A VHH That Neutralizes the Zinc Metalloproteinase Activity of Botulinum Neurotoxin Type A*

The current treatment of botulism is to administer animal-derived antitoxin, which frequently causes severe adverse reactions in the recipients. In this study, a heavy chain antibody fragment (VH/VHH) phage display library was constructed by amplification of the immunoglobulin genes of a nonimmune camel, Camelus dromedarius, using primers specific to human VH gene segments. A recombinant light chain of type A botulinum toxin, BoTxA/LC, with zinc endoprotease activity was used in phage bio-panning to select phage clones displaying BoTxA/LC-bound VH/VHH. Soluble VH/VHH were produced and purified from 10 VH/VHH phagemid-transformed E. coli clones. Complementary determining regions (CDRs) and immunoglobulin frameworks (FRs) of the 10 camel VH/VHH-deduced amino acid sequences were determined. FR2 sequences of two clones showed a hallmark of camel VHH, i.e. (F/Y)42E49R50(G/F)52. The remaining eight clones had an FR2 amino acid tetrad of conventional VH, i.e. V42G49L50W52. VHH of one clone (VHH17) neutralized the SNAP25 hydrolytic activity of BoTxA/LC, whereas mouse polyclonal anti-BoTxA/LC did not have such activity. Mimotope sequences of VHH17 matched with the 194–206 amino acid residues of BoTxA/LC, which are located near the S′1 subsite of the catalytic cleft of the enzyme. Molecular docking revealed that CDR3 of the VHH17 bound to epitope in the toxin enzymatic cleft. Therefore, the BoTxA/LC neutralization by the VHH17 should be due to the VHH insertion into the enzymatic cleft of the toxin, which is usually inaccessible to a conventional antibody molecule. This antibody fragment warrants further development as a therapeutic agent for botulism.

Botulism is a clinical manifestation characterized by generalized flaccid paralysis and respiratory insufficiency, which may be fatal if not treated properly. It is caused mainly by consumption of food contaminated with neurotoxins of Clostridium bot-ulinum (BoTxs). 4 The BoTxs are zinc-dependent endopeptidases that cleave SNARE proteins used for the exocytosis of the neurotransmitter at the motor nerve end plate (1,2). BoTxs are recognized as the most potent toxic substance of humans with a lethal dose as low as 1 ng/kg body weight (3)(4)(5) and are classified as a category A bio-weapon by the Centers for Disease Control and Prevention (6 -7). Presently, there are seven antigenic types of BoTxs, serotypes A-G (3)(4)(5). Among these, serotype A causes the most serious clinical manifestations in humans due to its prolonged localization within the cytoplasm of the affected neuron (8).
The molecular structure of BoTxs has been revealed by crystallography as an A-B toxin (9,10). The two peptides are synthesized as a single polypeptide, which is modified post-translationally to a 150-kDa, di-chain active holotoxin. Each molecule of the holotoxin is composed of an A subunit or light chain (ϳ50 kDa), which is linked to a B subunit or heavy chain (ϳ100 kDa) by a single disulfide bond. The heavy chain is responsible for receptor binding, internalization, and translocation of the holotoxin into the endosome of cholinergic neurons (11). After an early endosomal exit, the light chain hydrolyzes SNARE proteins such as SNAP25 (for types A, C, and E BoTxs), synaptobrevin (for types B, D, F, and G BoTxs), and syntaxin (type C BoTx) resulting in the disruption of the neurotransmission process (12,13).
A licensed BoTx antagonist is not available. Patients with botulism are treated with animal-derived anti-BoTx antibodies together with supportive measures, such as artificial respiration. There are several drawbacks of using the antitoxin of heterologous species. The animal antibodies often elicit allergic reactions, which may be as serious as fatal anaphylaxis, as well as an anti-isotype/ idiotype response that causes serum sickness (6). Besides, a prolonged immunization process of the donor animals is required before a satisfactory level of the antitoxin is reached.
