Idiotypic mimicry of a catalytic antibody active site

Molecular Immunology 39 (2002) 273–288

Idiotypic mimicry of a catalytic antibody active site Glynis Johnson∗ , Samuel W. Moore Departments of Paediatric Surgery and Medical Biochemistry, University of Stellenbosch, P.O. Box 19063, Tygerberg 7505, South Africa Received 1 January 2002; received in revised form 2 February 2002; accepted 23 April 2002

Abstract We have previously described three catalytic antibodies (Ab1s) raised against human erythrocyte acetylcholinesterase (AChE). These antibodies both recognise and resemble AChE in their reaction with substrates and appear with a relatively high frequency. We do not know, however, why catalytic activity should have developed in response to a ground state antigen. This question has implication for autoimmune disorders, which are frequently characterised by the presence of catalytic antibodies, many of which have cytotoxic effects. In this study, we raised anti-idiotypic (Ab2) and anti-anti-idiotypic (Ab3) antibodies to a catalytic Ab1 and examined their properties. None of the Ab2s showed catalytic activity, whereas four of the Ab3s did, an incidence of 1.26%. No contamination of antibody preparations with either AChE or butyrylcholinesterase (BChE) was found. Immunisation of mice with AChE, as well as AChE complexed with various inhibitors, resulted in a significant increase in catalytic immunoglobulins in the serum, compared with non-immunised mice and mice immunised with the Ab1. There appears to be considerable resemblance between Ab1s and Ab3s, but there are also significant differences between the two groups. All the antibodies were inhibited by phenylmethylsulphonyl fluoride (PMSF), indicating the presence of a serine residue in their active sites and were inhibited by the cholinesterase active site inhibitors tetraisopropyl pyrophosphoramide (iso-OMPA) and pyridostigmine. The Ab3s resembled the Ab1s in their ability to hydrolyse both acetylthiocholine (ATCh) and butyrylthiocholine (BTCh). However, the Ab3s appear to be better catalysts, having significantly reduced KM values (for ATCh but not BTCh) and increased turnover numbers (Kcat ), rate enhancements (Kcat /Kuncat ) and Kcat /KM ratios. The Ab3s also had reduced affinities for cholinesterase anionic site inhibitors (edrophonium, tetramethylammonium and BW284c51) and no affinity at all for the AChE peripheral anionic site (PAS) inhibitor fasciculin. All the antibodies recognise, to some degree, the PAS of AChE, shown by their ability to inhibit AChE, to compete with peripheral site inhibitors and to block AChE-mediated cell adhesion, a property of the site. These results indicate idiotypic mimicry of the catalytic antibody’s active site, suggesting that the catalytic activity is due to affinity maturation of immunoglobulin genes in response to a specific antigen, namely, the PAS of AChE. Further studies are required to determine the structural features of this ground state antigen responsible for the development of catalytic activity. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Catalytic antibody; Cholinesterase; Acetylcholinesterase; Anti-idiotypic; Peripheral anionic site

1. Introduction Catalytic antibodies are immunoglobulins that display an enzyme-like ability to catalyse chemical reactions. A characteristic of all catalysts (enzyme or antibody) is an affinity for their substrates’ transition state, the transient high energy conformation assumed by the substrate on binding to the catalyst. Many catalytic antibodies have been successfully raised by using transition state analogues as immunogens (Benkovic, 1992). Ground state antigens have also been found to elicit catalytic antibodies (Paul, 1994), but the mechanism by which they do this remains obscure. Catalytic antibody levels are increased in many autoimmune diseases (Schourov et al., 1995; Paul et al., 1997) and ∗

Corresponding author. Tel.: +27-21-938-9422; fax: +27-21-933-7999. E-mail address: [email protected] (G. Johnson).

the antibodies have been shown to have cytotoxic effects in vitro (Kozyr et al., 2000). Acetylcholinesterase (AChE; EC 3.1.1.7) is the serine hydrolase responsible for the breakdown of acetylcholine in the synapse and neuromuscular junction. It also has secondary, perhaps evolutionarily more ancient, functions associated with cell adhesion, aggregation and differentiation. These functions, which include cell adhesion (Johnson and Moore, 1999), the promotion of neurite outgrowth in primitive neural cells (Munoz et al., 1999) and the interaction with the amyloid ␤-peptide (Inestrosa et al., 1996), are mediated by the peripheral anionic site (PAS). The PAS is a charged, largely aromatic structure (Y70, D72, Y121, W279, Y334; Torpedo numbering) situated at the opening to the active site gorge of AChE. It binds, but does not hydrolyse, substrate, as well as specific inhibitors, and allosterically modulates catalysis at the active site. The latter (S200, E329, H440;

0161-5890/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 1 6 1 - 5 8 9 0 ( 0 2 ) 0 0 1 1 3 - X

274

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

Torpedo numbering) lies at the bottom of the gorge, 20A from the surface of the molecule (Sussman et al., 1991). Also within the gorge lies a second anionic site (W84, E199, Y330; Torpedo numbering) that is responsible for the binding and positioning of the choline moiety of the substrate. Butyrylcholinesterase (BChE) is a related enzyme, with, on average, 50% homology to AChE. It is thought to act as a scavenger, protecting acetylcholine-binding proteins from the effects of inhibitors (Schwarz et al., 1995). It is less specific than AChE; while AChE is only capable of hydrolysing acetylcholine, BChE can accommodate the larger members of the choline ester series (butyrylcholine, propinylcholine). This is a consequence of the structure of the acyl pocket; in AChE, this is composed of two Phe residues (F288 and F290) that protrude into the gorge (Ordentlich et al., 1993), restricting access. In BChE, these Phe residues are replaced by Leu, Ile or Val (Harel et al., 1993). BChE also does not have a PAS. Several years ago (Johnson and Moore, 1995), we accidentally raised a catalytic antibody (MAb E8) to AChE. This antibody was unusual in that it both recognised and resembled AChE, with relatively weak acetylcholine-hydrolysing ability (Kcat = 27.45 ± 5.1 s−1 ). No evidence of contamination of antibody preparations with either AChE or BChE was found. We subsequently raised two more such catalytic antibodies against the same antigen and characterisation (Johnson and Moore, 2000) showed them to have similar esterase activity and active sites with features of both AChE and BChE. Friboulet and co-workers (Izadyar et al., 1993; Kolesnikov et al., 2000) have raised similar cholinesterase-like catalytic antibodies, but as anti-idiotypic antibodies to a monoclonal antibody directed, like ours, against human erythrocyte AChE. An important question is that of how the antibodies developed catalytic activity. It is possible that the activity is innate, a property of minimally mutated V genes. Such activity has been observed in non-immune polyclonal immunoglobulins (Erhan and Greller, 1974) and in the L-chains of Bence Jones proteins (Gao et al., 1995). On the other hand, the activity may be the result of affinity maturation of V genes in response to a specific immunogen. AChE is, of course, not a transition state analogue and the question of how catalytic antibodies may develop from a ground state antigen arises. The relatively high frequency with which these AChE-like catalytic antibodies appear and the fact that two laboratories have, independently, raised such antibodies (albeit one as Ab1s and the other as Ab2s), to the same original antigen (human erythrocyte AChE), suggests that the occurrence of these antibodies is not due to chance alone and may be the result of structural peculiarities of AChE. A further, intriguing, speculation is the possibility of a link between anti-AChE antibodies and the anti-thyroglobulin autoantibodies observed in autoimmune thyroid disease; AChE is homologous to the C-terminal region of thyroglobulin (Schumacher et al., 1986) and cross-reaction of the antibodies has been observed (Mappouras et al., 1995). Interestingly,

many of these anti-thyroglobulin autoantibodies are catalytic and both recognise and hydrolyse thyroglobulin (Rose and Burek, 2000). Anti-AChE autoantibodies have been described in a number of autoimmune disorders (Conradi and Ronnevi, 1994; Haggstrom et al., 1997). It is not known whether any of these antibodies are catalytic, nor what their pathological role might be. We have previously observed (Johnson and Moore, 1999) that many anti-AChE antibodies (and especially those with catalytic activity) adversely affect neural cell adhesion in vitro, resulting in apoptosis. In this study, we have addressed the questions of whether the catalytic activity might be a chance occurrence, whether idiotypic mimicry of such activity occurs and its effects on cells in vitro and the part of the AChE molecule with which the antibodies react and how this might account for the development of catalytic activity.

