Characterization of a Neutralizing Monoclonal Antibody Against Botulinum ADP- Ribosyltransferase, C3 Exoenzyme

FULL PAPER Bacteriology

Characterization of a Neutralizing Monoclonal Antibody against Botulinum ADP- Ribosyltransferase, C3 Exoenzyme

Yoichi KAMATA1), Hidenobu HOSHI1), Hayato CHOKI1) and Shunji KOZAKI1)

1)Department of Veterinary Science, College of Agriculture, Osaka Prefecture University, 1–1 Gakuen-cho, Sakai, Osaka 599–8531, Japan

(Received 27 February 2002/Accepted 24 April 2002)

ABSTRACT. A monoclonal antibody, named C302, was prepared and characterized against botulinum ADP-ribosyltransferase C3 exoen- zyme that inactivates RhoA GTP-binding protein, resulting in the neurite outgrowth of human neuroblastoma GOTO cells. C302 bound not to the smaller fragments derived from the protease-treated C3 exoenzyme but to the intact C3 exoenzyme. It seems that the C302 epitope may depend on the three-dimensional structure of C3 exoenzyme molecule. C302 depressed the enzymatic and biological acti o n s of C3 exoenzyme. The dose-dependent depression pattern of C302 on the enzyme activity was similar to that to the biological one. C302 turned the neurite-bearing shape of the C3 exoenzyme-treated GOTO cells into the intact shape. By using of C302 mAb and C3 exoen- zyme, the research concerning GTP-binding proteins would be improved. KEY WORDS: ADP-ribosylation, biological effect, botulinum C3 exoenzyme, monoclonal antibody. J. Vet. Med. Sci. 64(9): 767–771, 2002