Because of their small size (ϳ15-20 kDa), high tissue-penetrating efficacy, and relative stability, single domain heavy * This work was supported by the Commission on Higher Education (CHE), chains (V H H) from a dromedary (Camelus dromedarius), which are devoid of a variable light chain domain have attractive molecular structures for a potent enzyme/toxin inhibitor (14 -20). V H H could directly recognize the conformational structure within the pocket of an enzyme active site, which can never be reached by the large sized conventional heavy light chain antibody (21)(22)(23). In this study, V H H produced from a phage clone derived from a VH/V H H phage display library constructed from immunoglobulin genes of B cells of a nonimmune Arabian camel, C. dromedarius, are used to bind specifically to the catalytic light chain of the type A botulinum neurotoxin and to inhibit the toxin endopeptidase activity. Experimental details and results are reported herein.

EXPERIMENTAL PROCEDURES
Production of a Full-length Recombinant Light Chain of Type A Botulinum Neurotoxin (BoTxA/LC)-Chromosomal DNA of serotype A C. botulinum was used as a template for amplifying a gene sequence encoding the full-length BoTxA/LC. The 1.4-kb DNA amplicon of the toxin gene segment was cloned into pQE30 expression vectors (Qiagen), and the recombinant expression vectors were introduced into competent SG13009 (pREP4) Escherichia coli cells by a heat-shock method. The transformed SG13009 (pREP4) E. coli cells were selected from an overnight Luria-Bertani (LB) agar plate containing 100 g/ml ampicillin and 25 g/ml kanamycin (LB-AK) and screened by PCR for the presence of the BoTxA/LC plasmid vectors. Selected transformed E. coli clones were individually grown in LB-AK broth at 25°C with shaking until the absorbance at 600 nm (A 600 nm ) was 0.2. Isopropyl-␤-D-thiogalactopyranoside (USB) was added to 0.5 mM, and the culture was incubated further at 25°C for 16 h. Bacterial cells were collected by centrifugation and sonicated in a lysis buffer (5 mM NaH 2 PO 4 , 300 mM NaCl, 10 mM imidazole, pH 8.3). The homogenate was centrifuged at 8,000 ϫ g at 25°C for 10 min. The recombinant BoTxA/LC in the bacterial lysate was purified by nickel-nitrilotriacetic acid-agarose (Invitrogen) according to the manufacturer's instruction.
Determination of the Enzymatic Activity of the Recombinant BoTxA/LC-The endopeptidase activity of the recombinant BoTxA/LC was determined by Western blot analysis and fluorescent assay. For Western blotting (24,25), 20 l of 10 nM recombinant BoTxA/LC were added to 200 g of a SK-N-MC human neuroblastoma cell lysate in a working buffer (40 mM HEPES, pH 7.4, and 0.5 mM ZnCl 2 ), and the mixture was incubated at 37°C for 24 h. The preparation was subjected to SDS-PAGE, transblotted onto a nitrocellulose membrane (NC), and probed with rabbit polyclonal anti-SNAP25 antibodies (Zymed Laboratories Inc.), which recognized only intact SNAP25. Goat anti-rabbit immunoglobulin-alkaline phosphatase (AP) conjugate (Southern Biotech) served as secondary anti-isotype antibody and 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium TM (KPL) was used as AP substrate. For the fluorescent assay, the methods described by Schmidt (26) and Palmier (27) were followed with modifications. Briefly, a fluorogenic oligopeptide substrate representing amino acid residues 187-203 of SNAP25, which is a cleavage motif of BoTxA, was synthesized (Anaspec, Inc.). The sequence of the synthetic peptide was SNRTRIDEAN(N-2,4-dinitrophenyl-K)RA(3-iodoacetamido-4-methyl-7-dimethylaminocoumarin-C-RML, which would be cleaved by BoTxA/LC between the Lys 197 and Arg 198 residues. The fluorescence signal from 3-iodoacetamido-4methyl-7-dimethylaminocoumarin was quenched by N-2,4-dinitrophenyl in the intact fluorogenic substrate but became detectable upon the substrate cleavage when exposed to active BoTxA/LC. Hydrolytic rates of the fluorogenic substrate by BoTxA/LC were measured as the following: mixtures of various amounts of fluorogenic substrate (10, 