- Email: [email protected]
High-Level Expression, Purification, and Characterization of Recombinant Type A Botulinum Neurotoxin Light Chain
Protein Expression and Purification 17, 339 –344 (1999) Article ID prep.1999.1138, available online at http://www.idealibrary.com on
High-Level Expression, Purification, and Characterization of Recombinant Type A Botulinum Neurotoxin Light Chain Li Li and Bal Ram Singh 1 Department of Chemistry and Biochemistry, University of Massachusetts Dartmouth, Dartmouth, Massachusetts 02747
Received January 6, 1999, and in revised form August 9, 1999
Botulinum neurotoxin light chain (BoNT LC, 50 kDa) is responsible for the zinc endopeptidase activity specific for proteins of neuroexocytosis apparatus. We describe the expression of recombinant type A BoNT LC in Escherichia coli as well as the purification and characterization of the recombinant protein. A high level of expression of BoNT/A LC was obtained by an extended postinduction time of 15 h at 30°C. Recombinant BoNT/A LC was isolated from an Ni 21 column. Due to its high pI (;8.7), purification was achieved by a single step of passing the protein through anionexchange chromatography at pH 8.0 without the need of elution. The purified recombinant BoNT/A LC retained proteolytic activity and had a secondary structure similar to that of native LC determined by CD measurement. © 1999 Academic Press Key Words: botulinum neurotoxin; Clostridium.
Botulinum neurotoxins (BoNTs) 2 are a group of extremely potent toxins, which are produced by various strains of Clostridium botulinum and in some cases by C. botirycum and C. barati (1, 2). Each of the seven serotypes of BoNTs (A–G) consists of a 50-kDa light chain (LC) and a ;100-kDa heavy chain (HC), which are linked through a disulfide bond. In its mode of action, the HC binds to presynaptic membrane at the nerve–muscle junction and helps translocate the LC 1 The Henry Dreyfus Teacher–Scholar and author to whom correspondence should be addressed at Department of Chemistry and Biochemistry, University of Massachusetts Dartmouth, 285 Old Westport Road, Dartmouth, MA 02747. Fax: 508-999-8451. E-mail: [email protected]. 2 Abbreviations used: BoNT/A, Clostridium botulinum type A neurotoxin; LC, light chain of BoNT; SNAP-25, synaptosomal-associated protein of 25 kDa; SNARE, soluble NSF attachment protein receptor, IPTG, isopropyl b-D-thiogalactopyranoside; Tris, tris(hydroxymethyl)aminomethane; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; IEF, isoelectrol focusing; CD, circular dichoism.
1046-5928/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
into the nerve cell. The LC acts as an endopeptidase against proteins involves in the exocytosis process, thus blocking the neurotransmitter release, which results in a flaccid muscle paralysis (3). The endopeptidase activity of BoNT LC has remained unique in more than one way (4). It contains a Zn 21 bound to the zinc-binding motif, HEXXH, which plays a both catalytic and structural role (5), in contrast to any other known Zn 21 protease. However, the most unique characteristic relates to its exclusive substrate specificity and extremely selective cleavage site. Each of the different BoNTs recognizes either a different protein substrate or a different cleavage site. For example, BoNT/A, /C, and /E cleave SNAP-25 but at exclusively different sites (6 – 8). BoNT/B, /D, /F, and /G cleave VAMP but at different sites (9), whereas BoNT/C exclusively cleaves syntaxin (10). Syntaxin, VAMP, and SNAP-25 are part of the SNARE complex involved in the docking and fusion of synaptic vesicles to plasma membrane at the neuromuscular junctions. Because such a widely varied group of substrate proteins are exclusively recognized by a structurally similar group of toxins, an extensive study of the structural features of the light chain is warranted to understand the molecular basis of such an unique substrate specificity. The availability of pure LC in high amounts is critical for such endeavors. Expression of a DNA fragment of the LC in a suitable Escherichia coli vector is a convenient way to obtain adequate amounts of the protein. Because of its extreme toxicity, BoNT is considered a major biological warfare threat (11). An understanding of its mechanism of action at the molecular level will allow design of antidotes against BoNT agents. In addition, BoNTs are currently being used as therapeutic reagents against several neuromuscular disorders, such as blepharospasm, strabismus, and torticollis (12). A clear and comprehensive understanding of its mechanism of action will provide the opportunity to explore the use of BoNTs against many more diseases. 339
340
LI AND SINGH
Expression of BoNT LC or HC in E. coli has remained a difficult problem because of the relatively high AT content of the C. botulinum genome (70%) (13). BoNT/A LC has been cloned and expressed in E. coli (14), but the yield has remained relatively low (;1 mg/L culture). In this article, we described a method of expressing BoNT/A LC in high yields (20 mg/L culture) and purifying it using a DEAE ion-exchange column. Such a high yield and purity (.99%) will help researchers to carry out experiments requiring large amounts of BoNT/A LC. EXPERIMENTAL PROCEDURES
