ryukyus, 1 Senbaru, Nishihara, Okinawa 903-0213, Japan

4 5 6

7 20

a r t i c l e

i n f o

a b s t r a c t

8 9 10 11 12 13

Article history: Received 9 September 2015 Received in revised form 8 December 2015 Accepted 17 January 2016 Available online xxx

14

19

Keywords: Cholera toxin B subunit Pentamer Multisubunit protein refolding Dialysis

21

1. Introduction

15 16 17 18

The cholera toxin B subunit (CTB) is secreted in its pentameric form from Escherichia coli if its leader peptide is replaced with one of E. coli origin. However, the secretion of the pentamer is generally severely impaired when the molecule is mutated or fused to a foreign peptide. Therefore, we attempted to regenerate pentameric CTB from the inclusion bodies (IBs) of E. coli. Stepwise dialysis of the IBs solubilized in guanidine hydrochloride predominantly generated soluble high-molecular-mass (HMM) aggregates and only a small fraction of pentamer. Three methods to reassemble homogeneous pentameric molecules were evaluated: (i) using a pentameric coiled-coil fusion partner, expecting it to function as an assembly core; (ii) optimizing the protein concentration during refolding; and (iii) eliminating contaminants before refolding. Coiled-coil fusion had some effect, but substantial amounts of HMM aggregates were still generated. Varying the protein concentration from 0.05 mg/mL to 5 mg/mL had almost no effect. In contrast, eliminating the contaminants before refolding had a robust effect, and only the pentamer was regenerated, with no detectable HMM aggregates. Surprisingly, the protein concentration at refolding was up to 5 mg/mL when the contaminants were removed, with no adverse effects on refolding. The regenerated pentamer was indistinguishable in its biochemical and immunological characteristics from CTB secreted from E. coli or choleragenoid from Vibrio cholerae. This study provides a simple but very efficient strategy for pentamerizing CTB with a highly homogeneous molecular conformation, with which it may be feasible to engineer CTB derivatives and CTB fusion antigens. © 2016 Published by Elsevier Ltd.

Q2 22 23 24 25 26 27 28 29

Cholera toxin (CT) is a prototypical AB5 enterotoxin, and its B subunit (CTB) becomes fully biologically functional when it adopts a pentameric form. CTB is a component of cholera and enterotoxigenic Escherichia coli vaccines [1,2], and is also expected to have other medical and veterinary applications. Therefore, methods to yield high levels of the biologically active CTB pentamer and its derivatives, including mutants [3] and fusion proteins [4–6], are required.

Abbreviations: COMP, cartilage oligomeric matrix protein; CT, cholera toxin; CTA, cholera toxin A subunit; CTB, cholera toxin B subunit; GdnHCl, guanidine hydrochloride; GM1, monosialoganglioside GM1; HMM, high-molecular-mass; IBs, inclusion bodies; RP, refolding procedure. ∗ Corresponding author. Tel.: +81 98 895 8974; fax: +81 98 895 8944. E-mail address: [email protected] (T. Arakawa). 1 Present address: Department of Biochemistry and Biotechnology, Faculty of Agriculture, Kagoshima University, 1-21-24 Korimoto, Kagoshima 890-0065, Japan.

The production of recombinant CTB has been intensively investigated in secretory expression systems, such as Vibrio cholerae [7,8], E. coli [9–11], and yeasts [12]. However, the modification of CTB severely impairs its secretion from host organisms [3,4,12]. An alternative to its expression as a secreted protein is the regeneration of the biologically active pentamer from inclusion bodies (IBs) with in vitro protein refolding. The recovery of biologically active proteins after protein refolding with dilution, solvent exchange, reversible adsorption to a solid support, or pressure treatment [13,14] has been an area of intensive research, because the final yields can significantly exceed those achieved with intra- or extracellular soluble protein expression if the refolding efficiency is optimized. There are also other advantages in refolding proteins from IBs, including the ease of separation from the soluble host proteins and low endotoxin contamination [15]. In this study, we evaluated various methods to regenerate the CTB pentamer from E. coli IBs, including its fusion to a coiled-coil domain [16] and modifying the concentration and purity of the protein [17–19]. We then analyzed the biochemical and immunological characteristics of the regenerated pentameric CTB.

http://dx.doi.org/10.1016/j.vaccine.2016.01.034 0264-410X/© 2016 Published by Elsevier Ltd.