12.5, 15, 17.5, 20, 25, and 50 M) and a fixed amount (0.25 M) of BoTxA/LC were separately made in 60 l of working buffer contained in wells of a clear bottom 96-well assay black plate (Costar, Corning) at 25°C. The excitation and emission maxima were 398 and 485 nm, respectively. The fluorescence signal of each reaction mixture was monitored at 2-s intervals using the VarioSkan Flash microplate reader (Thermo Fisher Scientific). Initial hydrolysis rates or velocity (V i ) were derived from 5 to 8 min of the reactions and expressed as fluorescence units/sec. The V i values were fitted with the Michaelis-Menten equation by using GraphPad Prism5 software to calculate maximum velocity (V max ) and K m . The V i units were from conversion of fluorescence units/min into pmol/min by using an equation derived from a standard curve. The standard curve was constructed by plotting varying amounts of fluorescent products, (RA(DA-CIA)CRML), from 5 pmol to 1 nmol against the fluorescence intensities. The k cat was then calculated by dividing the V max with molarity of the enzyme concentration.
Preparation of Mouse Anti-BoTxA/LC-Three ICR mice (from the National Laboratory Animal Center, Mahidol University, Nakhon-Pathom, Thailand) were housed for 1 week before commencing immunization. Each mouse received three injections of 10 g of purified BoTxA/LC at weekly intervals. The immunogen was mixed with an equal volume of Alum's adjuvant (Pierce, Thermo Fisher Scientific) and administrated intraperitoneally. One week after the last booster, the mice were bled, and serum samples were collected separately. The titers of the antibodies to BoTxA/LC in the immune sera were measured by indirect ELISA using goat anti-mouse immunoglobulin-horseradish peroxidase conjugate (Southern Biotech), and a chromogenic substrate, i.e. 2,2Ј-azino-di(3-ethylbenzthiazoline-6-sulfonate) (Kirkegaard and Perry Laboratories, Inc.), as detection reagents. The optical density (OD) of the content of each well was determined at A 405 nm against the background (wells with normal mouse serum). The immune sera were kept at Ϫ20°C until use.
Construction of the Phage Display C. dromedarius VH/V H H Library-Peripheral blood mononuclear cells were isolated from 50 ml of venous blood collected from an eight-month-old naïve male dromedary (C. dromedarius) from the Chockchai Farm, Nakhon Ratchaseema province, Thailand, using Ficoll-Paque TM (Amersham Biosciences). The total RNA was extracted from the cells by TRIzol TM reagent (Invitrogen), and the mRNA was reverse-transcribed to cDNA using Revert Aid TM (Fermentas Life Sciences). The gene fragments encoding variable domains of the dromedary VH/V H H were PCRamplified using the cDNA as template, as well as 14 forward and three reverse human immunoglobulin-specific primers (28). Each primer sequence was flanked with SfiI and NotI endonuclease restriction sites at the 5Ј and 3Ј ends, respectively. During PCR, the immunoglobulin forward primers annealed to the 5Ј ends of the VH/V H H exons and the reverse primers to the 3Ј ends of the JH exons of all immunoglobulin gene families. The amplified products were verified by agarose gel electrophoresis, and the VH/V H H DNA amplicons of ϳ400 bp were extracted from the agarose gel slabs and purified using GeneClean TM II kit (MP Medicals). The purified DNA was digested with SfiI and NotI endonucleases and ligated into a pCANTAB5E phagemid vector TM (Amersham Biosciences) precut with the same enzymes, and the ligation mixture was introduced into competent TG1 E. coli cells by the electroporation method. An aliquot of the transformation mixture was plated onto a sodium, bacto tryptone, ampicillin, and glucose agar plate, and the plate was incubated at 37°C overnight to estimate the transformation efficiency. The VH/V H H-displaying phage particles were rescued from the remaining portion of the transformation mixture by co-infecting the phagemidtransformed E. coli with M13K07 helper phages, and the titer of the rescued phage repertoire was determined (28).