Bacterial Strains and Plasmids The plasmids pBN3 and pBN10 were kindly provided by Dr. H. Niemann (Institut fur Tierzucht und Tierverhalten, Mariensee, Neustadt, Germany). The BoNT/A LC DNA fragment from C. botulinum neurotoxin type A was cloned with the histine tag at the C-terminal end and expressed under the control of the T5 promoter (15). The SNAP-25 expression clone was constructed similarly. The E. coli strain HB101 was purchased from Promega (Madison, WI). The culture was grown in 2YT medium (1.6% tryptone, 1% yeast extract, 0.5% NaCl) supplemented with 100 mg/ml ampicillin (Sigma, St. Louis, MO). Expression of Recombinant BoNT/A LC E. coli strain HB101 transformed with the T5 RNA polymerase expression plasmid pBN3 was used to inoculate a 50-ml aliquot of 2YT medium. The culture was grown under agitation overnight at 37°C and was stored as a glycerol stock in aliquots at 270°C until use. Frozen glycerol stock (1.5 ml) was used to inoculate 100 ml of 2YT medium at 37°C. When the absorbance at 600 nm was approximately 0.6, the 100-ml culture was added to 1 L medium in a 4-L flask, and the culture was grown at 30°C to OD 600 5 0.6 – 0.7. IPTG was then added to a final concentration of 0.5 mM, and the induction was allowed to proceed for 15 h. Purification E. coli cells were harvested by centrifugation at 4000g for 10 min at 4°C. All subsequent procedures were performed at 4°C. Cell pellet was resuspended in 50 ml lysis buffer (10 mM phosphate, 300 mM NaCl, 1 mM PMSF, and 5 mM benzamidine, pH 8.0) and frozen at 220°C until further use. The thawed cells were ruptured by pulse sonication for 2 3 1.5 min on ice and then centrifuged at 4000g for 10 min to pellet unbroken cells and other insoluble material. The supernatant was collected as a lysate. To ensure complete disruption of the soluble proteins, the pellet was resuspended in 20 ml lysis buffer, sonicated (2 3 1.5 min), and
centrifuged and the supernatant was pooled along with the lysate obtained in the previous step. The lysate was centrifuged at 50,000g for 1 h, and the resulting supernatant was loaded onto a His-Bind (Novagen, Inc., Madison, WI) column (2 ml bed volume), which was previously equilibrated with equilibration buffer (lysis buffer without protease inhibitors). The column was first washed with 40 ml equilibration buffer and then with 15 ml wash buffer (20 mM immidazole in lysis buffer). LC was eluted with 30 ml elution buffer (100 mM immidazole in lysis buffer). The last 15-ml wash and the 30-ml eluate were collected as pools A and B, respectively. LC pools were dialyzed overnight against 10 mM phosphate buffer containing 50 mM NaCl, pH 8.0. Pools B and A were sequentially loaded onto a DEAE– A50 column (1.5 3 5 cm) equilibrated with the same buffer. The flow through fractions were analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and pooled based on the contaminant present. SDS–PAGE Sodium dodecyl sulfate–polyacrylamide gel electrophoresis was carried out to estimate protein expression under different conditions and to check the purity of the isolated LC. The electrophoresis was carried out on a Mini Protein II System (Bio-Rad, Hercules, CA) using a 7.5% polyacrylamide gel, as described previously (16). To monitor the induction of LC, 1.5 ml of culture was sampled at various times after induction. The cells were centrifuged (4000g, 10 min) and resuspended in 300 ml 23 reducing SDS sample buffer (0.125 M Tris, 4% SDS, 20% glycerol, 2% 2-mercaptoethanol, pH 6.8). After heating at 100°C for 3 min and centrifugation at 5000g for 15 min, 5 ml of the sample (corresponding to 25 ml of culture) was subjected to SDS–PAGE, and the proteins were detected by Coomassie blue staining. Purified LC (2 mg/ml) and molecular markers were separately mixed with 23 SDS sample buffer and heated at 100°C for 3 min. These samples (10 ml) were applied to the SDS–PAGE gels. After electrophoresis, protein bands were detected by Coomassie blue staining. To estimate the percentage of total cellular protein represented by the recombinant BoNT/A LC, Coomassie-stained 7.5% SDS–PAGE gels resolving the total E. coli lysate proteins were scanned on an itti imager, plotted, and integrated for density using an itti Imager (itti, I.c., Petersburg, FL). The LC band was integrated for density in each lane and compared with the total integrated density of all proteins in the same lane to obtain the percentage of LC of total E. coli protein.
HIGH-LEVEL EXPRESSION OF BOTULINUM NEUROTOXIN LIGHT CHAIN
341
FIG. 1. SDS–PAGE analysis of total cell lysate of E. coli HB101 containing pBN3 showing the expression of LC at different induction times (lanes 2–9). Lysate samples were prepared from 1.5 ml of cell suspension induced with IPTG (see Experimental Procedures), microcentrifuged, resuspended in 300 ml 23 SDS gel loading buffer, heated to 100°C for 4 min, and loaded (5 ml) onto 12% polyacrylamide gels. Supernatants and pellets of the cell crude extract were also analyzed (lanes 10 and 11) to assess the solubility of the expression product. After induction at 30°C for 15 h, cells (1.5 L) were harvested by centrifugation and disrupted by sonication. The suspension was centrifuged, and the resulting supernatant and pellet were analyzed by SDS gel.