Please cite this article in press as: Tamaki Y, et al. Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination. Vaccine (2016), http://dx.doi.org/10.1016/j.vaccine.2016.01.034

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

G Model JVAC 17312 1–7

ARTICLE IN PRESS Y. Tamaki et al. / Vaccine xxx (2016) xxx–xxx

2 50

2. Materials and methods

2.3. Protein refolding

51

2.1. Construction of CTB expression plasmids

The IBs were solubilized by incubation with 6 M guanidine hydrochloride (GdnHCl) in 50 mM Tris buffer (pH 8.2) overnight at 37 ◦ C. The solution was then diluted to 3 M GdnHCl with 50 mM Tris buffer (pH 8.2) and transferred to a dialysis membrane (Spectra/Por CE dialysis tubing, 8–10 kDa MWCO; Spectrum Japan, Shiga, Japan). The protein concentration was adjusted to 0.5 mg/mL at this point, unless otherwise indicated. The membrane was immersed in dialysis buffer (50 mM Tris [pH 8.2], 2 M GdnHCl) for 4 h, and dialysis was continued by transferring the membrane stepwise into the following buffers for the indicated periods: 50 mM Tris (pH 8.2), 1 M GdnHCl, 1 M arginine (Arg) for 4 h; 50 mM Tris (pH 8.2), 0.5 M GdnHCl, 0.5 M Arg overnight; PBS (pH 7.4), 0.5 M Arg for 4 h; PBS (pH 7.4), 0.25 M Arg for 4 h; PBS (pH 7.4), 0.125 M Arg for 4 h; and PBS (pH 7.4) for 4 h. This protein refolding method was designated “refolding procedure 1” (RP1). A round of nickel-affinity purification was added to RP1 before the dialysis steps to remove any contaminants, and this method was designated “refolding procedure 2” (RP2). Nickel-affinity purification was conducted as part of RP2 as follows: the protein solubilized in 6 M GdnHCl was applied to a nickel-affinity column (HisTrap FF Ni Sepharose 6 Fast Flow) that had been pre-equilibrated with binding buffer (50 mM Tris [pH 8.2], 20 mM imidazole, 6 M GdnHCl), and was then washed with the same binding buffer. The bound protein was eluted with elution buffer (50 mM Tris [pH 8.2], 500 mM imidazole, 6 M GdnHCl) and refolded with RP1, as described above. Where “Pentamer-1” and “Pentamer-2” appear in the text, they refer to the mCTB pentamers regenerated with RP1 and RP2, respectively. Arginine was used in protein refolding as a protein-stabilizing agent to inhibit inappropriate intermolecular hydrophobic interactions, which would lead to undesirable protein aggregation [21,22].

52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70

71

72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110

pET expression plasmids encoding mature CTB lacking its leader peptide (mCTB, Thr22 –Asn124 ), mCTB fused to the pelB signal (pCTB), and full-length CTB with its original leader peptide (fCTB) were constructed. This was done with PCR subcloning using a template DNA plasmid carrying the CTB gene (GenBank accession number U25679) and the appropriate primer sets (Supplementary Table 1). The amplified genes were subcloned into the corresponding sites in pET-21d or pET-22b (Merck KGaA, Darmstadt, Germany) (Supplementary Fig. 1a). An expression plasmid encoding a fusion protein combining mCTB and the pentameric coiled-coil domain of cartilage oligomeric matrix protein (COMP; Protein Data Bank: 1VDF) was also constructed. The pET-22b expression plasmid encoding a (G4 S)3 spacer fused to COMP (Gly26 –Gly80 ) [20] was digested with NdeI and NcoI to remove the pelB signal. It was then replaced with a sequence encoding mCTB fused to a GPGP spacer, generating the mCTB–COMP fusion protein expression vector (Supplementary Fig. 2a). The plasmids were introduced into E. coli BL21(DE3) (Merck KGaA) for protein expression.

2.2. Expression analysis of recombinant CTB and its fusion protein Escherichia coli BL21(DE3) transformed with each CTB expression plasmid was cultured at 37 ◦ C in LB medium (with 100 ␮g/mL ampicillin) until the optical density at 600 nm (OD600 ) reached 0.5, at which point gene expression was induced with 1 mM isopropyl ␤-d-1-thiogalactopyranoside (IPTG). The culture supernatant was separated from the cells by centrifugation (9600 × g for 20 min), filtered (FastCap filter with 0.2 ␮m pore size; Nalgene Nunc International, Inc., Rochester, NY, USA), and subjected to batch purification with nickel-affinity resin (Ni Sepharose 6 Fast Flow resin; GE Healthcare, Uppsala, Sweden), according to manufacturer’s protocol. Briefly, the culture supernatant was applied to a 50% resin slurry that had been preequilibrated with binding buffer (phosphate-buffered saline [PBS, pH 7.4], 20 mM imidazole), and then shaken gently with a rotator at 4 ◦ C overnight. The resin was recovered by centrifugation (500 × g for 5 min), washed with the binding buffer, and eluted with elution buffer (PBS [pH 7.4], 500 mM imidazole). The cells were separated from the culture supernatant, as described above, and were disrupted with BugBuster Protein Extraction Reagent (Merck KGaA). The soluble proteins were separated from the IBs by centrifugation (16,000 × g for 20 min) and batch-purified with nickel-affinity resin, as described above. The IBs were dissolved in PBS containing 1% sodium dodecyl sulfate (SDS) by incubating them overnight at room temperature. They were then subjected to batch purification with nickel-affinity resin as before, except that both the binding buffer and elution buffer contained 1% SDS. The affinity-purified protein samples from supernatant, soluble or insoluble cellular fraction were mixed with SDS loading buffer (Tris buffer, 1% SDS, 10% sucrose, bromophenol blue; EzApply; ATTO Corp., Tokyo, Japan). The protein samples without heat treatment were then analyzed with SDS-polyacrylamide gel electrophoresis (PAGE)/Coomassie Brilliant Blue (CBB) staining or western blotting (WB). The protein samples were also analyzed with a monosialoganglioside GM1 enzyme-linked immunosorbent assay (GM1-ELISA) or sizeexclusion chromatography with a HiLoad 16/60 Superdex 75 pg column (GE Healthcare), as described previously [12,20]. The protein yield was determined with a bicinchoninic acid protein assay (BCA Protein Assay Reagent Kit; Thermo Scientific, Waltham, MA, USA).