2. Experimental 2.1. Materials The following reagents and chemicals were obtained from Sigma: human erythrocyte AChE (cholinesterase, acetyl, C5400), Torpedo AChE (C2888), recombinant human AChE (C1682), human BChE (cholinesterase, butyryl, C9971), acetylthiocholine (ATCh), butyrylthiocholine (BTCh), pyridostigmine, tetraisopropyl pyrophosphoramide (iso-OMPA), tetramethylammonium chloride, ethopropazine (10-[2-diethylaminopropyl]phenothiazine), gallamine triethiodide, fasciculin II, phenylmethylsulphonyl fluoride (PMSF), tetracaine (4-[butylamino]benzoic acid-2-[dimethyl amino ethyl ester]), propidium iodide, BW284c51 (1,5bis-[4-allyldimethyl ammoniumphenyl]pentan-3-one dibromide), edrophonium chloride (ethyl[m-hydroxyphenyl] dimethylammonium chloride), Freund’s complete and incomplete adjuvants, Protein-A-Sepharose 4B, serum- and protein-free hybridoma medium, mouse IgG and IgM, biotinylated anti-mouse IgG and IgM, streptavidin peroxidase and 5,5 -dithiobis[2-nitrobenzoic acid] (DNTB; Ellman’s reagent). Tissue culture medium Dulbecco’s MEM and RPMI) and fetal bovine serum were obtained from Delta. 2,2 -azinobis(3-ethylbenzthiazoline sulphonic acid (ABTS) was obtained from Boehringer, Mannheim. 2.2. Methods 2.2.1. Immunisation of mice BALB/c mice were injected with purified MAb E8. The initial immunisation was subcutaneous, in Freund’s complete adjuvant, and subsequent immunisations (at 3-week intervals) were intraperitoneal, in incomplete adjuvant. MAb E8 was purified from ascites fluid by standard protocols (Harlow and Lane, 1988) on Protein-A-Sepharose 4B, using low salt buffers, and appeared homogeneous on SDS-PAGE. Mice used for the generation of Ab2s (anti-idiotypic

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

antibodies) were sacrificed after 9–10 weeks and those for Ab3s, after >6 months. The latter relatively long period of immunisation allows for the development of anti-idiotypic antibodies to the immunised antigen (Shoenfeld, 1994). The care and handling of animals was in accordance with guidelines laid down by the Animal Research Review Committee of the University of Stellenbosch. In a separate series of experiments, mice (three in each group) were immunised with 90 ␮g human erythrocyte AChE complexed with the inhibitors fasciculin, tetramethylammonium and BW284c51, respectively, in stoichiometric amounts. The AChE used was from the same source as that used for the generation of Ab1s (Johnson and Moore, 2000), appeared homogeneous on SDS-PAGE and was used without further purification. The inhibitors were incubated with AChE for 2 h at room temperature and the complex passed through Sephadex G-25 to remove unbound inhibitor. AChE-containing fractions were identified by A280 spectrophotometry, diluted in phosphate-buffered saline to the required volume, mixed with adjuvant and used immediately. One control group was immunised with AChE only, the other remained unimmunised. Booster immunisations (in incomplete adjuvant) were at 3-week intervals and blood samples were taken 4 days after each immunisation. The possible presence of catalytic immunoglobulins was tested as follows: microtitre plates were coated with either anti-mouse IgM or IgG and incubated with serum. After blocking (1% bovine serum albumin in phosphate-buffered saline), wells were probed with Ellman’s reagent (see Section 2.2.3) for cholinesterase activity, using either ATCh or BTCh as substrate. 2.2.2. Culture of hybridomas Spleen lymphocytes were fused with mouse myeloma cells of the non-secreting SP2/0-Ag14 line, according to standard protocols (Harlow and Lane, 1988). Hybridomas were screened by ELISA for the production of IgM and IgG and anti-AChE antibodies (see Section 2.2.7) and for cholinesterase activity by the Ellman assay (see Section 2.2.3), using culture medium as control. Positive hybridomas were cloned twice by limiting dilution onto mouse spleen cell feeder layers and the cultures expanded. The last two passages before harvesting of antibody were conducted in serum- and protein-free tissue culture medium. Cells were washed twice with phosphate-buffered saline at subculture. IgG was purified on Protein-A-Sepharose 4B, using low salt buffers (Harlow and Lane, 1988). No significant IgM antibodies were found. The IgG-containing fractions were identified and quantitated by measuring their absorbance at 280 nm and ELISA using biotinylated anti-mouse IgG as probe. 2.2.3. Measurement of cholinesterase activity Cholinesterase activity was determined by the method of Ellman et al. (1961), using either 0.5 mM ATCh (AChE

275

and BChE substrate) or BTCh (BChE substrate) and 1 mM DTNB (Ellman’s reagent) in a microassay system (Doctor et al., 1987). Production of the yellow nitrobenzoate anion was measured at 405 nm and the reaction rate data calculated using its extinction coefficient of 1.36 × 104 M−1 cm−1 . 2.2.4. Kinetic measurements Monoclonal antibody or enzyme (7.9 × 10−7 M) was incubated in buffer (50 mM sodium phosphate, pH 7.4) containing 1 mM DTNB (Ellman’s reagent), at 25 ◦ C. Temperature and pH were monitored to ensure that no variation occurred. Reactions were initiated by the addition of varying amounts of substrate (ATCh or BTCh) to give 29–7500 ␮M final concentration. Absorbance was monitored spectrophotometrically at 405 nm. The first order rate constant (spontaneous hydrolysis), measured in the absence of antibody or enzyme, was used to correct the initial rate data. Michaelis constants (KM ) were determined from Lineweaver–Burk plots, as was Vmax for the calculation of the turnover number (Kcat ). 2.2.5. Reaction of antibodies with inhibitors The following inhibitors were used (the type, sensitive enzyme, site of action and inhibitor concentration range are given in parenthesis): iso-OMPA (organophosphate, BChE active site, 62.5 ␮M to 2 mM); pyridostigmine (carbamate, AChE and BChE active site, 3–192 ␮M); edrophonium (quaternary, AChE and BChE anionic site, 3–192 ␮M); tetramethylammonium (quaternary, AChE and BChE anionic sites, 0.3–20 mM); BW283c51 (bisquaternary, AChE peripheral site, 3–192 ␮M); propidium (cationic, AChE peripheral site, 3–192 ␮M); fasciculin II (peptide, AChE peripheral site, 0.006–0.2 nM) and PMSF (serine inhibitor, 31.25 ␮M to 1 mM). Inhibitors were incubated with 175 nM enzyme or antibody for 1 h at 25 ◦ C. Residual activity was measured by the Ellman assay. Inhibition constants (Ki ) were determined by Dixon plots (1/v versus [I]; Dixon, 1953) at three substrate concentrations. Ki values were determined by the intercept of graphs (competitive inhibition) and the negative intercept on the x-axis (non-competitive inhibition). 2.2.6. Sucrose density gradient sedimentation Samples of Protein-A-purified IgG, AChE and mixtures of the two were applied to 5–20% (w/v) sucrose density gradients in 50 mM Tris–HCl, pH 7.4, containing 0.1% Triton and 1 M NaCl. Gradients were centrifuged at 120,000 × g for 24 h at 4 ◦ C (Beckman SW 65 rotor) and fractionated by upward displacement. Gradients were calibrated by the inclusion of markers of known sedimentation values (bovine serum albumin 4.65S; fibrinogen 7.63S and catalase 11.2S) and also measured by direct densitometry. Fractions were scanned at 280 nm and assayed for IgG, AChE and BChE by ELISA (see Section 2.2.7) and for cholinesterase activity by the Ellman assay.