Isoelectric Focusing Isoelectric focusing (IEF) was carried out using the Mini Protein II system from Bio-Rad (Hercules, CA). Briefly, an aliquot of LC (2 mg/ml) was mixed with the same volume of sample buffer (6% Ampholytes, pH 3.5–10). The electrode buffers used were 0.02 M acetic acid anolyte and 0.02 M NaOH catholyte. After prerunning the gel (5.5% Ampholytes, pH 3.5–10, 5% acrylamide, and 10% glycerol) at 300 V for 30 min, 40 ml of mixture was applied to the gel. Eletrophoresis was TABLE 1 Purification Scheme of Recombinant BoNT/A LC
Step
Total LC (mg/L culture)
Total protein (mg/L culture) a
Purification factor b
Yield (%)
Extraction Ni 21 column DEAE-A50
25 21 20
622 22 20
1 28 31
100 84 80
a b
Estimated from band intensity on SDS–PAGE. Calculated based on total protein content.
performed at 300 V for 5 h at 8°C. The gel was fixed in 20% trichloroacetic acid and then in destaining solution (40% ethanol, 10% acetic acid) for 10 min. The gel was then stained with 0.1% Coomassie Blue R dissolved in 40% methanol and 10% acetic acid and destained until the background was clear. Endopeptidase Activity E. coli HB101 cells containing pBN10 were grown at 37°C to an A 590 of about 0.6. Expression of SNAP-25 was induced by addition of IPTG to a final concentration of 0.5 mM at 37°C for 3 h. Recombinant His 6tagged SNAP-25 was isolated using a His-Bind column. Assays for endopeptidase activity of the recombinant LC and native toxin were carried out in a 50-ml reaction mixture containing 8 mM SNAP-25, 20 nM LC, or reduced BoNT/A. After incubation in a reaction buffer (20 mM Tris buffer, pH 7.0, containing 50 mM NaCl) at 37°C, samples were withdrawn at various times and reactions were stopped by adding SDS– PAGE sample buffer. The products were resolved by electrophoresis on a 10% polyacrylamide gel.
342
LI AND SINGH
FIG. 2. SDS–PAGE analysis of sample obtained during the purification of LC from His-binding column by DEAE-A50 chomatography. The number on the left indicates molecular mass of markers in kDa.
loss, resulting in a final yield of 0.5 mg/L. In this article, we report the high-level expression of His 6tagged LC in E. coli and a two-step purification to obtain homogeneous preparations of BoNT/A light chain. E. coli cells containing pBN3 did not appear to grow as well as cells containing the control vector and typically required ;4 h to reach an A 600 of 0.6 at 37°C, whereas control cells reached this stage of growth within 2.5 h. The pBN3 cells produced smaller colonies on culture plates and lower turbidity in overnight cultures than the control cells. A low level of expression (;1 mg/liter culture) was obtained at 3 h of induction. The reasons for this observation are not clear to us at this time. It might be possible that the LC caused cellular proliferative and growth impairment, and LC product expressed even at the basal level before induction could affect the growth of the cells. We have not as yet pursued these possibilities any further since our major concern here is the high-level expression of the LC. An attempt at prolonged postinduction at lowered temperature showed surprising results. We used SDS– polyacrylamide gel eletrophoresis analysis to follow the appearance of LC production following induction with IPTG (Fig. 1). A maximum level was reached after 15 h of incubation, at which time the LC accumulated at 24 –28 mg/liter, accounting for 3.9% of the total protein, as determined by SDS–PAGE separation followed by densitometric scanning using purified LC as a stan-
Circular Dichoism Spectroscopy The recombinant LC (0.3 mg/ml) was exchanged into 10 mM sodium phosphate buffer, pH 7.4, by dialysis. The CD spectrum was recorded between 178 and 260 nm at 10°C in a 1-mm path cuvette using a Jasco J710 spectropolarimeter. The scanning speed was set at 20 nm/min and the response time was 4 s. Buffer contribution was corrected. The secondary structure calculation was carried out by SELCON using the Softsec program (Softwood Co.). On the basis of the amino acid sequence, the mean residue weight was calculated to be 113.81. RESULTS AND DISCUSSION
Purification of native LC from whole toxin is difficult because it involves its efficient separation from HC. In addition, working with large quantities of botulinum neurotoxin poses a health risk. To avoid these problems and to more economically obtain high yields of pure LC, alternative strategies are required. Low expression yield has remained a technical difficulty associated with the expression of LC in E. coli (17). Although BoNT/A LC fused with MBP has been successfully expressed in E. coli up to 5–10 mg/L (18), additional separation of MBP from LC causes a major
FIG. 3. Isoelectric focusing analysis of purified recombinant LC. Lane 1, isoelectric point marker as denoted in the left column. Lane 2, isolated recombinant LC.
HIGH-LEVEL EXPRESSION OF BOTULINUM NEUROTOXIN LIGHT CHAIN
343
FIG. 4. Cleavage of SNAP-25 by BoNT/A (■) and recombinant LC (Œ). SNAP-25 (12 mM) was incubated with BoNT/A (20 nM) or LC (20 nM) at 37°C. Samples were withdrawn at various times and reaction was terminated by the addition of SDS sample buffer. The cleaved SNAP-25 was analyzed by gel electrophoresis. The inset shows SDS–PAGE of SNAP-25 cleavage for 0 (lane 1), 5 (lane 2), 10 (lane 3), 20 (lane 4), and 30 (lane 5) min.