2.4. Intranasal immunization of mice Seven-week-old female BALB/c mice (five per group) (Japan SLC, Shizuoka, Japan) were each twice administered 30 ␮g of Pentamer-2, pCTB, or choleragenoid (List Biological Laboratories, Inc., Campbell, CA, USA) without adjuvant, in weeks 0 and 2, via the intranasal (i.n.) route. Their sera were collected in week 3 and the CTB-specific serum immunoglobulin G (IgG) was measured as described previously [23]. Endotoxin was removed from the protein preparations with Detoxi-Gel Endotoxin Removing Gel (Pierce, Rockford, IL, USA). The toxin levels were measured with Limulus Amebocyte Lysate (Pyrogent Single Test Vials; Lonza, Walkersville, MD, USA), and were < 0.05 endotoxin units/mg of protein. The animal experimental protocols were approved by the University of the Ryukyus Institutional Animal Care and Use Committee, and the animal experiments were conducted in accordance with the Institutional Ethical Guidelines for Animal Experiments. 2.5. GM1-binding competition assay A GM1-binding competition assay was performed to compare the GM1-binding capacities of competitors Pentamer-2, pCTB, and choleragenoid with the capacity of CT (choleragen; List Biological Laboratories). The competition assay was performed essentially as described previously for the GM1-ELISA [12], except that a step involving the application of CT to a microtiter plate and subsequent steps were added. In these steps, CT (2 ␮g/mL) mixed with increasing concentrations (1–80 ␮g/mL) of each competitor was applied to the plate and incubated at 37 ◦ C for 1 h. The bound CT was detected with rabbit anti-CTA antiserum (AB43; Advanced Targeting Systems, San Diego, CA, USA), alkaline phosphatase-conjugated anti-rabbit IgG antibody (A2556; SigmaAldrich, St. Louis, MO, USA), and p-nitrophenyl phosphate (Bio-Rad

Please cite this article in press as: Tamaki Y, et al. Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination. Vaccine (2016), http://dx.doi.org/10.1016/j.vaccine.2016.01.034

111

112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141

142

143 144 145 146 147 148 149 150 151 152 153 154 155 156 157

158

159 160 161 162 163 164 165 166 167 168 169 170 171

G Model JVAC 17312 1–7

ARTICLE IN PRESS Y. Tamaki et al. / Vaccine xxx (2016) xxx–xxx

Fig. 1. Expression of CTB constructs in Escherichia coli. (a) CTB constructs (i.e. mCTB, pCTB, and fCTB in Supplementary Fig. 1) were batch-purified from the culture supernatant with nickel-affinity resin (lane S), from soluble cellular extract (lane C), or from inclusion bodies (lane I), as described in the Materials and methods (Section 2.2). The affinity-purified proteins in buffer containing no reducing agent and without heat treatment were then subjected to 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The separated proteins were stained with Coomassie Brilliant Blue (CBB) or subjected to western blotting (WB). M, molecular marker; *, CTB monomer; **, CTB pentamer. (b) Size-exclusion chromatographic separation of nickel-affinity-purified pCTB from the supernatant (solid line) and mCTB refolded from IBs (dashed line). HMM, soluble high-molecularmass.

174

Laboratories Inc., Redmond, WA, USA). OD415 was measured and the inhibition rate (%) was calculated with the equation: (1 – [OD415 of CT with competitor]/[OD415 of CT without competitor]) × 100.

175

172 173

3

Fig. 2. Effect of coiled-coil fusion on CTB pentamer assembly. mCTB and the mCTB–COMP fusion protein were refolded with refolding procedure 1 (RP1), as described in the Materials and methods (Section 2.3). The refolded protein samples in phosphate-buffered saline (PBS) (containing no reducing agent) without heat treatment were then subjected to (a) size-exclusion chromatography, or (b) SDS-PAGE/CBB. M, molecular marker.

software ver. 12; SAS Institute, Inc., Cary, NC). P < 0.05 was considered statistically significant.