Selection and Production of BoTxA/LC-specific VH/V H H-
The recombinant BoTxA/LC was used as an antigen in a phagebio-panning process to select phage clones displaying VH/V H H that bound to the protein (28). The protein (5 g) was immobilized on the surface of each well of a microtiter ELISA plate (Costar, Corning). The so-constructed phage display VH/V H H library (ϳ5 ϫ 10 10 phage particles) was added into the antigencoated wells and kept at 25°C for 1 h. Unbound phage particles were removed by successive washing with a washing buffer (0.15 M phosphate-buffered saline, pH 7.4, containing 0.05% Tween 20; washing buffer). Bound phage particles were directly supplemented with exponential phase grown HB2151 E. coli cells. The phage transformed HB2151 E. coli cells were grown on LB-AG (LB-A containing 2% of glucose) selective agar plates. HB2151 E. coli colonies carrying VH/V H H-inserted phagemids were screened by PCR-based VH/V H H gene amplification. The PCR primers were: forward (R1) 5Ј-CCATGATTACGCCAAGC-TTT-3Ј and reverse (R2) 5Ј-GCTAGATTTCAAAACAGCAGA-AAGG-3Ј. The VH/V H H amplicon size including the plasmid vector sequence was ϳ600 bp. The transformants carrying the recombinant phagemid vectors were further screened for their ability to express VH/V H H by Western blot analysis (28). Briefly, each selected E. coli clone was grown in LB-A broth, and the VH/V H H was induced to express itself with 0.5 mM isopropyl-␤-D-thiogalactopyranoside. The VH/V H H in the bacterial lysate was detected by Western blot analysis using mouse anti-E Tag TM antibody (Amersham Biosciences), goat anti-mouse immunoglobulin-AP conjugate and 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium substrate, respectively. The VH/V H H in the bacterial lysates of the selected clones was purified using ion exchange chromatography. The bacterial lysate containing VH/V H H was equilibrated with 50 mM Tris-HCl, pH 8.0. The preparation was added to mix with DEAE-Sepharose TM Fast Flow packed beads (Pharmacia), and the tube was rotated for 1 h on a horizontal platform. The Sepharose beads were allowed to set, and the supernatant containing the VH/V H H was collected. Purity of the antibody was checked by SDS-PAGE and stained with Coomassie Brilliant Blue G-250 dye (USB). The preparation was dialyzed against phosphate-buffered saline. High purity VH/V H H was prepared by subjecting the preparation in phosphate-buffered saline to an affinity anti-E-tag column (GE Healthcare Bio-Sciences AB).
Detection of the Binding of the Phage-derived-VH/V H H to the BoTxA/LC-Indirect and dot-ELISAs were used to detect the binding of the VH/V H H derived from different recombinant phagemid-transformed HB2151 E. coli clones. For indirect ELISA, 1 g of purified recombinant full-length BoTxA/LC or BSA, which served as an antigen control, in 100 l of carbonatebicarbonate buffer, pH 9.6, was added to each well of an ELISA plate. The antigen-coated wells were blocked with 1% BSA in phosphate-buffered saline and were then incubated with individual HB2151 E. coli lysates. After washing, the amount of bound VH/V H H in each well was detected using mouse anti-E Tag antibody, goat anti-mouse immunoglobulin-horseradish peroxidase conjugate (Southern Biotech), and ABTS substrate. The OD of the content of each well was determined at A 405 nm against a blank (wells filled with phosphate-buffered saline instead of VH/V H H-containing E. coli lysate). A well filled with a lysate of normal HB2151 E. coli instead of a V H H-containing E. coli lysate served as a background control. The VH/V H H in the E. coli lysate that yielded an OD two times higher than that of the BSA control was selected. For the dot-ELISA, 3-l aliquots of the BoTxA/LC (ϳ1 g) were dotted on separated nitrocellulose membranes (NCs; Hybond ECL, Amersham Biosciences). The antigen-dotted membrane was immersed into a solution of 3% BSA in washing buffer for 1 h, washed with washing buffer, and cut into squares with one antigen dot on each square. subjected to SDS-PAGE and probed with rabbit polyclonal antibody to SNAP25 followed by goat anti-rabbit immunoglobulin-AP conjugate and substrate. The intensity of the SNAP25 protein band was quantified by densitometric analysis with AlphaDigiDoc TM 1201 software (version 3.3.0, Alpha Innotech). Percent cleavage inhibition of SNAP25 by VH/V H H was calculated: % inhibition ϭ (intensity value of sample Ϭ intensity value of 100% inhibition control) ϫ 100. The neutralizing activity of the selected clone was confirmed by mixing the antibody with the toxin either at a molar ratio of 3:1 or as otherwise indicated and incubated at 37°C for 1 h before subjecting to the fluorescent assay. A zinc-chelating agent, i.e. N,N,NЈ,NЈ-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN, Sigma) was used for positive toxin inhibition control.