dard. This expression level was considerably higher (2-fold) than the maximum expression at 37°C that was achieved at the induction time of 11 h (data not shown). A typical purification from a 1-L culture yielded 18 –22 mg, which was ;20-fold enhancement over the 3-h induction at 37°C method (Table 1). To assess the solubility of the pBN3 product, the distribution of LC among supernatants and pellets of cell sonicate lysate was examined on SDS–PAGE (Fig. 1). We observed that the expressed LC was associated with both soluble and insoluble fractions. LC in soluble fraction was estimated to be ;20 mg/liter culture and constituted ;80% of total expressed LC product. We attempted to solubilize the inclusion bodies from the cell pellet in 6 M guanidine hydrochloride to isolate the aggregated LC by Ni 21 column. However, this effort was not successful because a large amount of precipitation occurred in the process of LC refolding by dialysis removal of GnHCl. We used a 2-ml bed size Ni 21 column to isolate recombinant LC from 1 L cell culture. The production yield varied in experiments. When the yield was low, some minor contamination was noticeable in the eluted LC fraction (Fig. 2, lane 2) because these proteins adhered to excess charged binding sites on the Ni 21 resin. On the other hand, a substantial amount LC was observed in 20 mM immidazole wash fractions (Fig. 2, lane 4) when the LC was more abundant in the lysate. In any case, we need to remove minor impurities in LC preparations obtained though the Ni 21 column. Isoelectric focusing analysis of the recombinant LC revealed a
high pI of 8.7 (Fig. 3), which allowed us to use anionexchange chomatography at pH 8.0 so that the contaminants but not LC bound to the column. Purification of the LC eluate (pool B) and wash (pool A) was achieved by the single step of passing it though a DEAE-A50 column at pH 8.0. No contamination was visible in purified pool B even with heavy loading of LC sample to the gels (Fig. 2, lanes 3), and the extent of purity of the pool A was over 99% (Fig. 2, lanes 5). Recovery of LC from DEAE-A50 purification was close to 100%. The LC prepared with the procedure described above
FIG. 5. Circular dichoism spectrum of recombinant LC. The CD spectrum was recorded as described under Experimental Procedures. [u] is the mean residue weight ellipticity.
344
LI AND SINGH
was able to readily cleave SNAP-25, in a way similar to that for the native reduced BoNT/A (Fig. 4), suggesting that it retained a functional structure. The structural integrity of recombinant LC was further conformed by CD spectral analysis. The plot of mean residue molar ellipticity of recombinant LC exhibits two minima at 209 and 220 nm (Fig. 5). Curve fitting analysis of the CD spectrum revealed that recombinant LC consists of 27% a-helix, 17% b-sheet, 20% turns, and 36% random coil, which is similar to the secondary structure content of the native LC (19). Thus, the recombinant LC is highly structured and functionally active, similar to the native LC. In summary, we have developed a procedure to express a large quantity of BoNT/A light chain and a two-step method to purify it in a functionally active form. ACKNOWLEDGMENT This work was supported by Grant NS33740 from the National Institute of Neurological Disorders and Stroke, NIH.
REFERENCES 1. Hall, J. D., McCroskey, L. M., Pincomb, B. J., and Hatheway, C. L. (1985) J. Clin. Microbiol. 21, 654 – 655. 2. Aureli, P., Fenicia, L., Pasolini, B., Gianfranceschi, M., McCroskey, L. M., and Hatheway, C. L. (1986) J. Infect. Dis. 154, 207–211. 3. Ahnert-Hilger, G., Mach, W., Fo¨hr, K. J., and Gratzl, M. (1989) Methods Cell Biol. 31, 63–90. 4. Li, L., and Singh, B. R. (1999) J. Nat. Toxin-Toxin Rev. 18, 95–112.
5. Fu, F. N., Lomneth, R. B., Cai, S., and Singh, B. R. (1998) Biochemistry 37, 5267–5278. 6. Blasi, J., Chapman, E. R., Link, E., Binz, T., Yamasaki, S., De Camilli, P., Su¨dhof, T. C., Niemann, H., and Jahn, R. (1993) Nature 365, 160 –163. 7. Schiavo, G., Santucci, A., DasGupta, B. R., Mehta, P. P., Jontes, J., Benfenati, F., Wilson, M. C., and Montecucco, C. (1993) FEBS Lett. 335, 99 –103. 8. Binz, T., Blasi, J., Yamasaki, S., Baumeister, A., Link, E., Su¨dhof, T. C., Jahn, R., and Niemann, H. (1994) J. Biol. Chem. 269, 1617–1620. 9. Niemann, H., Blasi, J., and Jahn, R. (1994) Trends Cell Biol. 4, 179 –185. 10. Blasi, J., Chapman, E. R., Yamasaki, S., Binz, T., Niemann, H., and Jahn, R. (1993) EMBO J. 12, 4821– 4828. 11. Atlas, R. M. (1998) Crit. Rev. Microbiol. 24, 157–168. 12. Wheeler, A. H. (1997) Am. Fam. Physician 55, 541–545, 548. 13. Thompson, D. E., Brehm, J. K., Oultram, J. D., Swinfield, T. J., Shone, C. C., Atkinson, T., Melling, J., and Minton, N. P. (1990) Eur. J. Biochem. 189, 73– 81. 14. Li, Y., Foran, P., Fairweather, N. F., de Paiva, A., Weller, U., Dougan, G., and Dolly, J. O. (1994) Biochemistry 33, 7014 –7020. 15. Vaidyanathan, V. V., Yoshino, K., Jahnz, M., Dorries, C., Bade, S., Nauenburg, S., Niemann, H., and Binz, T. (1999) J. Neurochem. 72, 327–337. 16. Fu, F. N., Sharma, S. K., and Singh, B. R. (1998). J. Protein. Chem. 17, 53– 60. 17. Clayton, M. A., Clayton, J. M. Brown, D. R., and Middlebrook, J. L. (1996) Infect. Immun. 63, 2738 –2742. 18. Zhou, L., de Paiva, A., Liu, D., Aoki, R., and Dolly, J. O. (1995) Biochemistry 34, 15175–15181. 19. Singh, B. R., and DasGupta, B. R. (1989) Mol. Cell. Biochem. 85, 67–73.