In 1987, an ADP-ribosylation activity was discovered in useful for the structure-function relationship of C3 exoen- Clostridium botulinum type C1 [16] and D toxin preparation zyme. In this communication, we report the preparation and [17]. First, this enzymatic activity has been recognized as characterization of a mAb against C3 exoenzyme with a the ultimate toxic function of botulinum neurotoxin, higher neutralizing activity. because such enzyme activity was found to migrate into the toxin molecule [10]. Finally, however, Aktories et al. clari- MATERIALS AND METHODS fied that the activity was from a contaminated protein in the toxin preparation [2, 3]. Since an ADP-ribosylating toxin C3 exoenzyme and a monoclonal antibody: The purifica- among the toxins that C. botulinum produces has already tion procedure for C3 exoenzyme was described elsewhere been reported to be C2 toxin, this enzyme was named C3 [2, 11]. In brief, C3 exoenzyme was isolated from a culture exoenzyme [3]. C3 exoenzyme is a protein with 23-kDa supernatant of Clostridium botulinum type C1 strain 003-9 molecular weight [14]. The target for C3 exoenzyme in by ammonium sulfate precipitation, chromatography on mammalian cells was identified as a GTP-binding protein CM-Sephadex (Amarcham Biosciences, Tokyo), and gel fil- named RhoA [1, 13]. RhoA is the famous protein as a cell- tration of Sephacryl S-100 HR (Amarcham). C3 exoen- signal switching molecule, involved in many cell functions zyme was dialyzed against 0.1 M phosphate buffer, pH 7.0, such as stress fiber formation, cell motility and migration, and filtered through a 0.2 µm membrane to sterile and then smooth muscle contraction, etc. [6, 12]. A primal function stored at –20°C. A hybridoma producing a monoclonal anti- of RhoA is the organization of intracellular non-muscular body (mAb) was established by a general procedure using actin. C3 exoenzyme transfers an ADP-ribose moiety of mouse myeloma cells (X63Ag.653), spleen cells of nicotinamide adenine dinucleotide (NAD) to Asn41 of BALB/c mouse immunized with C3 exoenzyme, and poly- RhoA protein [19]. The ADP-ribosylation results in dys- ethylene glycol 4000 (Merck, Darmstadt). Screening has function of RhoA. The inactivated RhoA induced neurite done by enzyme-linked immunosorbent assay (ELISA) and formation in human neuroblastoma GOTO cells [7] and rat western blotting. An mAb-producing hybridoma was estab- pheochromocytoma PC12 cells [15]. C3 exoenzyme lished, and its mAb was purified from the ascitic fluid by evoked the outgrowth of neurites from a chicken ganglion protein-A (Amarcham) affinity chromatography. The [7] and induced acetylcholinesterase activity in PC12 cells immunoglobulin class and subclass and the type of the light [15], NG108-15, and C6 cell lines [8]. The primary culture chain were determined by an Immunoglobulin typing kit cells of mouse cerebrum treated with C3 exoenzyme (Wako Pure Chemicals, Osaka). Protein concentration was showed a response similar to that of the control matured determined by Bradford’s method [4]. The molecular cells against various drugs such as muscarine, and acetyl- weight of mAb was defined as 150-kDa. choline as a neurotransmitter [21]. These observations then ELISA was done by the following procedures. C3 exoen- indicate that C3 exoenzyme is regarded as a kind of neuro- zyme or bovine serum albumin (Sigma Chemicals, St. tropic factor. Although the research concerning RhoA has Louis) was added into a well of an ELISA plate (Iwaki, Fun- been markedly improved by the use of C3 exoenzyme as a abashi). After washing with 0.05% tween 20 (Wako) in specific inhibitor, the properties of C3 exoenzyme itself has PBS, the wells were blocked by incubation of 3% bovine not been elucidated. Monoclonal antibodies (mAbs) will be serum albumin (BSA, Fraction V, Sigma) to minimize non- 768 Y. KAMATA, H. HOSHI, H. CHOKI AND S. KOZAKI specific reactions. Purified mAb was added into the wells at various concentrations. After washing, horseradish peroxi- dase (HRP)-labeled anti-mouse IgG (Bio-rad, Tokyo) was added. Hydrogen peroxide and o-phenylenediamine (Wako) were used. The reaction products were measured for their absorbance at 450 nm using a microplate reader (Model 550, Bio-rad). ADP-ribosylation of RhoA with C3 exoenzyme: RhoA protein was kindly provided by Dr. Yasuhiro Ohashi (Japan Shelling Pharmaceutical, Osaka). ADP-ribosylation by C3 exoenzyme was done by the method