Affinity Measurement of the VH/V H H-The affinity (dissociation constant, K d ) of the VH/V H H was determined by using ELISA as described previously (30). Briefly, 10 nM of the VH/V H H was mixed with various amounts of soluble BoTxA/ LC, and the tubes containing the reaction mixtures were kept at 20°C for 16 h. Free VH/V H H was detected by ELISA. The contents of each tube (100 l) were transferred to a well of a microtiter plate previously coated with 1 g of BoTxA/LC. After incubation, VH/V H H captured by the immobilized toxin was detected by using mouse monoclonal anti-E Tag, goat antimouse IgG-horseradish peroxidase conjugate, and 3,3Ј,5,5Јtetramethylbenzidine substrate (Zymed Laboratories Inc.), respectively. OD of the content of each well was determined at A 450 nm against a blank (well filled with diluent instead of VH/V H H). K d was then calculated from a Klotz plot by linear regression analysis.
Determination of the VH/V H H Mimotope Sequence-The mimotopes of the VH/V H H that gave the highest BoTxA/LCneutralizing activity were determined by using a phage display 12-mer peptide library (Ph.D.-12 TM Phage Display Peptide Library, New England Biolabs) as described previously (28). VH/V H H (1 g) was immobilized on the surface of each ELISA well at 4°C overnight. After washing the well with Tris-buffered saline, pH 7.5, containing 0.1% Tween 20, the empty sites in the well were blocked by adding 200 l of 0.5% BSA in Tris-buffered saline for 1 h. After washing to remove the blocking reagent, 100 l of the phage display 12-mer peptide library (diluted 1:10 with Tris-buffered saline, pH 7.5, containing 0.1% Tween 20), which contained ϳ1.5 ϫ 10 11 plaque-forming units were added to the well, and the plate was kept at 25°C for 1 h. Unbound phages were washed away with Tris-buffered saline, pH 7.5, containing 0.1% Tween 20, the bound phages were eluted with 0.2 M glycine-HCl solution, and the pH of the solution was immediately brought to neutral by adding 2 M Tris base. The eluted phages were allowed to amplify by infecting 20 ml of log phase-grown ER2738 E. coli (OD at A 600 nm ϳ 0.3). After removing the bacterial cells by centrifugation at 12,000 ϫ g, the phages in the supernatant were concentrated by adding polyethylene glycol/NaCl. The phages in the pellet were resuspended in phosphate-buffered saline and used in the next round of bio-panning. Three rounds of bio-panning were performed. The eluted phages from the third round of bio-panning were used to infect the ER2738 E. coli in top agarose overlaid on the LB agar containing isopropyl-␤-D-thiogalactopyranoside and 5-bromo-4-chloro-3-indolyl-␤-D-galactopyranoside (X-gal). Well isolated blue plaques were randomly picked, inoculated individually in 1 ml of 1:100 overnight diluted ER2738 E. coli culture in LB broth, and incubated at 37°C with shaking for 4 h. After removing the bacterial cells by centrifugation as above, the DNA of the individual phage clones were extracted from the culture supernatant using the phenol/chloroform method. The DNA of each phage clone was sequenced, and the mimotope peptides were   1, 2, 7, 10, 11, 19

RESULTS