High-Level Expression, Purification, and Characterization of Recombinant Type A Botulinum Neurotoxin Light Chain Li Li and Bal Ram Singh 1 Department of Chemistry and Biochemistry, University of Massachusetts Dartmouth, Dartmouth, Massachusetts 02747
Received January 6, 1999, and in revised form August 9, 1999
Botulinum neurotoxin light chain (BoNT LC, 50 kDa) is responsible for the zinc endopeptidase activity specific for proteins of neuroexocytosis apparatus. We describe the expression of recombinant type A BoNT LC in Escherichia coli as well as the purification and characterization of the recombinant protein. A high level of expression of BoNT/A LC was obtained by an extended postinduction time of 15 h at 30°C. Recombinant BoNT/A LC was isolated from an Ni 21 column. Due to its high pI (;8.7), purification was achieved by a single step of passing the protein through anionexchange chromatography at pH 8.0 without the need of elution. The purified recombinant BoNT/A LC retained proteolytic activity and had a secondary structure similar to that of native LC determined by CD measurement. © 1999 Academic Press Key Words: botulinum neurotoxin; Clostridium.
Botulinum neurotoxins (BoNTs) 2 are a group of extremely potent toxins, which are produced by various strains of Clostridium botulinum and in some cases by C. botirycum and C. barati (1, 2). Each of the seven serotypes of BoNTs (A–G) consists of a 50-kDa light chain (LC) and a ;100-kDa heavy chain (HC), which are linked through a disulfide bond. In its mode of action, the HC binds to presynaptic membrane at the nerve–muscle junction and helps translocate the LC 1 The Henry Dreyfus Teacher–Scholar and author to whom correspondence should be addressed at Department of Chemistry and Biochemistry, University of Massachusetts Dartmouth, 285 Old Westport Road, Dartmouth, MA 02747. Fax: 508-999-8451. E-mail: [email protected]. 2 Abbreviations used: BoNT/A, Clostridium botulinum type A neurotoxin; LC, light chain of BoNT; SNAP-25, synaptosomal-associated protein of 25 kDa; SNARE, soluble NSF attachment protein receptor, IPTG, isopropyl b-D-thiogalactopyranoside; Tris, tris(hydroxymethyl)aminomethane; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; IEF, isoelectrol focusing; CD, circular dichoism.
1046-5928/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
into the nerve cell. The LC acts as an endopeptidase against proteins involves in the exocytosis process, thus blocking the neurotransmitter release, which results in a flaccid muscle paralysis (3). The endopeptidase activity of BoNT LC has remained unique in more than one way (4). It contains a Zn 21 bound to the zinc-binding motif, HEXXH, which plays a both catalytic and structural role (5), in contrast to any other known Zn 21 protease. However, the most unique characteristic relates to its exclusive substrate specificity and extremely selective cleavage site. Each of the different BoNTs recognizes either a different protein substrate or a different cleavage site. For example, BoNT/A, /C, and /E cleave SNAP-25 but at exclusively different sites (6 – 8). BoNT/B, /D, /F, and /G cleave VAMP but at different sites (9), whereas BoNT/C exclusively cleaves syntaxin (10). Syntaxin, VAMP, and SNAP-25 are part of the SNARE complex involved in the docking and fusion of synaptic vesicles to plasma membrane at the neuromuscular junctions. Because such a widely varied group of substrate proteins are exclusively recognized by a structurally similar group of toxins, an extensive study of the structural features of the light chain is warranted to understand the molecular basis of such an unique substrate specificity. The availability of pure LC in high amounts is critical for such endeavors. Expression of a DNA fragment of the LC in a suitable Escherichia coli vector is a convenient way to obtain adequate amounts of the protein. Because of its extreme toxicity, BoNT is considered a major biological warfare threat (11). An understanding of its mechanism of action at the molecular level will allow design of antidotes against BoNT agents. In addition, BoNTs are currently being used as therapeutic reagents against several neuromuscular disorders, such as blepharospasm, strabismus, and torticollis (12). A clear and comprehensive understanding of its mechanism of action will provide the opportunity to explore the use of BoNTs against many more diseases. 339
340
LI AND SINGH
Expression of BoNT LC or HC in E. coli has remained a difficult problem because of the relatively high AT content of the C. botulinum genome (70%) (13). BoNT/A LC has been cloned and expressed in E. coli (14), but the yield has remained relatively low (;1 mg/L culture). In this article, we described a method of expressing BoNT/A LC in high yields (20 mg/L culture) and purifying it using a DEAE ion-exchange column. Such a high yield and purity (.99%) will help researchers to carry out experiments requiring large amounts of BoNT/A LC. EXPERIMENTAL PROCEDURES