276

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

2.2.7. Immunoassays Immunoassays were conducted by the ELISA sandwich method, using biotinylated anti-mouse IgG and IgM for the IgG and IgM assays, three anti-AChE antibodies (one polyclonal and two monoclonals) for detecting AChE and two anti-BChE antibodies (one polyclonal and one monoclonal) in the case of the BChE assay. These anti-AChE antibodies were raised in our laboratory and the anti-BChE antibodies were gifts from A.S. Balasubramanian and S. Brimijoin. The blocking solution was 5% bovine serum albumin in phosphate-buffered saline, which was also used for dilution of reagents. The washing solution was 0.1% Triton X-100 in phosphate-buffered saline. For the screening of the anti-idiotypic antibodies, antibody E8, mouse IgG and IgM were used as plate coatings.

was measured by the Ellman assay, using either ATCh (substrate for AChE and antibodies) or BTCh (substrate for antibodies only). Controls used were non-specific mouse IgG, MAb E413D8 (which does not recognise the PAS; Saxena et al., 1998) and MAb E62A1 (which partially recognises the PAS). These MAbs were a gift from M. Gentry. 2.2.10. Statistics Statistical determinations were carried out using Student’s t-test. 2.3. Results 2.3.1. The antibodies Ab1: These antibodies were raised against human erythrocyte AChE and have been described previously (Johnson and Moore, 1995, 2000). Three catalytic antibodies (E8, C2 and 43B4C) and one non-catalytic antibody as control are included in this group (Table 1). These antibodies recognise AChE, but not the Ab1 used as antigen for the anti-idiotypic antibodies, nor mouse IgG (Fig. 1a). Ab2: The seven antibodies included in this classification were raised against MAb E8 with a short (10 weeks) immunisation period. None of these antibodies exhibited catalytic behaviour. The antibodies recognise MAb E8 and mouse IgG, but show little reaction with AChE (Fig. 1b). Ab3: Included in this group are six antibodies, four of which are catalytic (MAbs 13B9C, 13B9F, 12D11F and 13A). They were the result of immunisation with MAb E8 for a relatively long period of time (>6 months) that allows for the development of anti-idiotypic antibodies to the immunised antigen. Fig. 1c shows that, although they recognise mouse IgG and there is also some recognition of AChE itself, they do not interact with MAb E8.

2.2.8. Competition ELISA The 96-well microtitre plates were coated with 143 nM AChE in 50 mM sodium carbonate, pH 9.25 (calculation based on a 70 kDa AChE monomer, containing one active site). The cholinergic inhibitors, propidium iodide, BW284c51 and edrophonium, were used in a range of concentrations between 0 and 400 ␮M, and incubated with the AChE for 2 h at room temperature. After removal of these inhibitor solutions, MAbs were added at a concentration of 560 nM (4× AChE active site concentration) and incubated for 2.5 h, also at room temperature. The amount of MAb bound was probed with biotinylated anti-mouse IgG and streptavidin peroxidase. The blocking and washing solutions used were those described in Section 2.2.7. 2.2.9. Binding of antibodies to AChE and determination of inhibition of activity Antibody (1.5 ␮g) were incubated overnight at 4 ◦ C with varying concentrations of AChE (0–6 ␮g). Residual activity Table 1 Description of the antibodies Ab classification

MAb

Raised against

Specificity

Catalytic

Ab1 (idiotypic)

E8 C2 43B4C Control

AChE AChE AChE AChE

AChE AChE AChE AChE

Yes Yes Yes No

Ab2 (anti-idiotypic)

211E 17B 11C 210G 29H 18A 210H

MAb MAb MAb MAb MAb MAb MAb

E8, E8, E8, E8, E8, E8, E8,

No No No No No No No

Ab3 (anti-anti-idiotypic)

13B9C 13B9F 12D11F 13A Control

Ab2 Ab2 Ab2 Ab2 Ab2

E8 E8 E8 E8 E8 E8 E8

in in in in in

vivo vivo vivo vivo vivo

IgG IgG IgG IgG IgG IgG IgG

IgG, IgG, IgG, IgG, IgG,

AChE AChE AChE, E8 AChE, E8 AChE

Yes Yes Yes Yes No

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

2.3.2. Frequency of appearance of catalytic antibodies Ab1: MAb E8 was derived from a fusion done in 1994 and has been previously described (Johnson and Moore, 1995). C2 and 43B4C were derived from three fusions, which yielded 29 IgG-secreting hybridoma lines. Of these, two showed catalytic activity, giving a frequency of 6.9%. These fusions were not screened at the pre-cloning stage,

277

nor were clones screened for catalytic activity. If this had been done, before clones were either lost or selected for, the frequency of catalytic antibody-secreting clones is likely to have been much less. Ab2: Two fusions yielded 46/96 (47.9%) IgG-positive wells (screened 11 days post-fusion). Two of these had borderline catalytic activity (measured on tissue culture

Fig. 1. Recognition of AChE, MAb E8 and mouse IgG by antibodies: (a) Ab1s; (b) Ab2s; (c) Ab3s. Microtitre plates were coated with 2 ␮g ml−1 (100 ng per well) recombinant human AChE (column 1), MAb E8 (column 2) or mouse IgG (column 3), respectively. Antibodies were added at 375 ng per well. The amount of antibody bound was probed with biotinylated anti-mouse IgG, streptavidin peroxidase and H2 O2 /ABTS colour reagent. Values are given as A405. The control Ab in (a) is a non-catalytic antibody from the same fusion as MAb C2 and the control Ab in (c) is a non-catalytic antibody from the same fusion as MAbs 13B9C and 13B9F. Background values (in the absence of MAb) were 0.120 ± 0.003 (a), 0.115 ± 0.002 (b) and 0.106 ± 0.002 (c); n = 8 in all cases.