described previously [7, 17]. One µg of RhoA was mixed in 50 µl of reaction buffer (0.1 M Tris-HCl buffer, pH 7.5, 50 mM nicotinamide, 10 µM of [32P]-NAD (1.25 µCi, New England Nuclear, Bos- ton) with 1 ng of C3 exoenzyme in the presence of C302. After addition of 50 µl of 1% BSA, the mixture was incu- bated at 30°C for 15 min and then 100 µl of 20% trichloro- acetic acid was added to terminate the reaction. The ° Fig. 1. Titration of a monoclonal antibody C302 against bot- mixture was centrifuged at 10,000 rpm for 15 min at 4 C, ulinum ADP-ribosyltransferase C3 exoenzyme in ELISA and the supernatant was discarded. The precipitate was dis- and Western blotting (inlet). Open circle; C3 exoenzyme, solved in a scintillation cocktail (Scintisol EX-H, Wako), Closed circle; Bovine serum albumin. Inlet: Lane 1, 30 ng; and radioactivity was determined in duplicate. The experi- Lane 2, 10 ng; lane 3, 3 ng; Lane 4, 1 ng of C3 exoenzyme. ments were repeated twice. The experimental procedures were in the text. Treatment of GOTO human neuroblastoma cells with C3 exoenzyme: GOTO cells (Japan Health Science Foundation, Osaka) were grown in Dulbecco’s modified Eagle’s exoenzyme: The cell fusion was performed twice, followed medium (DMEM; Gibco BRL, New York) containing 10% by the screening and five clones were selected. After the fetal bovine serum (FBS; Gibco). GOTO cells were plated purification, the five mAbs were examined for their neutral- on a 24-well culture plate at 2 × 104 cells/ml at 0.5 ml/well izing activity against the biological action of C3 exoen- and then cultured in 5% CO2 atmosphere at 37°C. After 24 zyme. Among five, only one mAb named C302 neutralized hr, the culture supernatant was aspirated, and C3 exoenzyme the activity of C3 exoenzyme. C302 belonged to IgG1 with (1 µg/ml) or a mixture of C3 exoenzyme and an mAb diluted a κ-light chain. Figure 1 shows the titration of C302 by in DMEM-10% FBS was added into the wells. The GOTO ELISA and Western blotting. C302 did not react to a control cells were photographed after 24-hr incubation, and the rate protein, but a positive signal was observed at 1 ng/ml of of the neurite-bearing cells was measured among 100 or C302 to C3 exoenzyme on ELISA. The western blotting more cells in duplicate. The experiments were repeated showed 1 ng/lane of C3 exoenzyme indicating a positive twice. signal to C302. To the author’s knowledge, C302 would be Protease digestion of C3 exoenzyme: Staphylococcus ranked as the highest-responding mAb in the system using aureus V8 protease (Sigma), lysyl-endopeptidase (Wako) o-phenylenediamine and 3–3’ diaminobenzidine. and chymotrypsin (Sigma) were added to a C3 exoenzyme The response of C302 mAb to the fragments of the solution (500 µg/ml) at a ratio of 100 (C3 exoenzyme) to 1 digested C3 exoenzyme was examined to determine where (protease) independently. After an appropriate incubation the C302 epitope located on C3 exoenzyme molecule (Fig. period at 37°C, sodium dodecylsulfate-polyacrylamide gel 2). The protease treatment of C3 exoenzyme resulted in electrophoresis (SDS-PAGE) sample buffer was added, and many fragments showing various molecular sizes. The size the mixture was boiled briefly. The digested products were ranged from 23-kDa to less than 5 kDa. When C302 was separated by tricine SDS-PAGE [18], transferred onto a allowed to react to the fragments, although many bands gave western blotting membrane (PVDF membrane, Millipore, positive signals, a 10-kD fragment in the chymotrypsin- Tokyo), and the membrane was treated with 3% skim milk digested C3 exoenzyme products was the smallest. No pos- in PBS. After washing with 0.1% tween 20 in PBS, the itive signals were observed among the fragments of less membrane was immersed in a mAb solution, and then HRP- than 10-kDa molecular weight. Because the size of this 10- labeled anti-mouse IgG, hydrogen peroxide, and 3–3’ kDa fragment was large enough to present many epitopes, diaminobenzidine (Sigma) were used to visualize the reac- the C302 epitope may alter or disassemble due to a decrease tion signal. in the molecular size of C3 exoenzyme. The portions which were digested by the protease treatment might help to RESULTS AND DISCUSSION present the C302 epitope. This suggested that the C302 epitope would depend on the three-dimensional structure of Immunological characterization of a mAb against C3 C3 exoenzyme molecule. NOVEL MAB TO BOTULINUM C3 EXOENZYME 769