Bacterial Strains and Plasmids The plasmids pBN3 and pBN10 were kindly provided by Dr. H. Niemann (Institut fur Tierzucht und Tierverhalten, Mariensee, Neustadt, Germany). The BoNT/A LC DNA fragment from C. botulinum neurotoxin type A was cloned with the histine tag at the C-terminal end and expressed under the control of the T5 promoter (15). The SNAP-25 expression clone was constructed similarly. The E. coli strain HB101 was purchased from Promega (Madison, WI). The culture was grown in 2YT medium (1.6% tryptone, 1% yeast extract, 0.5% NaCl) supplemented with 100 mg/ml ampicillin (Sigma, St. Louis, MO). Expression of Recombinant BoNT/A LC E. coli strain HB101 transformed with the T5 RNA polymerase expression plasmid pBN3 was used to inoculate a 50-ml aliquot of 2YT medium. The culture was grown under agitation overnight at 37°C and was stored as a glycerol stock in aliquots at 270°C until use. Frozen glycerol stock (1.5 ml) was used to inoculate 100 ml of 2YT medium at 37°C. When the absorbance at 600 nm was approximately 0.6, the 100-ml culture was added to 1 L medium in a 4-L flask, and the culture was grown at 30°C to OD 600 5 0.6 – 0.7. IPTG was then added to a final concentration of 0.5 mM, and the induction was allowed to proceed for 15 h. Purification E. coli cells were harvested by centrifugation at 4000g for 10 min at 4°C. All subsequent procedures were performed at 4°C. Cell pellet was resuspended in 50 ml lysis buffer (10 mM phosphate, 300 mM NaCl, 1 mM PMSF, and 5 mM benzamidine, pH 8.0) and frozen at 220°C until further use. The thawed cells were ruptured by pulse sonication for 2 3 1.5 min on ice and then centrifuged at 4000g for 10 min to pellet unbroken cells and other insoluble material. The supernatant was collected as a lysate. To ensure complete disruption of the soluble proteins, the pellet was resuspended in 20 ml lysis buffer, sonicated (2 3 1.5 min), and
centrifuged and the supernatant was pooled along with the lysate obtained in the previous step. The lysate was centrifuged at 50,000g for 1 h, and the resulting supernatant was loaded onto a His-Bind (Novagen, Inc., Madison, WI) column (2 ml bed volume), which was previously equilibrated with equilibration buffer (lysis buffer without protease inhibitors). The column was first washed with 40 ml equilibration buffer and then with 15 ml wash buffer (20 mM immidazole in lysis buffer). LC was eluted with 30 ml elution buffer (100 mM immidazole in lysis buffer). The last 15-ml wash and the 30-ml eluate were collected as pools A and B, respectively. LC pools were dialyzed overnight against 10 mM phosphate buffer containing 50 mM NaCl, pH 8.0. Pools B and A were sequentially loaded onto a DEAE– A50 column (1.5 3 5 cm) equilibrated with the same buffer. The flow through fractions were analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and pooled based on the contaminant present. SDS–PAGE Sodium dodecyl sulfate–polyacrylamide gel electrophoresis was carried out to estimate protein expression under different conditions and to check the purity of the isolated LC. The electrophoresis was carried out on a Mini Protein II System (Bio-Rad, Hercules, CA) using a 7.5% polyacrylamide gel, as described previously (16). To monitor the induction of LC, 1.5 ml of culture was sampled at various times after induction. The cells were centrifuged (4000g, 10 min) and resuspended in 300 ml 23 reducing SDS sample buffer (0.125 M Tris, 4% SDS, 20% glycerol, 2% 2-mercaptoethanol, pH 6.8). After heating at 100°C for 3 min and centrifugation at 5000g for 15 min, 5 ml of the sample (corresponding to 25 ml of culture) was subjected to SDS–PAGE, and the proteins were detected by Coomassie blue staining. Purified LC (2 mg/ml) and molecular markers were separately mixed with 23 SDS sample buffer and heated at 100°C for 3 min. These samples (10 ml) were applied to the SDS–PAGE gels. After electrophoresis, protein bands were detected by Coomassie blue staining. To estimate the percentage of total cellular protein represented by the recombinant BoNT/A LC, Coomassie-stained 7.5% SDS–PAGE gels resolving the total E. coli lysate proteins were scanned on an itti imager, plotted, and integrated for density using an itti Imager (itti, I.c., Petersburg, FL). The LC band was integrated for density in each lane and compared with the total integrated density of all proteins in the same lane to obtain the percentage of LC of total E. coli protein.
HIGH-LEVEL EXPRESSION OF BOTULINUM NEUROTOXIN LIGHT CHAIN
341
FIG. 1. SDS–PAGE analysis of total cell lysate of E. coli HB101 containing pBN3 showing the expression of LC at different induction times (lanes 2–9). Lysate samples were prepared from 1.5 ml of cell suspension induced with IPTG (see Experimental Procedures), microcentrifuged, resuspended in 300 ml 23 SDS gel loading buffer, heated to 100°C for 4 min, and loaded (5 ml) onto 12% polyacrylamide gels. Supernatants and pellets of the cell crude extract were also analyzed (lanes 10 and 11) to assess the solubility of the expression product. After induction at 30°C for 15 h, cells (1.5 L) were harvested by centrifugation and disrupted by sonication. The suspension was centrifuged, and the resulting supernatant and pellet were analyzed by SDS gel.