278

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

Fig. 1. (Continued ).

supernatants, which contain AChE, against suitable controls). These two fusions resulted in 143 viable clones, none of which produced catalytic IgG. Ab3: Two fusions resulted in 87/96 (90.6%) wells positive for IgG. Nine of these culture supernatants reacted with AChE also. Of these 318 clones produced, 4 showed catalytic activity, an incidence of 1.26%. 2.3.3. Catalytic IgG in mouse sera The cholinesterase-like activity associated with immunoglobulins (after 9-week immunisation) is shown in Fig. 2. Analysis of the results indicates that mice immunised with AChE (without addition of inhibitors) showed a significantly higher incidence of catalytic IgG than non-immunised mice (P < 0.01). No significant difference was observed between AChE and (AChE + BW284c51)-immunised mice; the incidence of catalytic activity for mice immunised with AChE + fasciculin and AChE + tetramethylammonium was in both cases significantly less than for those with AChE alone, with this difference being less marked in the tetramethylammonium group (P < 0.01 for fasciculin; P < 0.05 for tetramethylammonium). No significant difference was found between mice immunised with MAb E8 and the non-immunised group. 2.3.4. Demonstration of antibody purity Our major concern was that the antibody preparations had become contaminated with either AChE or BChE. Demonstration of antibody purity on polyacrylamide gels would be of little value in this case, as the denaturing conditions could cause the breaking of any antibody–AChE (or BChE)

complex. Furthermore, the AChE dimer is located in a very similar position to IgG on SDS-PAGE. It was, therefore, decided to use sucrose density gradients (5–20%) which would be capable of separating IgG itself from any IgG complexed with AChE or BChE. Individual gradients containing antibody IgG, with and without AChE; irrelevant IgG, with and without AChE, and AChE itself, were run. The IgG fraction sedimented in a single peak at approximately 7S (Fig. 3a), identified by A280 and the anti-IgG assay. The anti-IgG antibody does not cross-react with either AChE or BChE (data not shown). This peak was not recognised as AChE by three anti-AChE antibodies (Fig. 3a) nor as BChE by two anti-BChE antibodies (data not shown). In the case of the antibodies showing catalytic activity, the peak was recognised by the Ellman assay, using ATCh as substrate. In contrast, when AChE was added to the antibody sample (Fig. 3b), IgG immunoreactivity is present, again in the smaller 7S peak, with the bulk of activity shifting to a position at approximately 11S. This latter peak was recognised by the AChE immunoassay. The first peak presumably represents uncomplexed IgG and the second, the AChE–IgG complex. A further, smaller peak is evident at higher density: this may be a complex of two AChE molecules with IgG. Note that for the sake of brevity only one antibody (13B9F) is shown here; no significant differences were observed in the sedimentation behaviour of any of the antibodies. As a further precaution, spent medium from the hybridoma cultures was run on gradients. Although the IgG fraction at 7S was apparent, no evidence of either AChE or BChE was found (data not shown).

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

279

Fig. 2. Cholinesterase-like catalytic activity associated with IgG from mouse sera. Mice (n = 3 for each group) were immunised with AChE (“AChE”), AChE + fasciculin (“+FAS”), AChE + tetramethylammonium (“+TMA”), AChE + BW284c51 (“+BW”), MAb E8 (“E8”) or received no immunisation (“Non-Imm.”). IgG from serum was immobilised on microtitre plates coated with anti-mouse IgG and the cholinesterase activity measured with the Ellman assay, using ATCh as substrate.

2.3.5. Kinetic parameters of antibodies Kinetic parameters (KM , Kcat , Kcat /Kuncat and Kcat /KM ) for the Ab3s, together with the previously described Ab1s, are shown in Tables 2 and 3. Table 2 shows parameters using ATCh (hydrolysed by both AChE and BChE) as substrate and Table 3 gives those using BTCh (BChE substrate only). It can be seen that all the antibodies are able to hydrolyse both substrates, with few significant differences. The only

exceptions are the Kcat values for the Ab1s, which are significantly (P < 0.01) higher for BTCh. Differences in kinetic parameters are, however, observed between the Ab1s and Ab3s. The Ab3s have significantly lower KM values (P < 0.01), using ATCh, but not BTCh, as substrate. Turnover numbers (Kcat ) for both substrates are significantly higher (P < 0.01), as are the rate enhancements (Kcat /Kuncat ) and Kcat /KM ratios.

Table 2 Kinetic parameters for catalytic antibodies (Ab1s and Ab3s) and Torpedo AChE using ATCh as substrate Catalyst

KM ∗∗ (mM)

Kcat ∗ (s−1 )

Kcat /Kuncat ∗ (×108 )

E8 C2 43B4C

7.69 ± 0.40 4.55 ± 0.98 4.08 ± 0.42

27.40 ± 5.51 23.42 ± 2.68 20.21 ± 0.98

1.61 ± 0.32 1.38 ± 0.15 1.19 ± 0.06

13B9C 13B9F 12D11F 13A

0.63 0.80 0.83 2.56

AChE Control Ab

0.14 0

± ± ± ±

0.07 0.21 0.19 0.18

69.62 85.56 38.98 25.06 10000 0

± ± ± ±

4.21 7.29 3.57 1.29

3.98 4.89 2.23 1.43

± ± ± ±

588.2 0

0.42 0.81 0.31 0.09

Kcat /KM ∗ (mM−1 s−1 ) 3.74 ± 0.94 6.28 ± 2.59 4.97 ± 0.66 110.51 106.95 40.94 9.79

± ± ± ±

5.45 7.89 3.27 1.25

71428.5 0

Values for KM (Michaelis constant), Kcat (turnover number), Kcat /Kuncat (rate enhancement) and the Kcat /KM ratio are shown as mean ± S.D. (n = 3). The value for Kuncat (the reaction in the absence of catalyst) was determined as 1.74 × 10−7 s−1 . The control Ab is a non-catalytic antibody from the same fusion as MAb C2. ∗ Significant differences between the Ab1s and the Ab3s (P < 0.05). ∗∗ Significant differences between the Ab1s and the Ab3s (P < 0.01).

280

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

Fig. 3. Density gradient sedimentation of MAb 13B9F with and without the addition of AChE: (a) 35 ␮g MAb 13B9F alone; (b) 35 ␮g MAb 13B9F with addition of 100 ␮g human erythrocyte AChE. Gradients were fractionated by upward displacement, hence the lower density fractions are on the LHS of the graph. The positions of the sedimentation markers bovine serum albumin (4.5S), fibrinogen (7.63S) and catalase (11.2S) are at fractions 4, 7 and 14, respectively. Fractions were assayed for mouse IgG, AChE and BChE (not shown) by ELISA, and for cholinesterase activity by the Ellman assay, using ATCh as substrate. IgG (䉬); Ellman assay ( ); AChE immunoassay ( ).

2.3.6. Reaction of the antibodies with inhibitors Fig. 4 shows the effects of various cholinesterase inhibitors, including PMSF, on the antibodies. All the antibodies are inhibited by PMSF, indicating the presence of a Ser residue in their active sites. Pyridostigmine, a carbamate interacting with the active sites of both AChE and BChE, affects Ab1s and Ab3s almost identically, as shown also by the inhibition constants (Table 4). On the other hand, iso-OMPA,

a large organophosphate specific for the active site of BChE, is more effective at inhibiting the Ab1s (P < 0.01). Dixon plots (not shown) indicate competitive inhibition for both pyridostigmine and iso-OMPA, for both sets of antibodies. None of the antibodies is affected by the AChE PAS inhibitor, fasciculin. The Ab1s are partially inactivated, non-competitively, by the cholinesterase anionic site inhibitors BW284c51 and tetramethylammonium, suggesting

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

281

Table 3 Kinetic parameters for catalytic antibodies (Ab1s and Ab3s) and Torpedo AChE, using BTCh as substrate Catalyst

KM (mM)

Kcat ∗∗ (s−1 )

Kcat /Kuncat ∗∗ (×108 )