192 193

3. Results

194

2.6. GM1-binding inhibition assay

3.1. Expression of mCTB, pCTB, and fCTB

195

187

A GM1-binding inhibition assay was performed to evaluate the capacity of the induced mouse antisera to block the binding of CT to GM1, essentially as described previously for the GM1-ELISA [12], except that a step involving the application of the anti-CTB antiserum to the microtiter plate and subsequent steps were added. In these steps, CT (0.2 ␮g/mL) was mixed and incubated with antiCTB antiserum (diluted 1:10 or 1:20) that had been induced with Pentamer-2, pCTB, or choleragenoid, before it was applied to the plate and incubated continuously at 37 ◦ C for 1 h. The bound CT was detected essentially as described in Section 2.5. The inhibition rate (%) was calculated with the equation: (1 – [OD415 of antiserum]/[OD415 of unimmunized serum]) × 100.

188

2.7. Statistical analysis

The protein expression analysis indicated that mCTB existed exclusively in IBs (Fig. 1a), whereas pCTB was only found in the culture supernatant, exclusively as a pentamer (Fig. 1a), with binding affinity for GM1 (Supplementary Fig. 1c). In contrast, fCTB was found in no compartment analyzed, indicating that the original CTB leader is nonfunctional in E. coli. A size-exclusion chromatographic analysis of the secreted pCTB showed a single peak with a molecular mass of 43 kDa, indicating that it was the homogeneous pentamer (Fig. 1b) with an intact intrasubunit disulfide bond (Supplementary Fig. 1d) [23]. Its N-terminal amino acid sequence, determined with Edman degradation, was MAPQNITDL, indicating the proper removal of the pelB signal. Mice administered pCTB showed serum IgG (Supplementary Fig. 1e) and mucosal IgA titers (Supplementary Fig. 1f) similar to those induced with choleragenoid. These results suggest that pCTB was structurally and immunologically indistinguishable from choleragenoid. The final yield of pCTB was approximately 100 mg/L of bacterial culture.

176 177 178 179 180 181 182 183 184 185 186

189 190 191

The Mann–Whitney U test was used to compare the antibody titers of the unimmunized group and the individual immunization groups, or those of two different immunization groups (JMP

Please cite this article in press as: Tamaki Y, et al. Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination. Vaccine (2016), http://dx.doi.org/10.1016/j.vaccine.2016.01.034

196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213

G Model JVAC 17312 1–7

ARTICLE IN PRESS Y. Tamaki et al. / Vaccine xxx (2016) xxx–xxx

4

Fig. 3. Effect of the protein concentration on CTB pentamer assembly. mCTB at the indicated protein concentrations (0.05–5 mg/mL) was refolded with RP1. The refolded protein samples in PBS (containing no reducing agent) without heat treatment were then subjected to (a) size-exclusion chromatography, or (b) SDSPAGE/CBB. M, molecular marker.

225

As mentioned above, CTB secretion is severely hampered when the protein is modified [3,4,12]. Therefore, we attempted to regenerate pentameric CTB from IBs because pentamer regeneration from IBs may not have been affected by protein modification. Refolding (RP1, described in Section 2.3) the mCTB protein predominantly generated soluble HMM aggregates (Fig. 1b). However, interestingly, like pCTB and choleragenoid, the mCTB monomer refolded with RP1 already contained an intact intrasubunit disulfide bond (Supplementary Fig. 1d). This was unexpected and implies that devising a proper refolding method should regenerate homogeneous pentamer from IBs because the intrasubunit disulfide is a prerequisite for CTB pentamerization [24].

226

3.2. Pentamer regeneration with protein refolding

214 215 216 217 218 219 220 221 222 223 224

227 228 229 230 231 232 233 234 235 236

To improve the regeneration of the pentamer from IBs, three methods were evaluated: (i) using a pentameric coiled-coil fusion, which may facilitate the orderly assembly of the CTB pentamer [16]; (ii) optimizing the protein concentration during protein refolding [17]; and (iii) removing all contaminants before refolding [18,19]. First, to evaluate effect of an ␣-helical coiled-coil fusion on CTB pentamerization, the COMP coiled coil was genetically fused to CTB (Supplementary Fig. 2), based on our recent finding that a CTB–COMP fusion protein expressed in Pichia pastoris displayed robust physicochemical stability [16], implying that an ␣-helical

Fig. 4. Effect of protein purity on CTB pentamer assembly. mCTB at the indicated protein concentrations (0.05–5 mg/mL) was refolded with RP2, as described in the Materials and methods (Section 2.3). The refolded protein samples in PBS (containing no reducing agent) without heat treatment were then subjected to (a) size-exclusion chromatography, or (b) SDS-PAGE/CBB. M, molecular marker.