Isoelectric Focusing Isoelectric focusing (IEF) was carried out using the Mini Protein II system from Bio-Rad (Hercules, CA). Briefly, an aliquot of LC (2 mg/ml) was mixed with the same volume of sample buffer (6% Ampholytes, pH 3.5–10). The electrode buffers used were 0.02 M acetic acid anolyte and 0.02 M NaOH catholyte. After prerunning the gel (5.5% Ampholytes, pH 3.5–10, 5% acrylamide, and 10% glycerol) at 300 V for 30 min, 40 ml of mixture was applied to the gel. Eletrophoresis was TABLE 1 Purification Scheme of Recombinant BoNT/A LC
Step
Total LC (mg/L culture)
Total protein (mg/L culture) a
Purification factor b
Yield (%)
Extraction Ni 21 column DEAE-A50
25 21 20
622 22 20
1 28 31
100 84 80
a b
Estimated from band intensity on SDS–PAGE. Calculated based on total protein content.
performed at 300 V for 5 h at 8°C. The gel was fixed in 20% trichloroacetic acid and then in destaining solution (40% ethanol, 10% acetic acid) for 10 min. The gel was then stained with 0.1% Coomassie Blue R dissolved in 40% methanol and 10% acetic acid and destained until the background was clear. Endopeptidase Activity E. coli HB101 cells containing pBN10 were grown at 37°C to an A 590 of about 0.6. Expression of SNAP-25 was induced by addition of IPTG to a final concentration of 0.5 mM at 37°C for 3 h. Recombinant His 6tagged SNAP-25 was isolated using a His-Bind column. Assays for endopeptidase activity of the recombinant LC and native toxin were carried out in a 50-ml reaction mixture containing 8 mM SNAP-25, 20 nM LC, or reduced BoNT/A. After incubation in a reaction buffer (20 mM Tris buffer, pH 7.0, containing 50 mM NaCl) at 37°C, samples were withdrawn at various times and reactions were stopped by adding SDS– PAGE sample buffer. The products were resolved by electrophoresis on a 10% polyacrylamide gel.
342
LI AND SINGH
FIG. 2. SDS–PAGE analysis of sample obtained during the purification of LC from His-binding column by DEAE-A50 chomatography. The number on the left indicates molecular mass of markers in kDa.
loss, resulting in a final yield of 0.5 mg/L. In this article, we report the high-level expression of His 6tagged LC in E. coli and a two-step purification to obtain homogeneous preparations of BoNT/A light chain. E. coli cells containing pBN3 did not appear to grow as well as cells containing the control vector and typically required ;4 h to reach an A 600 of 0.6 at 37°C, whereas control cells reached this stage of growth within 2.5 h. The pBN3 cells produced smaller colonies on culture plates and lower turbidity in overnight cultures than the control cells. A low level of expression (;1 mg/liter culture) was obtained at 3 h of induction. The reasons for this observation are not clear to us at this time. It might be possible that the LC caused cellular proliferative and growth impairment, and LC product expressed even at the basal level before induction could affect the growth of the cells. We have not as yet pursued these possibilities any further since our major concern here is the high-level expression of the LC. An attempt at prolonged postinduction at lowered temperature showed surprising results. We used SDS– polyacrylamide gel eletrophoresis analysis to follow the appearance of LC production following induction with IPTG (Fig. 1). A maximum level was reached after 15 h of incubation, at which time the LC accumulated at 24 –28 mg/liter, accounting for 3.9% of the total protein, as determined by SDS–PAGE separation followed by densitometric scanning using purified LC as a stan-
Circular Dichoism Spectroscopy The recombinant LC (0.3 mg/ml) was exchanged into 10 mM sodium phosphate buffer, pH 7.4, by dialysis. The CD spectrum was recorded between 178 and 260 nm at 10°C in a 1-mm path cuvette using a Jasco J710 spectropolarimeter. The scanning speed was set at 20 nm/min and the response time was 4 s. Buffer contribution was corrected. The secondary structure calculation was carried out by SELCON using the Softsec program (Softwood Co.). On the basis of the amino acid sequence, the mean residue weight was calculated to be 113.81. RESULTS AND DISCUSSION
Purification of native LC from whole toxin is difficult because it involves its efficient separation from HC. In addition, working with large quantities of botulinum neurotoxin poses a health risk. To avoid these problems and to more economically obtain high yields of pure LC, alternative strategies are required. Low expression yield has remained a technical difficulty associated with the expression of LC in E. coli (17). Although BoNT/A LC fused with MBP has been successfully expressed in E. coli up to 5–10 mg/L (18), additional separation of MBP from LC causes a major
FIG. 3. Isoelectric focusing analysis of purified recombinant LC. Lane 1, isoelectric point marker as denoted in the left column. Lane 2, isolated recombinant LC.
HIGH-LEVEL EXPRESSION OF BOTULINUM NEUROTOXIN LIGHT CHAIN
343
FIG. 4. Cleavage of SNAP-25 by BoNT/A (■) and recombinant LC (Œ). SNAP-25 (12 mM) was incubated with BoNT/A (20 nM) or LC (20 nM) at 37°C. Samples were withdrawn at various times and reaction was terminated by the addition of SDS sample buffer. The cleaved SNAP-25 was analyzed by gel electrophoresis. The inset shows SDS–PAGE of SNAP-25 cleavage for 0 (lane 1), 5 (lane 2), 10 (lane 3), 20 (lane 4), and 30 (lane 5) min.