E8 C2 43B4C

1.92 ± 0.35 5.00 ± 1.27 16.71 ± 3.29

52.74 ± 2.69 62.03 ± 3.49 89.03 ± 5.25

1.51 ± 0.99 1.77 ± 0.40 2.54 ± 0.82

13B9C 13B9F 12D11F 13A

0.38 0.45 0.76 0.17

BChE Control Ab

0.01 0

± ± ± ±

0.12 0.05 0.07 0.02

33.92 28.10 43.10 26.33

± ± ± ±

1.35 3.42 2.59 4.27

0.92 0.76 1.18 0.72

1335.0 0

± ± ± ±

Kcat /KM ∗∗ (mM−1 s−1 ) 27.47 ± 2.10 12.41 ± 0.81 5.53 ± 0.72

0.09 0.19 0.23 0.21

89.26 62.44 57.63 154.88

381.4 0

± ± ± ±

5.37 7.28 4.29 10.95

133500 0

Values for KM (Michaelis constant), Kcat (turnover number), Kcat /Kuncat (rate enhancement) and the Kcat /KM ratio are shown as mean ± S.D. (n = 3). The value for Kuncat (the reaction in the absence of catalyst) was determined as 3.50 × 10−7 s−1 . The control Ab is a non-catalytic antibody from the same fusion as MAb C2. ∗∗ Significant differences between the Ab1s and the Ab3s (P < 0.01).

Table 4 Inhibition constants for catalytic antibodies (AChE and BChE) with the active site inhibitors (pyridostigmine and iso-OMPA) and the anionic site inhibitor (tetramethylammonium) Catalyst

Pyridostigmine (␮M)

iso-OMPAa (␮M)

Tetramethylammonium (␮M)

E8 C2 43B4C

47.1 ± 2.7 36.0 ± 4.3 19.2 ± 2.1

129.3 ± 10.2 65.5 ± 5.7 80.4 ± 6.7

17.1 ± 1.3 48.5 ± 2.7 29.2 ± 3.0

13B9C 13B9F 12D11F 13A

28.5 35.5 18.3 38.9

± ± ± ±

2.2 5.1 4.9 4.5

270.4 403.6 324.6 193.2

8.0 ± 2.9 7.0 ± 2.3

AChE BChE

± ± ± ±

12.9 20.3 27.5 6.9

457.1 ± 10.3 0.2 ± 0.05

ND ND ND ND 61.3 ± 2.2 57.4 ± 4.9

Ki values were determined from Dixon plots; they are shown as mean ± S.D., n = 3 in all cases. ND: not done. a Denotes significant differences between Ab1s and Ab3s.

the presence of an anionic site-like structure, separate from the active site. The Ab3s, show less inhibition by these compounds, significantly so (P < 0.01) in the case of tetramethylammonium.

2.3.7. Inhibition of AChE by antibodies and inhibition of antibodies by AChE The effects of the antibodies on the activity of AChE is shown in Table 5. The substrate used here was ATCh

Table 5 Inhibition of activity by AChE-antibody binding, using ATCh as substrate Ab

Ab only

Ab + AChE (initial reading)

E8 C2 43B4C

3.61 ± 0.88 1.49 ± 0.26 2.76 ± 0.36

13.93 ± 1.19 11.81 ± 0.73 13.08 ± 0.49

13B9C 13B9F 12D11F 13A

2.50 3.15 2.57 3.59

E413D8 E62A1

– –

± ± ± ±

0.05 0.27 0.04 0.52

12.82 13.48 12.90 13.91

± ± ± ±

1.13 1.01 0.38 1.98

10.40 ± 0.09 10.46 ± 0.31

Ab + AChE (after incubation) 8.36 ± 0.86 5.41 ± 0.57 9.30 ± 0.60 9.04 9.70 9.12 10.13

± ± ± ±

0.70 0.76 1.21 0.64

10.35 ± 0.41 7.00 ± 0.23

An amount of 750 ng antibody was added to 12 ng human recombinant AChE in a total volume of 100 ␮l buffer containing 1 mM DTNB (Ellman’s reagent) and 0.5 mM ATCh were added, and the absorbance at 405 nm measured immediately (“initial reading”). Duplicate samples of AChE and antibodies were incubated for 24 h at 4 ◦ C and the same procedure followed (“after incubation”). MAb E413D8 is a non-catalytic antibody that does not interact with the PAS; MAb E62A1 is also a non-catalytic antibody that partially recognises the PAS (Saxena et al., 1998). Results are in mIU ml−1 ; mean ± S.D., n = 4 for all.

282

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

because this is hydrolysed by both AChE and the antibodies, it is not possible to distinguish between the two sets of catalysts. Complexation of the antibodies with AChE results in a significant (P < 0.01) decrease in the activity of the complex; this is seen for all antibodies. MAb E62A1, which partially recognises the PAS, also results in a decrease in

AChE activity; MAb E413D8, which does not react with the PAS, does not. Table 6 shows the results of a parallel study, using BTCh as substrate. As BTCh is hydrolysed by the antibodies, but not by AChE, we are able to distinguish the respective activities. Significant inhibition (P < 0.01)

Fig. 4. Activity remaining after incubation of antibodies, AChE and BChE with inhibitors. (a) Antibodies: column 1 = mean Ab1 values; column 2 = mean Ab3 values; standard deviations are given in the text; (b) AChE; (c) BChE. Antibody (175 nM), AChE or BChE were incubated with the inhibitors pyridostigmine (192 ␮M; Pyrido), fasciculin (0.2 nM; FAS), iso-OMPA (2 mM; OMPA), tetramethylammonium (20 mM; TMA), edrophonium (192 ␮M; Edro), BW284c51 (192 ␮M; BW) and PMSF (1 mM; PMSF) for 1 h at 25 ◦ C and the residual activity measured on the Ellman assay. Values are given as this activity expressed as a percentage of the catalyst in the absence of inhibitor.

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

283

Fig. 4. (Continued ).

of activity is seen only in three antibodies: the Ab1s E8 and C2, and the Ab3, 13B9F. Comparing the results in Tables 5 and 6, we can, therefore, conclude that all the antibodies, both Ab1s and Ab3s, inhibit AChE to a certain extent. 2.3.8. Competition between antibodies and inhibitors It is apparent (Fig. 5) that all the antibodies compete with the PAS inhibitors propidium and BW284c51 (Fig. 5a and b, respectively) to varying extents, but not with the non-PAS inhibitor, edrophonium (Fig. 5c). Taking the amount of antibody bound in the absence of inhibitor as 100%, the Ab1s showed a mean percentage of 14.74 ± 8.58% (n = 3) bound

in the presence of the maximum dose (390 ␮M) propidium. For the Ab3s, in the presence of propidium, this figure was 21.22 ± 11.40% (n = 4). A similar trend was seen in the presence of the PAS inhibitor BW284c51, with a mean percentage bound of 24.42 ± 11.11 (n = 3) for the Ab1s and 39.96 ± 13.15% for the Ab3s. In contrast, in the presence of edrophonium, a mean of 80.66 ± 10.39% (n = 3) of the Ab1s bound to AChE and 75.42 ± 6.82% (n = 4) of the Ab3s. The values for MAb E62A1, which reacts partially with the PAS, were 43.38 ± 3.99% for propidium, 81.75 ± 2.75% for BW284c51 and 96.39 ± 4.01% for edrophonium. Values for MAb E413D8, which does not recognise the PAS, were

Table 6 Inhibition of activity by AChE-antibody binding, using BTCh as substrate Ab

Ab only

Ab + AChE (initial reading)