coiled coil may function as a CTB-pentamer-stabilizing scaffold. Therefore, the coiled coil was also expected to facilitate CTB pentamer reassembly from IBs in vitro. Supporting this expectation, the B subunit of Shiga toxin type 2 (Stx2B), when fused to the COMP coiled coil, was efficiently reassembled from E. coli IBs with dialysis, whereas the unfused Stx2B was not, instead forming insoluble aggregates (unpublished data). The mCTB–COMP fusion protein was expressed in the IBs of E. coli, as observed for mCTB. However, the fusion protein formed pentamers after refolding, and the ratio of pentamer to soluble HMM aggregate was higher than for mCTB (Fig. 2a). Thus, the ␣-helical coiled coil facilitated CTB pentamerization to some extent, but the process still required improvement. The SDS-PAGE results (Fig. 2b) indicated a relatively uniform pattern of pentameric bands for both mCTB and its coiled-coil fusion protein. However, it is likely that the HMM aggregates disassembled into pentamers, then further into monomers in the presence of SDS because they probably consisted of the B subunit with an atypical configuration, which is more vulnerable to disassembly. For the coiled coil to be effective in CTB assembly, a pentameric valency is essential because a trimeric or tetrameric coiled coil, such as that of chicken cartilage matrix protein or tetrabrachion, respectively, did not improve the pentamerization of CTB (data not shown). Second, the effect of the protein concentration on CTB pentamerization was evaluated because the rate of aggregation is

Please cite this article in press as: Tamaki Y, et al. Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination. Vaccine (2016), http://dx.doi.org/10.1016/j.vaccine.2016.01.034

237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260

G Model JVAC 17312 1–7

ARTICLE IN PRESS Y. Tamaki et al. / Vaccine xxx (2016) xxx–xxx

5

Fig. 5. Biochemical and immunological analyses of refolded mCTB. (a) GM1-ELISA of refolded mCTB, secreted pCTB, and choleragenoid. HMM aggregates-1, soluble highmolecular-mass aggregates generated with RP1; Pentamer-1 and Pentamer-2, mCTB pentamers regenerated with RP1 and RP2, respectively. HMM aggregates-1 and Pentamer1 were separated with size-exclusion chromatography after the mCTB protein was refolded with RP1. (b) GM1-binding competition assay. Cholera toxin (CT, 2 ␮g/mL) mixed with various amounts (1–80 ␮g/mL) of Pentamer-2, pCTB, or choleragenoid was detected with a GM1-ELISA using CTA-specific antiserum. Inhibition rate (%) was calculated as: (1–[OD415 of CT with competitor]/[OD415 of CT without competitor]) × 100. (c) Serum IgG titers of antisera obtained from mice intranasally (i.n.) immunized with Pentamer-2, pCTB, or choleragenoid, as described in the Materials and methods (Section 2.4). Titers were defined as the serum dilution that gave an OD415 of 0.2, or the serum dilution for which a one-point higher dilution (two-fold) gave an OD415 < 0.2. n.s., not significant (P > 0.05, Mann–Whitney U test). (d) GM1-binding inhibition assay. CT (0.2 ␮g/mL) mixed with antiserum (diluted 1:10 or 1:20) from mice immunized i.n. with Pentamer-2, pCTB, or choleragenoid was detected with a GM1-ELISA. Inhibition rate (%) was calculated as: (1 – [OD415 of antiserum]/[OD415 of unimmunized serum]) × 100.

261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287

known to increase with increasing concentrations of an unfolded polypeptide chain [17,18]. The initial protein concentration at dialysis was adjusted to 0.5 mg/mL, as described in Section 2.3. However, it was varied from 0.05 mg/mL to 5 mg/mL to test the concentration effect. As shown in Fig. 3, protein concentrations as low as 0.05 mg/mL had no positive effect on pentamer assembly, and such low protein concentrations were impractical because the scale of the dialysis becomes too large to handle easily. Third, the effect of protein purity on CTB pentamerization was evaluated because it is conceivable that contaminants interfere with protein folding, increasing aggregate formation [18], as previously demonstrated for monomeric proteins [19]. The contaminants of the mCTB protein present in IBs were removed with nickel-affinity chromatography after the IBs were solubilized with 6 M GdnHCl, a process referred to as RP2, as described in Section 2.3. A size-exclusion chromatographic analysis of all the protein concentrations, in a range of 0.05–5 mg/mL, showed a single peak of pentameric CTB, with no detectable soluble HMM aggregates or other multimeric or monomeric forms (Fig. 4a). The pentamer band observed on SDS-PAGE also appeared more homogeneous (Fig. 4b) than the pentamer bands evident in Figs. 2b and 3b. This result indicates that contaminants severely hampered the assembly of the pentamer, and that their removal robustly improved the pentamerization of CTB from IBs. RP2 also regenerated homogeneous mCTB–COMP fusion protein (compare Fig. 2a and Supplementary Fig. 2c), suggesting that the removal of contaminants before