Recombinant BoTxA/LC-A full-length recombinant light chain of type A botulinum neurotoxin (BoTxA/LC), ϳ50 kDa, was successfully produced and purified. The nucleotide sequence of the BoTxA/LC was verified by DNA sequencing, and the deduced amino acid sequence showed 100% homology to the sequence deposited in the GenBank IM database (accession number M30196). The produced recombinant BoTxA/LC had endopeptidase activity for SNAP25 as determined by Western blot analysis (Fig. 1), which digested SNAP25 in the neuroblastoma cell lysate treated with the BoTxA/LC, indicating that the SNARE protein was completely degraded by the toxin, whereas the untreated cell lysate (negative cleavage control) showed the SNAP25 band at ϳ27 kDa. The lysate treated with trypsin (positive cleavage control) showed no SNAP25 band. From fluorescent assay, the kinetic parameters of the recombinant BoTxA/LC including k cat and K m were 2.03 s Ϫ1 and 109.7 Ϯ 32.38 M, respectively.

V H H That Neutralizes Botulinum Toxin A
Selection of Phage Clones Displaying VH/V H H That Bound to BoTxA/LC-After a single round of phage bio-panning with immobilized recombinant BoTxA/LC, 60 transformed HB2151 E. coli were randomly picked from the selective agar plate and screened for the presence of the VH/V H H-coding sequences by PCR using the R1 and R2 primers. It was found that 39 of 60 clones (65%) were positive for VH/V H H amplicons. The lysates of these E. coli clones grown under isopropyl-␤-D-thiogalactopyranoside induction were subjected to Western blotting, and 28 clones (46.6%) could express the VH/V H H proteins seen as bands at ϳ15-20 kDa (Fig. 3) (Fig. 4A).
Binding of the VH/V H H of the 10 clones to BoTxA/LC was confirmed by dot-ELISA (Fig. 4B). The VH/ V H H did not bind to recombinant light chain of type B botulinum neurotoxin, zinc-dependent metalloprotease lethal factor of anthrax toxin, neuraminidase of H5N1 virus, and lysate of the SK-N-MC neuroblastoma human cell line (data not shown). Fig. 5 shows the RFLP patterns of the VH/V H H nucleotide sequences of the 10 clones, which revealed completely different banding patterns. Fig. 6 shows the ClustalW multiple sequence alignment of the deduced amino acid sequences including the immunoglobulin frameworks and the CDRs of the 10 VH/V H H clones. It was found that the FR2 sequences of clones 17 and 21 carried an amino acid tetrad, i.e. (F/Y) 42 E 49 R 50 (G/F) 52 , which is a hallmark of camel V H H. The  remaining eight clones lacked the tetrad but, instead, had the features of FR2 of conventional VH of mammals including human, mouse, and camelid, i.e. Val 42 -Gly 49 -Leu 50 -Trp 52 . Previous data have demonstrated that the highest similarity between camel V H H and human VH was 82.6% (32). In our data, the deduced amino acid sequences of the amplified camel VH using human primers showed 73.68 -100% homology (average 91.26%) with the human VH, whereas the V H H using human primers showed 58.82-91.39% homology (average 76.13%) with the human VH (Table 1). A marked difference between the V H H and human VH was found at the tetrad amino acids of FR2, which determines the hydrophilicity of the former and hydrophobicity at the variable light chain-binding site of the latter.