dard. This expression level was considerably higher (2-fold) than the maximum expression at 37°C that was achieved at the induction time of 11 h (data not shown). A typical purification from a 1-L culture yielded 18 –22 mg, which was ;20-fold enhancement over the 3-h induction at 37°C method (Table 1). To assess the solubility of the pBN3 product, the distribution of LC among supernatants and pellets of cell sonicate lysate was examined on SDS–PAGE (Fig. 1). We observed that the expressed LC was associated with both soluble and insoluble fractions. LC in soluble fraction was estimated to be ;20 mg/liter culture and constituted ;80% of total expressed LC product. We attempted to solubilize the inclusion bodies from the cell pellet in 6 M guanidine hydrochloride to isolate the aggregated LC by Ni 21 column. However, this effort was not successful because a large amount of precipitation occurred in the process of LC refolding by dialysis removal of GnHCl. We used a 2-ml bed size Ni 21 column to isolate recombinant LC from 1 L cell culture. The production yield varied in experiments. When the yield was low, some minor contamination was noticeable in the eluted LC fraction (Fig. 2, lane 2) because these proteins adhered to excess charged binding sites on the Ni 21 resin. On the other hand, a substantial amount LC was observed in 20 mM immidazole wash fractions (Fig. 2, lane 4) when the LC was more abundant in the lysate. In any case, we need to remove minor impurities in LC preparations obtained though the Ni 21 column. Isoelectric focusing analysis of the recombinant LC revealed a
high pI of 8.7 (Fig. 3), which allowed us to use anionexchange chomatography at pH 8.0 so that the contaminants but not LC bound to the column. Purification of the LC eluate (pool B) and wash (pool A) was achieved by the single step of passing it though a DEAE-A50 column at pH 8.0. No contamination was visible in purified pool B even with heavy loading of LC sample to the gels (Fig. 2, lanes 3), and the extent of purity of the pool A was over 99% (Fig. 2, lanes 5). Recovery of LC from DEAE-A50 purification was close to 100%. The LC prepared with the procedure described above
FIG. 5. Circular dichoism spectrum of recombinant LC. The CD spectrum was recorded as described under Experimental Procedures. [u] is the mean residue weight ellipticity.
344
LI AND SINGH
was able to readily cleave SNAP-25, in a way similar to that for the native reduced BoNT/A (Fig. 4), suggesting that it retained a functional structure. The structural integrity of recombinant LC was further conformed by CD spectral analysis. The plot of mean residue molar ellipticity of recombinant LC exhibits two minima at 209 and 220 nm (Fig. 5). Curve fitting analysis of the CD spectrum revealed that recombinant LC consists of 27% a-helix, 17% b-sheet, 20% turns, and 36% random coil, which is similar to the secondary structure content of the native LC (19). Thus, the recombinant LC is highly structured and functionally active, similar to the native LC. In summary, we have developed a procedure to express a large quantity of BoNT/A light chain and a two-step method to purify it in a functionally active form. ACKNOWLEDGMENT This work was supported by Grant NS33740 from the National Institute of Neurological Disorders and Stroke, NIH.
REFERENCES 1. Hall, J. D., McCroskey, L. M., Pincomb, B. J., and Hatheway, C. L. (1985) J. Clin. Microbiol. 21, 654 – 655. 2. Aureli, P., Fenicia, L., Pasolini, B., Gianfranceschi, M., McCroskey, L. M., and Hatheway, C. L. (1986) J. Infect. Dis. 154, 207–211. 3. Ahnert-Hilger, G., Mach, W., Fo¨hr, K. J., and Gratzl, M. (1989) Methods Cell Biol. 31, 63–90. 4. Li, L., and Singh, B. R. (1999) J. Nat. Toxin-Toxin Rev. 18, 95–112.
5. Fu, F. N., Lomneth, R. B., Cai, S., and Singh, B. R. (1998) Biochemistry 37, 5267–5278. 6. Blasi, J., Chapman, E. R., Link, E., Binz, T., Yamasaki, S., De Camilli, P., Su¨dhof, T. C., Niemann, H., and Jahn, R. (1993) Nature 365, 160 –163. 7. Schiavo, G., Santucci, A., DasGupta, B. R., Mehta, P. P., Jontes, J., Benfenati, F., Wilson, M. C., and Montecucco, C. (1993) FEBS Lett. 335, 99 –103. 8. Binz, T., Blasi, J., Yamasaki, S., Baumeister, A., Link, E., Su¨dhof, T. C., Jahn, R., and Niemann, H. (1994) J. Biol. Chem. 269, 1617–1620. 9. Niemann, H., Blasi, J., and Jahn, R. (1994) Trends Cell Biol. 4, 179 –185. 10. Blasi, J., Chapman, E. R., Yamasaki, S., Binz, T., Niemann, H., and Jahn, R. (1993) EMBO J. 12, 4821– 4828. 11. Atlas, R. M. (1998) Crit. Rev. Microbiol. 24, 157–168. 12. Wheeler, A. H. (1997) Am. Fam. Physician 55, 541–545, 548. 13. Thompson, D. E., Brehm, J. K., Oultram, J. D., Swinfield, T. J., Shone, C. C., Atkinson, T., Melling, J., and Minton, N. P. (1990) Eur. J. Biochem. 189, 73– 81. 14. Li, Y., Foran, P., Fairweather, N. F., de Paiva, A., Weller, U., Dougan, G., and Dolly, J. O. (1994) Biochemistry 33, 7014 –7020. 15. Vaidyanathan, V. V., Yoshino, K., Jahnz, M., Dorries, C., Bade, S., Nauenburg, S., Niemann, H., and Binz, T. (1999) J. Neurochem. 72, 327–337. 16. Fu, F. N., Sharma, S. K., and Singh, B. R. (1998). J. Protein. Chem. 17, 53– 60. 17. Clayton, M. A., Clayton, J. M. Brown, D. R., and Middlebrook, J. L. (1996) Infect. Immun. 63, 2738 –2742. 18. Zhou, L., de Paiva, A., Liu, D., Aoki, R., and Dolly, J. O. (1995) Biochemistry 34, 15175–15181. 19. Singh, B. R., and DasGupta, B. R. (1989) Mol. Cell. Biochem. 85, 67–73.