Table 1 Final yields of refolded mCTB and mCTB–COMP proteins. Protein

mCTB mCTB–COMP

Refolding procedurea

RP1 RP2 RP1 RP2

Protein yield (mg/L of culture)b Unfractionated total protein

Pentamerc

125 – 135 –

32 40 30 45

a

Details of refolding procedures (RPs) RP1 and RP2 are described in Section 2. Final protein yields (in milligrams per liter of bacterial culture) were calculated based on a bicinchoninic acid protein assay after each protein was refolded and purified. c Amount of pentamer regenerated by RP1 was calculated after size-exclusion chromatographic purification of the pentamer. RP2 produced only pentamers of both mCTB and mCTB–COMP, with no soluble high-molecular-mass aggregates, so the total amount of protein was considered the total amount of pentamer. b

refolding is a valuable tool for refolding multisubunit proteins and their fusion proteins. As shown in Table 1, the recovery of the pentamer was 30–40 mg/L of bacterial culture for mCTB and the mCTB–COMP fusion protein. The optimization of fermentation, such as with high-density cultures, should increase the accumulation of the CTB protein in IBs, and consequently increase the final yield. RP2 did not require size-exclusion chromatography, which represents a significant advantage over RP1.

Please cite this article in press as: Tamaki Y, et al. Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination. Vaccine (2016), http://dx.doi.org/10.1016/j.vaccine.2016.01.034

288 289 290 291 292 293 294 295 296

G Model JVAC 17312 1–7

Y. Tamaki et al. / Vaccine xxx (2016) xxx–xxx

6 297

298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319

320

321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358

ARTICLE IN PRESS

3.3. Biochemical and immunological analyses of refolded CTB Except for the HMM aggregates generated with RP1, all the CTB constructs (Pentamer-1, Pentamer-2, pCTB, and choleragenoid) displayed similar binding curves in a GM1-ELISA (Fig. 5a), indicating that their receptor-binding affinities are similar. The GM1-binding competition assay suggested that the GM1-binding capacity of Pentamer-2 was slightly lower than those of pCTB and choleragenoid (Fig. 5b), particularly in its low concentration range. However, Pentamer-2 was as immunogenic as pCTB when administered mucosally to mice, indicating that these two forms are immunologically indistinguishable (Fig. 5c). Choleragenoid induced slightly higher levels of anti-CTB antibodies, probably because choleragenoid is contaminated with choleragen (i.e. CT), which is a strong mucosal antigen. Finally, the GM1-bindinginhibition capacities of the antisera induced with Pentamer-2 and pCTB were similar (Fig. 5d). Again, the choleragenoid-induced antisera exerted a stronger inhibitory effect than the recombinant CTBs because choleragenoid is more immunogenic. Taking these results together, we conclude that secreted pCTB and mCTB refolded with RP2 (Pentamer-2) were essentially indistinguishable in their biochemical and immunological properties, and functionally very similar to V. cholerae-derived choleragenoid.

4. Discussion The high-level production of CTB with the biologically active pentameric structure has been reported previously. For example, its secretion from V. cholerae approached 1 g/L of culture [8], and CTB with the ompA signal was secreted from E. coli at 1 g/L of culture [11]. In this study, we demonstrated that CTB, in which the leader sequence was replaced with the pelB signal, was secreted from E. coli solely as the pentamer, at 100 mg/L of culture. The secretion of CTB or its intracellular production as a soluble molecule has proved valuable. However, the production of the pentameric form was severely reduced in E. coli [3], V. cholerae [4], and yeast [12] when the protein was modified, including by the introduction of mutations or its conjugation with heterologous peptides at its N or C terminus. Therefore, the reassembly of the CTB pentamer from bacterial IBs through a process of protein denaturation–renaturation should be a valuable tool for the development of B-subunit-based toxoids or heterologous protein antigen carriers. The regeneration of pentamers from acid-denatured CTB monomers by neutralizing the protein has been documented previously [24,25], indicating that is possible to refold denatured CTB monomers to assemble pentamers. Although several studies have reported the regeneration of CTB pentamers from E. coli IBs with a protein denaturation–renaturation cycle [9,10], it must be noted that no soluble HMM CTB aggregates were detected on SDS-PAGE in this study (Figs. 2b, 3b, 4b) because they are vulnerable to disassembly to pentamers or monomers in the presence of SDS. Therefore, we stress that the CTB pentamer must be analyzed under nondenaturing conditions to evaluate the homogeneity of the protein (Figs 2a, 3a, 4a). We infer that the previously reported refolding of CTB generated considerable amounts of soluble HMM aggregates. In this study, we developed a simple but very efficient assembly method to regenerate highly structurally homogeneous pentameric CTB, the bioactivity of which is essentially indistinguishable from that of CTB secreted from E. coli or that of choleragenoid. Protein refolding in vitro, particularly of multimeric proteins, generally requires a high protein concentration to increase the frequency of productive physical collisions between