Neutralization Tests-VH3, VH15, VH20, VH22, VH26, and VH27 and V H H17 and V H H21 were screened for their ability to neutralize the enzymatic activity of the recombinant BoTxA/ LC. Fig. 7 shows the results of the Western blot assay testing the neutralizing activities of the VH/V H H of the eight clones. Unfortunately, clones 5 and 11 lost their ability to express VH. On an equal weight basis, VH15 and VH22 and V H H17 efficiently inhibited (83.33, 75.0, and 83.33%) the hydrolysis of SNAP25 by 10 nM recombinant BoTxA/LC, as shown by the presence of intensely stained SNAP25 bands in Fig. 7 (lanes 4, 8, and 5, respectively) and the densitometer readouts at the bottom of Fig. 7. VH20, VH26, and VH27 had less BoTxA/LC inhibitory activity (54.17, 41.67, and 50.0%) than the former three clones (lanes 6, 9, and 10). The VH of clone 3 showed trace toxin-inhibitory activity (29.17%; Fig. 7, lane 3), whereas   V H H21 showed the lowest toxin-inhibitory activity (8.33%) (Fig. 7, lane 7). VH15, VH22, and V H H17 were tested further for their toxin neutralization by using a fluorogenic substrate. The results of the fluorescent-based assay showed that V H H17 exhibited the highest % inhibition of hydrolysis of SNAP25 (73% inhibition), whereas VH15 and VH22 exerted 59 and 64% inhibition, respectively (Fig. 8). The mouse polyclonal antibody to BoTxA/LC at dilution 1:1,000 did not have any BoTxA/LCneutralizing activity. V H H17 inhibited the endopeptidase activity of the BoTxA/LC in a dose-dependent manner. At 4.5 and 10 M (contained 13.5 ϫ 10 13 and 30.1 ϫ 10 13 molecules, respectively), the V H H17 exhibited 73 and 92% inhibition of hydrolysis of SNAP25 by 1.5 M of BoTxA/LC (contained 4.5 ϫ 10 13 molecules), respectively. Affinity of V H H-The dissociation constant (K d ) of the V H H17 derived from the Klotz plot was 11.6 nM.
Mimotope of V H H-10 randomly picked blue plaques were separately added to log phase-grown ER2738 E. coli. The DNA of each amplified 12-mer peptide display phage clone was extracted and sequenced, and the amino acids were deduced. The 10-amino acid mimotopes of the V H H17 (designated M1-M10) were aligned with the reference sequence of the BoTxA/LC of the GenBank TM database (accession number AAA23262.1). It was found that 7 of 10 sequences, i.e. M1 and M3-M8, matched with the amino acid residues 194 -206 of BoTxA/LC (Fig. 9).
Molecular Docking-Interface binding of BoTxA/LC and V H H17 is shown in Fig. 10. The best docking result (the lowest docking energy of Ϫ20.39661 kcal/mol), which was obtained from using the ZDOCK and RDOCK modules, showed that the CDR3 of the V H H17 bound at the BoTxA/LC enzymatic groove.

DISCUSSION
Because of their small size (ϳ15-20 kDa) and long CDR3 sequences, which can be readily extended into the enzymatic cleft, the variable domains of camelid heavy chain antibodies (V H H domain) have been shown to be potent enzyme inhibitors (22,23,33,34). V H H-conjugated with ␤-lactamase has been shown to exert anti-cancer activity (35). V H H human trypanolytic factor conjugate was effective in the treatment of experimental African trypanosomiasis (36). Nevertheless, data in the literature are limited concerning the use of the V H H as a therapeutic agent for intoxications caused by bacterial toxic enzymes such as botulism. Current botulism immunotherapy based on animal-derived antitoxin is facing several obstacles. The polyclonal antibodies neutralize only circulating extracellular BoTx but are ineffective for the internalized toxin. Besides, the animal proteins induce anti-isotype/ anti-idiotype responses in the antitoxin-treated human patients (6). Antibody that specifically interferes with the enzymatic activity of the botulinum toxin has never been available.
Recently, camel V H H (Nanobody, Ablynx) specific to von  The red loop represents the V H H17 CDR3 region (residues 97-116). The toxin-binding interface is shown in yellow, blue, and brown; yellow represents the enzymatic groove. Blue represents histidine and glutamic acid in the zinc-binding motif (HELIH) of the toxin. Brown represents a phenylalanine residue in the TFGFEESLEVDTNP sequence, which represents the mimotope-binding region.