Ab + AChE (after incubation)

Inhibition (%)

E8 C2 43B4C

3.56 ± 0.91 1.47 ± 0.25 2.66 ± 0.17

3.65 ± 0.73 1.49 ± 0.19 2.69 ± 1.00

1.85 ± 0.09 0.88 ± 0.02 2.47 ± 0.61

48.25 40.00 7.18

13B9C 13B9F 12D11F 13A

2.52 3.26 2.42 3.49

± ± ± ±

4.37 41.22 0.31 0.00

E413D8 E62A1

– –

± ± ± ±

0.39 0.47 0.29 0.61

2.51 3.22 2.45 3.51

± ± ± ±

0.24 0.35 0.24 0.59

0.06 ± 0.01 0.07 ± 0.01

2.41 1.92 2.41 3.52

0.49 0.19 0.40 0.27

0.01 ± 0.00 0.04 ± 0.01

An amount of 750 ng antibody was added to 12 ng human recombinant AChE in a total volume of 100 ␮l buffer containing 1 mM DTNB (Ellman’s reagent) and 0.5 mM BTCh were added and the absorbance at 405 nm measured immediately (“initial reading”). Duplicate samples of AChE and antibodies were incubated for 24 h at 4 ◦ C and the same procedure followed (“after incubation”). MAb E413D8 is a non-catalytic antibody that does not interact with the PAS; MAb E62A1 is alsoa non-catalytic antibody that partially recognises the PAS (Saxena et al., 1998). Results are in mIU ml−1 ; mean ± S.D., n = 4 for all.

284

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

99.59±0.25, 96.17±2.79 and 99.80±0.04% for propidium, BW284c51 and edrophonium, respectively. 2.3.9. Effects of antibodies on cell adhesion All of the catalytic antibodies, when added directly to the culture medium of the neuroblastoma cells, inhibited cell-substrate adhesion significantly (Table 7). The results

in the table are expressed for the highest concentration of IgG (32 ␮g/ml); dose-dependent effects were observed at the lower concentrations. The cells remained rounded and in suspension; if the antibodies were added after the cells had been plated, the latter rounded up and detached from the surface. In contrast, neither MAb E413D8 nor the non-specific mouse IgG had any significant effect on the cells.

Fig. 5. Competition ELISA between antibodies and inhibitors: (a) competition between propidium and antibodies; (b) competition between BW284c51 and antibodies; (c) competition between edrophonium and antibodies. Inhibitors were incubated with 143 nM AChE for 2 h at 25 ◦ C and removed. Antibodies were added at concentrations of 560 nM, incubated for 2.5 h and the amount of antibody bound probed with biotinylated anti-mouse IgG. Ab1s (䉬); Ab3s ( ); MAb E62A1 ( ); MAb E413D8 ( ).

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

285

Fig. 5. (Continued ).

Table 7 The effects of antibodies on neuroblastoma cell adhesion in vitro Ab

Adherent cells (%) 3.6∗∗

Comment

E8 C2 43B4C

8.0 ± 6.9 ± 2.5∗∗ 12.8 ± 4.2∗∗

Ab1 Ab1 Ab1

13B9C 13B9F 12D11F 13A

51.4 15.0 63.4 9.2

5.7∗∗ 2.9∗∗ 5.1∗∗ 3.7∗∗

Ab3 Ab3 Ab3 Ab3

E62A1 E413D8 Mouse IgG

41.2 ± 8.5∗∗ 93.4 ± 2.1 95.2 ± 3.7

Control (no antibody)

95.3 ± 4.1

± ± ± ±

Partial reaction with PASa No reaction with PASa Non-specific mouse IgG

Values are given for the maximum dose of antibody (32 ␮g ml−1 ) after 4-day incubation. a Saxena et al. (1998). ∗∗ Denotes significantly different (P < 0.01) from control (no antibody).

3. Discussion As the catalytic activity of the antibodies resembles that of known enzymes, it is of great importance to demonstrate a lack of contamination of the antibody preparations. To demonstrate this, we have shown that the shift in sedimentation position of the IgG fractions on sucrose density gradients and the lack of recognition of these fractions by both anti-AChE and -BChE antibodies, indicating that the fractions are indeed IgG. In addition, the limit of detection on the immunoassays is approximately 10 ng enzyme;

specific activities of the possible contaminating cholinesterase species range from 0.061 IU mg−1 for fetal bovine serum AChE (De la Hoz et al., 1986) to 0.25–1.0 IU mg−1 for both human erythrocyte AChE and human BChE (Sigma). It is thus apparent that none of these cholinesterase species are capable of producing measurable enzyme activity (on the Ellman assay) below 10 ng. A potential source of AChE contamination in our experimental method is from the fetal bovine serum used in cell culture. We controlled against this by culturing the hybridomas for the last two passages before harvesting in serum-free medium and thoroughly washing the cells on subculture. Furthermore, fetal bovine serum AChE exists as a tetramer, sedimenting at 11S (not 7S as do the IgG fractions) and no reducing procedures were used in preparations which might have disrupted intersubunit disulphide bonds. In addition, sedimentation of serum-free spent medium showed no evidence of AChE or BChE. There is no evidence that either B cells or hybridoma cells express cholinesterases, nor is BChE present in bovine serum at all (Carroll et al., 1995). Catalytic IgG fractions react as neither AChE nor BChE as indicated by the kinetic and inhibition data. A mixture of contaminating AChE and BChE would also not be possible, as, at the very small amounts of enzyme that could account for the activity, the stoichiometric ratio of inhibitor to enzyme would be so great that complete inhibition would occur, which is not the case. The antibodies follow Michaelis–Menten kinetics, with no evidence of substrate inhibition, a feature of all AChE species (Rosenberry, 1975). The KM values are also significantly different from AChE and BChE; the Michaelis constant is a property of

286

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

the enzyme and is independent of enzyme concentration. Minor variations may occur with differences in temperature and pH; however, all our measurements were conducted under strictly controlled conditions and also the differences in KM are too great to be accounted for in this manner. There are both similarities and differences between the Ab1s and the Ab3s raised in this study. Within each group, however, the antibodies appear similar in kinetic behaviour and presumed active site structure, as deduced from their reactions with substrates and inhibitors. All the antibodies are inhibited by PMSF, indicating the presence of a Ser residue in their active sites. Furthermore, all the antibodies are inactivated by an organophosphate (iso-OMPA) and a carbamate (pyridostigmine); both inhibitor types interact directly with the active site Ser. Interestingly, the Ab3s show significantly less inhibition with iso-OMPA; a possible reason being that they may have a more constrained active site, similar to AChE, which is not inhibited by the organophosphate at all. The serine hydrolases contain Ser as the nucleophile, a neighbouring His as the proton acceptor and a nearby acidic residue (Glu in the case of AChE and Asp in other enzymes) that forms a transient salt bridge with the His (Dodson and Wlodawar, 1998). Resolution of the structure of an esteratic antibody (Buchbinder et al., 1998) showed a Ser–His dyad. It would thus appear that the antibodies contain Ser as the nucleophile, very likely His as the base, and a residue or residues taking the part of the acid. The Ab3s raised in this study appear to be better catalysts than the Ab1s. This is seen by their reduced KM values (for ATCh) and increased Kcat , Kcat /Kuncat and Kcat /KM . This suggests a refinement of the catalytic potential by the idiotypic network. Although Ab1s and Ab3s hydrolyse both ATCh and BTCh, indicating a relatively unconstrained active site (compared to AChE), there is a slight preference for ATCh in the Ab3s. This fits with the reduced inhibition by iso-OMPA, suggesting that the Ab3s may have less accessible active sites. Another difference between the Ab1s and Ab3s is the reduced affinity for cholinesterase anionic site inhibitors (edrophonium, tetramethylammonium and BW284c51) in the latter. We have included BW284c51 in this category even though it is an AChE PAS inhibitor. Structurally, it is a quaternary compound and is capable of interacting with any of the anionic sites; steric hindrance prevents it from entering the active site gorge, resulting in binding to the PAS only. None of the antibodies are inactivated by the AChE PAS inhibitor fasciculin, a peptide that interacts very specifically with the PAS and surrounding structures, indicating the lack of a similar peripheral site. The differences between the antibodies and the cholinesterases, seen by their reactions with substrates and inhibitors, also confirms that the activity of the antibodies is not due to AChE or BChE contamination. The frequency of appearance and distribution of activity between the Ab1, Ab2 and Ab3 groups suggests that the activity may not be due to chance. It should be noted, however, that the frequency of 6.9% that was observed for the Ab1s is