the monomeric molecules, which are necessary for their assembly. This is primarily why enterotoxins or their B subunits are more efficiently assembled in vitro at high protein concentrations [25], or within a concentrated subcellular milieu, such as the Gram-negative bacterial periplasm [26] or eukaryotic endoplasmic reticulum [12]. Therefore, in vitro protein refolding should mimic such subcellular microenvironmental conditions. However, a high protein concentration concomitantly presents a formidable disadvantage because high protein concentrations at the refolding stage potentially increase the frequency of partially folded or misfolded monomers, favoring the formation of soluble or even insoluble HMM aggregates. To our surprise, the dialysis of the protein at concentrations as high as 5 mg/mL had no apparent adverse effect on pentamer regeneration, although only in the absence of contaminants. This is particularly valuable because the experimental scale can be minimized without sacrificing the final yield. Moreover, further purification steps, such as those involving size-exclusion chromatography, are not required. Although contaminants are known to interfere with protein folding [18,19,27], those findings were based on single-subunit proteins, and to the best of our knowledge, there have been no previous reports of the robust effect of contaminant removal on the refolding of multisubunit proteins. It is conceivable that contaminants interfere with monomer folding per se or with the necessary intersubunit contacts by interacting with the proteins in various states of folding (unfolded, semifolded, or misfolded monomers, or oligomeric intermediates), thus preventing the completion of the folding and assembly processes to produce a stable tertiary structure. The intrasubunit disulfide bond (Cys9 –Cys86 ) is known to be required for CTB pentamer assembly in vivo and in vitro [24], presumably because it stabilizes a certain monomeric configuration that allows the molecule to proceed to the downstream assembly cascade. We found that this bond was intact in the mCTB monomer after the protein was refolded with RP1 (Supplementary Fig. 1d). To determine whether this disulfide is required for the regeneration of the pentamer, we disrupted the bond with a reducing agent before proceeding to RP2. We found that a mixture of pentameric molecules and HMM aggregates was generated (data not shown), as was seen when mCTB was refolded with RP1 (Fig. 1b). This indicates that the intrasubunit disulfide bond is required for the regeneration of homogeneously pentameric CTB in vitro, but is not an absolute requirement because the pentamer was partially regenerated and an intact disulfide bond was restored. Interestingly, the intrasubunit disulfide bond was also observed in the HMM aggregates of mCTB, which implies that soluble aggregates are generated in a relatively late stage of refolding. The in vitro pentameric assembly of enterotoxin B subunits, including CTB, is a valuable alternative to the expression of the secreted or intracellular soluble protein to meet the need to produce mutant or hybrid molecules carrying foreign epitopes for the development of enterotoxin-B-subunit-based toxoids or vaccine antigen carriers. Our study provides important insight into the creation of such antigens.

Acknowledgments This work was supported by a JSPS KAKENHI grant (grant num- Q3 ber 15K08428).

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.vaccine.2016.01. 034.

Please cite this article in press as: Tamaki Y, et al. Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination. Vaccine (2016), http://dx.doi.org/10.1016/j.vaccine.2016.01.034