Fig. 4. Effects of C302 mAb on the biological activity of C3 exoenzyme with human neuroblastoma GOTO cells. The activ- Fig. 2. SDS-PAGE and Western blotting analyses for C302 ity of C3 in the absence of C302 was defined as 100%. The epitope. C3 exoenzyme was digested with V8 protease (panel ratio of C302 to C3 exoenzyme was calculated from the molec- A), chymotrypsin (panel B), and lysyl-endopeptidase (panel ular weight: C3 exoenzyme, 23 kDa; C302, 150 kDa. C). Lane 1; SDS-PAGE, Lane 2; Western blotting.

Fig. 5. Effect of C302 mAb on the neurite-bearing GOTO cells Fig. 3. Effects of C302 mAb on ADP-ribosyltransferase treated with C3 exoenzyme. After a complete change in the activity of C3 exoenzyme. The activity of C3 in the shape of the GOTO cells with C3 exoenzyme (1 µg/ml, 24-hr absence of C302 was defined as 100%. The ratio of C302 incubation), the cells were washed with DMEM-10% FCS to C3 exoenzyme was calculated from the molecular and were treated with C302 mAb. The rate of the neurite- weight: C3 exoenzyme; 23 kDa, C302; 150 kDa. The pre- bearing cells was measured after a further 24-hr incubation. cise experimental procedure was in the text. Toratani et al. [20] reported four neutralizing mAbs for the ADP-ribosylation activity of C3 exoenzyme. The IC50 of Effects of C302 mAb on ADP-ribosyltransferase activity the four mAbs was between 500 to 3,000 ng/ml to 1 ng/ml of C3 exoenzyme: The ADP-ribosylating activity of C3 of C3 exoenzyme, suggesting a ratio of 70 to 450 of mAb to exoenzyme was depressed by C302 dose-dependently (Fig. C3 exoenzyme (Fig. 2 of ref. 20). This indicated that C302 3). When C302 was added to the C3 exoenzyme-containing possessed a 23- to 150-times higher neutralizing activity mixture, the activity quickly decreased up to a ratio (C302 to than that of the four mAbs. Figure 3 showed two phases of C3) of 5, and then a gradual decrease occurred at a ratio of the C302 depression of the ADP-ribosylation activity of C3 more than 5. When C302 was added at a ratio of more than exoenzyme. ADP-ribosylation activity is distinguishable 10, almost no activity was observed. The IC50 value of from glucosidase and transferase activities to NAD. C302 was estimated to be 3 based on the ratio from Fig. 3. Although no experimental results exist, C302 might affect 770 Y. KAMATA, H. HOSHI, H. CHOKI AND S. KOZAKI both enzymatic actions, resulting in the two-phase depres- ulinum ADP-ribosyltransferase C3. Biochem. Biophys. Res. sion. Commun. 158: 209–213. Effects of C302 mAb on the biological activity of C3 2. Aktories, K., Rösener, S., Blashke, U. and Chhatwal, G. S. exoenzyme on GOTO cells: The cell-shape-change activity 1988. Purification of the enzyme and characterization of the of C3 exoenzyme was examined in the presence of C302 ADP-ribosylation reaction in platelet membranes. Eur. J. Bio- chem. 172: 445–450. (Fig. 4). When C302 was added, the activity quickly and 3. Aktories, K., Weller, U. and Chhatwal, G. S. 1987. Clostridium markedly decreased to about 25% at a ratio of 2. At a ratio botulinum type C produces a novel ADP-ribosyltransferase dis- of up to 7, the activity was maintained but then gradually tinct from botulinum C2 toxin. FEBS Lett. 212: 109–113. decreased at a ratio of 20. As seen in Fig. 4, the inhibition 4. Bradford, M. M. 1976. A rapid and sensitive method for the pattern of C302 on the biological activity of C3 exoenzyme quantitation of microgram quantities of protein utilizing the showed three phases. The quick and gradual depression in principle of protein-dye binding. Anal. Biochem. 72: 248–254. Fig. 4 was similar to those in Fig. 3, suggesting that such 5. Han, S., Arvai, A. S., Clancy, S. B. and Tainer, J. A. 2001. depression of the biological activity results from the depres- Crystal structure and novel recognition motif of Rho ADP- sion of the enzyme activity. ribosylating C3 exoenzyme from Clostridium botulinum: struc- After a complete change in the shape of the GOTO cells tural insights for recognition specificity and catalysis. J. Mol. Biol. 305: 95–107. with C3 exoenzyme, C302 was added, and the rate was then 6. Kaibuchi, K., Kuroda, S. and Amano, M. 1999. Regulation of determined after a further 24-hr incubation. Although a the cytoskeleton and cell adhesion by the Rho family GTPases much higher concentration of C302 was needed, the shape- in mammalian cells. Annu. Rev. Biochem. 68: 459–486. changed cells returned to their original shape in a dose- 7. Kamata, Y., Nishiki, T., Matsumura, K., Hiroi, T. and Kozaki, dependent manner (Fig. 5). Such an effect was not observed S. 1994. Morphological effects, rate of incorporation, and the with the addition of an mAb non-related to C3 exoenzyme enzymatic action of botulinum ADP-ribosyltransferase, known (data not shown). Our previous observation indicated that as C3 exoenzyme, on human neuroblastoma GOTO cells. C3 exoenzyme binds to the cell membrane first and very few Microbiol. Immunol. 