probably artificially high, as the cultures were not screened at the pre-cloning stage. Post-fusion screening of Ab3s gave a lower figure of 1.26%. Mice immunised with AChE showed a significantly higher incidence of catalytic immunoglobulins in their sera than both non-immunised mice and mice immunised with MAb E8. Those immunised with AChE complexed with BW284c51 were very similar to those with AChE only; interestingly, the AChE + fasciculin group was significantly less. Possible explanations could be the size of fasciculin and the fact that it obscures a relatively large section of the AChE surface surrounding the PAS (200A2 ; Le Du et al., 1992), compared to the smaller compounds that interact directly with the PAS only, as well as possible differences in dissociation of the AChE–inhibitor complexes. Either way, the results suggest a possible involvement of the PAS and surrounding structures in the development of catalysis (see later). The results also appear to negate our previous hypothesis (Johnson and Moore, 2000) that complexation of AChE with an inhibitor bound at the PAS may have been responsible for the catalytic behaviour. A striking feature of all the antibodies is that they all, to varying degrees, recognise the PAS of AChE. This was shown by the ability of the antibodies to inhibit AChE, by their competition with PAS inhibitors and by their ability to block neuroblastoma cell adhesion, a property of the PAS (Johnson and Moore, 1999; Munoz et al., 1999). It is also suggested by the mouse sera results discussed earlier. This common recognition of the PAS, together with the apparent similarity of the antibodies’ active sites, would suggest that the development of catalytic activity might be a consequence of structural features of the site. The work of Friboulet and co-workers (Izadyar et al., 1993; Joron et al., 1992) is very interesting in this regard. These authors obtained catalytic anti-idiotypic AChE-like antibodies after immunisation with the anti-AChE MAb AE-2. AE-2 inhibits AChE and evidence indicates that it does this by binding to an anionic site, almost certainly the PAS (Sorensen et al., 1987; Wolfe et al., 1993). The anti-idiotypic antibody described by Izadyar et al. (1993) appears similar to ours, albeit a somewhat better catalyst, in its reaction with substrates and inhibitors, suggesting a structurally similar active site to our antibodies. It, therefore, seems possible that structural peculiarities of the PAS may be responsible for the development of catalysis. The PAS is an interesting structure that is involved in a variety of functions: modulation of hydrolysis at the active site (mediated by aromatic residues and its anionic character), interaction with the amyloid ␤-peptide (mediated by a hydrophobic loop; De Ferrari et al., 2001), interaction with an omega loop on an adjacent AChE subunit (Bourne et al., 1999), electrostatically mediated cell adhesion (Botti et al., 1998) and a non-hydrophobic interaction with laminin-1 and collagen IV (Johnson and Moore, in preparation). The PAS is also anionic, as is the transition state of esterase substrates. It is possible that at least part of the antibodies’ activity is germline-encoded and that AChE provides the antigenic stimulus for selection and refinement of the activity. Our

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

results indicate an above-background level of immunoglobulin catalytic activity in non-immunised and Ab2 mouse sera; there was also some activity present in the Ab2 hybridomas at the pre-cloning stage; unfortunately, these were lost after cloning. These results suggest that an innate potential may indeed exist.

4. Conclusions We have described cholinesterase-like catalytic activity and similar active sites in a group of anti-anti-idiotypic antibodies to a catalytic anti-AChE antibody, suggesting idiotypic mimicry of the catalytic antibody’s active site. It appears likely that the development of catalytic activity in the original antibodies to AChE may be a consequence of structural peculiarities of the PAS of AChE; the precise nature of these remains to be determined. These results extend our understanding of the mechanisms by which catalysis may arise in antibodies to ground-state antigens. This is of relevance to autoimmune disease, where the deleterious effects of catalytic autoantibodies are well-documented and, particularly, of autoimmune neuropathies characterised by high levels of anti-AChE autoantibodies.

Acknowledgements This work was supported by grants from the Medical Research Council of South Africa and the Harry Crossley Foundation of the University of Stellenbosch. We thank Drs. A.S. Balasubramanian (anti-BChE antibody), S. Brimijoin (anti-BChE antibody) and M.K. Gentry (anti-AChE antibodies E413D8 and E62A1) for antibody donations. References Benkovic, S.J., 1992. Catalytic antibodies. Annu. Rev. Biochem. 61, 29–54. Botti, S.A., Felder, C.E., Sussman, J.L., Silman, I., 1998. Electrotactins: a class of adhesion proteins with conserved electrostatic and structural motifs. Protein Eng. 11, 415–420. Bourne, Y., Taylor, P., Bougis, P.E., Marchot, P., 1999. Crystal structure of mouse acetylcholinesterase. A peripheral site-occluding loop in a tetrameric assembly. J. Biol. Chem. 274, 2963–2970. Buchbinder, J.L., Stephenson, R.C., Scanlan, T.S., Fletterick, R.J., 1998. A comparison of the crystallographic structures of two catalytic antibodies with esterase activity. J. Mol. Biol. 282, 1033–1041. Carroll, R.T., Grimm, J.L., Hepburn, T.W., Emmerling, M.R., 1995. Purification of acetylcholinesterase by tacrine affinity chromotography. Protein Exp. Purif. 6, 389–393. Conradi, S., Ronnevi, L.O., 1994. Further studies on the occurrence of serum autoantibodies against a membrane-bound acetylcholinesterase fraction in ALS/MND patients and controls. J. Neurol. Sci. 124 (Suppl.), 167–169. De Ferrari, G.V., Canales, M.A., Shin, I., Weiner, L.M., Silman, I., Inestrosa, N.C., 2001. A structural motif of acetylcholinesterase that promotes amyloid ␤-peptide fibril formation. Biochemistry 40, 10447–10457.