359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411

412

413 414

415

416 417 418

G Model JVAC 17312 1–7

ARTICLE IN PRESS Y. Tamaki et al. / Vaccine xxx (2016) xxx–xxx

419

420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457

References [1] Jertborn M, Svennerholm A-M, Holmgren J. Safety and immunogenicity of an oral recombinant cholera B subunit—whole cell vaccine in Swedish volunteers. Vaccine 1992;10:130–2. [2] Jelinek T, Kollaritsch H. Vaccination with Dukoral® against travelers’ diarrhea (ETEC) and cholera. Expert Rev Vaccines 2008;7:561–7. [3] Miyata T, Oshiro S, Harakuni T, Taira T, Matsuzaki G, Arakawa T. Physicochemically stable cholera toxin B subunit pentamer created by peripheral molecular constraints imposed by de novo-introduced intersubunit disulfide crosslinks. Vaccine 2012;30:4225–32. [4] Sanchez J, Svennerholm A-M, Holmgren J. Genetic fusion of a non-toxic heatstable enterotoxin-related decapeptide antigen to cholera toxin B-subunit. FEBS Lett 1988;241:110–4. [5] Liljeqvist S, Ståhl S, Andréoni C, Binz H, Uhlén M, Murby M. Fusions to the cholera toxin B subunit: influence on pentamerization and GM1 binding. J Immun Methods 1997;210:125–35. [6] Arêas APM, Oliveira MLS, Miyaji EN, Leite LCC, Aires KA, Dias WO, et al. Expression and characterization of cholera toxin B—pneumococcal surface adhesin A fusion protein in Escherichia coli: ability of CTB-PsaA to induce humoral immune response in mice. Biochem Biophys Res Commun 2004;321:192–6. [7] Sanchez J, Holmgren J. Recombinant system for overexpression of cholera toxin B subunit in Vibrio cholerae as a basis for vaccine development. Proc Nat Acad Sci 1989;86:481–5. [8] Lebens M, Johansson S, Osek J, Lindblad M, Holmgren J. Large-scale production of Vibrio cholerae toxin B subunit for use in oral vaccines. Nature Biotechnol 1993;11:1574–8. [9] L’hoir C, Renard A, Martial JA. Expression in Escherichia coli of two mutated genes encoding the cholera toxin B subunit. Gene 1990;89:47–52. [10] de Mattos Arêas AP, de Oliveira MLS, Ramos CRR, Sbrogio-Almeida ME, Raw Ia, Ho PL. Synthesis of cholera toxin B subunit gene: cloning and expression of a functional 6XHis-tagged protein in Escherichia coli. Protein Expression Purification 2002;25:481–7. [11] Slos P, Speck D, Accart N, Kolbe HV, Schubnel D, Bouchon B, et al. Recombinant cholera toxin B-subunit in Escherichia coli: high-level secretion, purification, and characterization. Protein Expression Purification 1994;5:518–26. [12] Harakuni T, Sugawa H, Komesu A, Tadano M, Arakawa T. Heteropentameric cholera toxin B subunit chimeric molecules genetically fused to a vaccine antigen induce systemic and mucosal immune responses: a potential new strategy to target recombinant vaccine antigens to mucosal immune systems. Infection Immun 2005;73:5654–65.

7

[13] Jungbauer A, Kaar W. Current status of technical protein refolding. J Biotechnol 2007;128:587–96. [14] Li M, Su Z-G, Janson J-C. In vitro protein refolding by chromatographic procedures. Protein Expression Purification 2004;33:1–10. [15] Liu S, Tobias R, McClure S, Styba G, Shi Q, Jackowski G. Removal of Endotoxin from recombinant protein preparations. Clin Biochem 1997;30:455–63. [16] Arakawa T, Harakuni T. Cholera toxin B subunit—Five-stranded ␣-helical coiled-coil fusion protein: Five-to-five molecular chimera displays robust physicochemical stability. Vaccine 2014;32:5019–26. [17] Rudolph R, Lilie H. In vitro folding of inclusion body proteins. FASEB J 1996;10:49–56. [18] Lilie H, Schwarz E, Rudolph R. Advances in refolding of proteins produced in E. coli. Curr Opin Biotechnol 1998;9:497–501. [19] Maachupalli-Reddy J, Kelley BD, Clark EDB. Effect of inclusion body contaminants on the oxidative renaturation of hen egg white lysozyme. Biotechnol Progress 1997;13:144–50. [20] Miyata T, Harakuni T, Tsuboi T, Sattabongkot J, Ikehara A, Tachibana M, et al. Tricomponent immunopotentiating system as a novel molecular design strategy for malaria vaccine development. Infection Immun 2011;79:4260–75. [21] De Bernardez Clark E, Schwarz E, Rudolph R. Inhibition of aggregation side reactions during in vitro protein folding. Methods Enzymol 1999;309: 217–36. [22] Ho JGS, Middelberg APJ, Ramage P, Kocher HP. The likelihood of aggregation during protein renaturation can be assessed using the second virial coefficient. Protein Sci 2003;12:708–16. [23] Miyata T, Harakuni T, Taira T, Matsuzaki G, Arakawa T. Merozoite surface protein-1 of Plasmodium yoelii fused via an oligosaccharide moiety of cholera toxin B subunit glycoprotein expressed in yeast induced protective immunity against lethal malaria infection in mice. Vaccine 2012;30:948–58. [24] Hardy S, Holmgren J, Johansson S, Sanchez J, Hirst TR. Coordinated assembly of multisubunit proteins: oligomerization of bacterial enterotoxins in vivo and in vitro. Proc Nat Acad Sci 1988;85:7109–13. [25] Lesieur C, Cliff MJ, Carter R, James RF, Clarke AR, Hirst TR. A kinetic model of intermediate formation during assembly of cholera toxin B-subunit pentamers. J Biol Chem 2002;277:16697–704. [26] Hirst TR, Holmgren J. Conformation of protein secreted across bacterial outer membranes: a study of enterotoxin translocation from Vibrio cholerae. Proc Nat Acad Sci 1987;84:7418–22. [27] Tsumoto K, Ejima D, Kumagai I, Arakawa T. Practical considerations in refolding proteins from inclusion bodies. Protein Expression Purification 2003;28:1–8.

Please cite this article in press as: Tamaki Y, et al. Cholera toxin B subunit pentamer reassembled from Escherichia coli inclusion bodies for use in vaccination. Vaccine (2016), http://dx.doi.org/10.1016/j.vaccine.2016.01.034

458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497