38: 421–428. C3 molecules enter the cytosol, then inactivates RhoA [7]. 8. Komagome, R. 1994. Effect of botulinum ADP-ribosyltrans- ferase (Exoenzyme C3) on neural cells. Jpn. J. Med. Res. 42: Many C3 molecules would be retained on the cell mem- 44. brane. The antibody is a large molecule not able to enter the 9. Maehara, T., Takahashi, K., Ohoka Y., Ohtsuka, T., Ui, M. and cells; therefore, C302 may neutralize the biological function Katada, T. 1991. Identification of a botulinum C3-like enzyme of C3 exoenzyme, resulting from the binding to C3 mole- in bovine brain that catalyzed ADP-ribosylation of GTP-bind- cules fixed on the outside of the cell membrane. Further ing proteins. J. Biol. Chem. 266: 10062–10065. experiments are needed to understand the total neutraliza- 10. Morii, N., Ohashi, Y., Nemoto, Y., Fujiwara, M., Ohnishi, Y., tion mechanism of C302 against the biological action of C3 Nishiki, T., Kamata, Y., Kozaki, S., Narumiya, S. and Sakagu- exoenzyme. chi, G. 1990. Immunochemical identification of the ADP-ribo- Recently the sites in the C3 molecule to which RhoA and syltransferase in botulinum C1 neurotoxin as C3 exoenzyme- NAD bind were reported [5, 22]. The RhoA-binding site like molecule. J. Biochem. 107: 769–775. was far from that of NAD under the first dimensional struc- 11. Moriishi, K., Syuto, B., Yokozawa, N., Oguma, K. and Saito, M. 1991. Purification and characterization of ADP-ribosyl- ture, but both sides were adjacent on the three-dimensional transferase (exoenzyme C3) of Clostridium botulinum type C [5]. The author proposes that the C302 epitope is the site to and D strains. J. Bacteriol. 173: 6025–6029. which RhoA and NAD bind, and the binding of C302 causes 12. Narumiya, S. 1996. The small GTPase Rho: cellular functions a greater depression of the activity described in this report. and signal transduction. J. Biochem. 120: 215–228. Now we are preparing a C3 exoenzyme and C302 complex 13. Narumiya, S., Sekine, A. and Fujiwara, M. 1988. Substrate for as a sample for crystallographic analysis to verify our work- botulinum ADP-ribosyltransferase, Gb, has an amino acid ing hypothesis. As C3 exoenzyme was a very powerful tool sequence homologous to a putative rho gene product. J. Biol. for analyzing cell functions through the GTP-binding pro- Chem. 263: 17255–17257. tein; therefore, elucidation of C3 exoenzyme molecule itself 14. Nemoto, Y., Namba, T., Kozaki, S. and Narumiya, S. 1991. and the use of the neutralizing mAb will activate the GTP- Clostridium botulinum C3 ADP-ribosyltransferase gene. J. Biol. Chem. 266: 19312–19319. binding protein research further. Furthermore, Maehara et 15. Nishiki, T., Narumiya, S., Morii, N., Yamamoto, M., Fujiwara, al. reported the presence of a C3 exoenzyme-like-ADP- M., Kamata, Y., Sakaguchi, G. and Kozaki, S. 1990. ADP- ribosyltransferase in bovine brain [9]. C302 recognized ribosylation of the rho/rac proteins induces growth inhibition, some proteins originating from the cytosol fraction of neurite outgrowth and acetylcholine esterase in cultured PC-12 bovine and mouse brains on western blotting analysis cells. Biochem. Biophys. Res. Commun. 167: 265–272. (results will be published). C302 will help to solve the role 16. Ohashi, Y., Kamiya, T., Fujiwara, M. and Narumiya, S. 1987. of the endogenous C3 exoenzyme-like enzyme. ADP-ribosylation by type C1 and D botulinum neurotoxins: stimulation by guanine nucleotides and inhibition by guani- REFERENCES dino-containing compounds. Biochem. Biophys. Res. Commun. 142: 1032–1038. 1. Aktories, K., Braun, U., Rösener, S., Just, J. and Hall, A. 1989. 17. Ohashi, Y. and Narumiya, S. 1987. ADP-ribosyltransferase of The rho gene product expressed in E. coli is a substrate of bot- a Mr 21,000 membrane protein by type D botulinum toxin. J. Biol. Chem. 262: 1430–1433. NOVEL MAB TO BOTULINUM C3 EXOENZYME 771

18. Schägger, H. and von Jagow, G. 1987. Tricine-sodium dode- lation of small GTP-binding proteins catalyzed by Clostridium cylsulfate-polyacrylamide gel electrophoresis for the separa- botulinum ADP-ribosyltransferase C3. FEBS Lett. 324: 353– tion of proteins in the range from 1 to 100 kDa. Anal. Biochem. 357. 166: 368–379. 21. Watanabe, Y., Morimatsu, M. and Syuto, B. 2000. The evalua- 19. Sekine, A., Fujiwara, M. and Narumiya, S. 1989. Asparagine tion of the potential of botulinum C3 enzyme as an exogenous residue in the rho gene product is the modification site for bot- differentiation inducing factor to neurons. J. Vet. Med. Sci. 62: ulinum ADP-ribosyltransferase. J. Biol. Chem. 264: 8602– 473–478. 8605. 22. Wilde, C., Genth, H., Aktories, K. and Just, I. 2000. Recogni- 20. Toratani, S., Sekine, N., Katada, T. and Yokosawa, H. 1993. tion of RhoA by Clostridium botulinum C3 exoenzyme. J. Biol. Production of monoclonal antibodies that inhibit ADP-ribosy- Chem. 275: 16478–16483.