287

De la Hoz, D., Doctor, B.P., Ralston, J.S., Rush, R.S., Wolfe, A.D., 1986. A simplified procedure for the purification of large quantities of fetal bovine serum acetylcholinesterase. Life Sci. 39, 195–199. Dixon, M., 1953. The determination of enzyme inhibitor constants. Biochem. J. 55, 170–171. Doctor, B.P., Toker, L., Roth, E., Silman, I., 1987. Microtiter assay for acetylcholinesterase. Anal. Biochem. 166, 399–403. Dodson, G., Wlodawar, A., 1998. Catalytic triads and their relatives. Trends Biochem. Sci. 23, 347–352. Ellman, G.L., Courtney, K.D., Andres, V., Featherstone, R.M., 1961. A new and rapid determination of acetylcholinesterase activity. Biochem. Pharmacol. 7, 88–95. Erhan, S., Greller, L.D., 1974. Do immunoglobulins have proteolytic activity? Nature 251, 353–355. Gao, Q.S., Sun, M., Rees, A.R., Paul, S., 1995. Site-directed mutagenesis of proteolytic antibody light chain. J. Mol. Biol. 253, 658–664. Haggstrom, B., Andersen, P.M., Hjalmarsson, K., Binzer, M., Forsgren, L., 1997. Autoimmunity and ALS: studies on antibodies to acetylcholinesterase in sera. Acta Neurol. Scand. 95, 111–114. Harel, M., Schalk, I., Ehrat-Sabatier, L., Bouet, F., Goeldner, I., Sussman, J.L., 1993. Quaternary ligand binding to aromatic residues in the active site gorge of acetylcholinesterase. Proc. Natl. Acad. Sci. U.S.A. 90, 9031–9035. Harlow, E., Lane, D., 1988. Antibodies: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor. Inestrosa, N.C., Alvarez, A., Perez, C.A., Moreno, R.D., Vicente, M., Linker, C., Casanueva, O.I., Soto, C., Garrido, J., 1996. Acetylcholinesterase accelerates assembly of amyloid ␤-peptides into Alzheimer’s fibrils: possible role of the peripheral site of the enzyme. Neuron 16, 881–891. Izadyar, L., Friboulet, A., Remy, M.-H., Roseto, A., Thomas, D., 1993. Monoclonal anti-idiotypic antibodies as functional internal images of enzyme active sites: production of a catalytic antibody with a cholinesterase activity. Proc. Natl. Acad. Sci. U.S.A. 90, 8876–8880. Johnson, G., Moore, S.W., 1995. Anti-acetylcholinesterase antibodies display cholinesterase-like activity. Eur. J. Immunol. 25, 25–29. Johnson, G., Moore, S.W., 1999. The adhesion function of acetylcholinesterase is located at the peripheral anionic site. Biochem. Biophys. Res. Commun. 258, 758–762. Johnson, G., Moore, S.W., 2000. Cholinesterase-like catalytic antibodies: reaction with substrates and inhibitors. Mol. Immunol. 37, 707–719. Joron, L., Izadyar, L., Friboulet, A., Remy, M.-H., Pancino, G., Roseto, A., Thomas, D., 1992. Anti-idiotypic antibodies exhibiting an acetylcholinesterase enzyme activity. Ann. N. Y. Acad. Sci. 672, 216–223. Kolesnikov, A.V., Kozyr, A.V., Alexandrova, E.S., Kovalewski, F., Demin, A.V., Titov, M.I., Avalle, B., Tramontano, A., Paul, S., Thomas, D., Gabibov, A.G., Friboulet, A., 2000. Enzyme mimicry by the anti-idiotypic antibody approach. Proc. Natl. Acad. Sci. U.S.A. 97, 13526–13531. Kozyr, A.V., Kolesnikov, A.V., Zelenova, N.A., Sashchenko, L.P., Mikhalap, S.V., Bulina, M.E., Ignatova, A.N., Favorov, P.V., Gabibov, A.G., 2000. Autoantibodies to nuclear antigens: correlation between cytotoxicity and DNA-hydrolysing activity. Appl. Biochem. Biotechnol. 83, 255–270. Le Du, M.M., Marchot, P., Bougis, P.E., Fontecilla-Camps, J.C., 1992. Binding of 125 I-fasciculin to rat brain acetylcholinesterase. The complex still binds diisopropyl fluorophosphate. J. Biol. Chem. 267, 22122–22130. Mappouras, D.G., Philippou, G., Haralambous, S., Tzartos, S.J., Balafas, A., Souvatzoglou, A., Lymberi, P., 1995. Antibodies to acetylcholinesterase cross-reacting with thyroglobulin in myasthenia gravis and Graves’ disease. Clin. Exp. Immunol. 100, 336–343. Munoz, F.J., Aldunate, R., Inestrosa, N.C., 1999. Peripheral binding site is involved in the neurotrophic activity of acetylcholinesterase. Neuroreport 10, 3621–3625. Ordentlich, A., Barak, D., Kronman, C., Flashner, Y., Leitner, M., Segall, Y., Ariel, N., Cohen, S., Velan, B., Shafferman, A., 1993. Dissection of

288

G. Johnson, S.W. Moore / Molecular Immunology 39 (2002) 273–288

the human acetylcholinesterase active center determinants of substrate specificity. Identification of residues constituting the active site, the hydrophobic site and the acyl pocket. J. Biol. Chem. 268, 17083– 17095. Paul, S., 1994. Catalytic activity of anti-ground state antibodies, antibody subunits and human autoantibodies. Appl. Biochem. Biotechnol. 47, 241–253. Paul, S., Li, L., Kalaga, R., O’Dell, J., Dannenbring, R.E., Swindells, S., Hinrichs, S., Caturegli, P., Rose, N.R., 1997. Characterisation of thyroglobulin-directed and polyreactive catalytic antibodies in autoimmune disease. J. Immunol. 159, 1530–1536. Rose, N.R., Burek, C.L., 2000. Autoantibodies to thyroglobulin in health and disease. Appl. Biochem. Biotechnol. 83, 245–251. Rosenberry, T.L., 1975. Acetylcholinesterase. Adv. Enzymol. Rel. Areas Mol. Biol. 43, 103–218. Saxena, A., Hur, R., Doctor, B.P., 1998. Allosteric control of acetylcholinesterase activity by monoclonal antibodies. Biochemistry 37, 145–154. Schourov, D.V., Gololobov, G.V., Mkarevich, O.I., Yadav, R.P., Chernova, E.A., Nevinsky, G.A., Prokaeva, T.B., Alekberova, Z.S., Gabibov, A.G., 1995. DNA-hydrolysing autoantibodies in autoimmune pathologies. Ann. N. Y. Acad. Sci. 750, 255–264.

Schumacher, M., Camp, S., Maulet, Y., Newton, M., MacPhee-Quigley, K., Taylor, S.S., Friedmann, T., Taylor, P., 1986. Primary structure of Torpedo californica acetylcholinesterase derived from its cDNA sequence. Nature 319, 407–409. Schwarz, M., Glick, D., Loewenstein, Y., Soreq, H., 1995. Engineering of human cholinesterases explains and predicts diverse consequences of administration of various drugs and poisons. Pharmacol. Ther. 76, 283–322. Shoenfeld, Y., 1994. Idiotypic induction of autoimmunity: a new aspect of the idiotypic network. FASEB J. 8, 1296–1301. Sorensen, K., Brodbeck, U., Rasmussen, A.G., Norgaard-Pedersen, B., 1987. An inhibitory monoclonal antibody to human acetylcholinesterase. Biochim. Biophys. Acta 912, 56–62. Sussman, J.L., Harel, M., Frolow, F., Oefner, C., Goldman, A., Toker, L., Silman, I., 1991. Atomic structure of acetylcholinesterase from Torpedo californica: a prototypic acetylcholine-binding protein. Science 253, 872–879. Wolfe, A.D., Chiang, P.K., Doctor, B.P., Fryar, N., Rhee, J.-P., Saeed, M., 1993. Monoclonal antibody AE-2 modulates carbamate and organophosphate inhibition of fetal bovine serum acetylcholinesterase. Mol. Pharmacol. 44, 1152–1157.