<<

PERSPECTIVES

be answered regarding AChE ; for OPINION example, the mechanism behind the extremely fast turnover rate of the . Despite the fact that the has to navi- — new roles for gate a relatively long distance to reach the , AChE is one of the fastest an old actor enzymes14. One theory to explain this phe- nomenon has to do with the unusually strong electric field of AChE. It has been argued that Hermona Soreq and Shlomo Seidman this field assists catalysis by attracting the cationic substrate and expelling the anionic The discovery of the first — understanding of AChE functions beyond the acetate product15. Site-directed mutagenesis, — was soon followed by the classical view and suggest the molecular basis however, has indicated that reducing the elec- discovery of its hydrolysing enzyme, for its functional heterogeneity. tric field has no effect on catalysis16.However, acetylcholinesterase. The role of the same approach has indicated an effect on acetylcholinesterase in terminating From early to recent discoveries the rate of association of , a acetylcholine-mediated The unique biochemical properties and phys- that can inhibit AChE17. made it the focus of intense research for iological significance of AChE make it an much of the past century. But the complexity interesting target for detailed structure–func- of acetylcholinesterase regulation and tion analysis. AChE-coding sequences have recent evidence for some of the long- been cloned so far from a range of evolution- a Peripheral suspected ‘non-classical’ actions of this arily diverse vertebrate and invertebrate binding enzyme have more recently driven a species that include insects, nematodes, fish, site profound revolution in acetylcholinesterase , and several , among Active site research. Although our understanding of the them man. Sequence data were shortly fol- gorge additional roles of acetylcholinesterase is lowed by the first crystal model for AChE incomplete, the time is ripe to summarize the from Torpedo californica 9, which historically evidence on a remarkable diversity of had been one of the main sources of AChE for acetylcholinesterase functions. research. Later on, crystal structures from Active site O mouse10, Drosophila11 and man12 were Acetylcholine-mediated neurotransmission1,2 obtained and found to be fundamentally sim- OH –O is fundamental for nervous system function. ilar. Surprisingly for an enzyme with an extra- N NH Its abrupt blockade is lethal and its gradual ordinarily rapid catalytic reaction, the acetyl- loss, as in Alzheimer’s disease3, multiple sys- choline site was found to reside at the bottom tem atrophy4 and other conditions5, is asso- of a deep, narrow gorge (FIG. 1a). Site-directed ciated with progressive deterioration of cog- mutagenesis studies13 have also delineated nitive, autonomic and neuromuscular many of the ligand-binding features of this b Ser–AChE functions. Acetylcholinesterase (AChE) enzyme, particularly a peripheral binding site HO CH O 3 AChE O hydrolyses (FIG. 1) and inactivates acetyl- that had been identified in kinetic studies and CH3 choline, thereby regulating the concentration that seems to be fundamental for some of the N+ O CH3 O– of the transmitter at the (BOX 1). ‘non-classical’ functions of AChE. Acetylcholine Acetate Termination of activation is normally depen- AChE can be classified in several ways. CH3 dent on dissociation of acetylcholine from the Mechanistically, it is a . Its cat- CH3 OH– N+ receptor and its subsequent and alytic site contains a catalytic triad — serine, HO CH3 O Choline hydrolysis, except in diseases where acetyl- and an acidic residue (TABLE 1) — as Ser–AChE choline levels are limiting or under AChE do the catalytic sites of the serine proteases O inhibition, conditions that increase the dura- such as trypsin, several blood clotting factors, Acetyl–AChE tion of receptor activation6. and others. However, the acidic group in Figure 1 | Acetylcholinesterase. a | Structural Acetylcholine hydrolysis can also be catal- AChE is a glutamate, whereas in most other features of the enzyme. X-ray crystallography has ysed by a related, less-specific enzyme — cases it is an aspartate residue. The nucle- identified an active site at the bottom of a narrow (BuChE, also known ophilic nature of the carboxylate is trans- gorge, lined with hydrophobic amino-acid side chains. At the time, the catalytic triad was unique as serum or pseudo-cholin- ferred through the imidazole ring of histidine among serine in having a glutamate 7 ) . BuChE can replace AChE by to the hydroxyl group of serine, allowing it to side chain in lieu of the familiar aspartate side hydrolysing acetylcholine and it can also act as displace the choline moiety from the sub- chain. A choline-binding site featured hydrophobic a molecular decoy for natural anti-AChEs by strate, forming an acetyl–enzyme intermedi- residues instead of the expected reacting with these toxins before they reach ate (FIG. 1b). A subsequent hydrolysis step frees anionic groups; a peripheral binding site has also AChE8. However, AChE seems to have many the acetate group. Understanding of the cat- been identified by site-directed mutagenesis. b | The acetylcholinesterase (AChE) reaction. more functions than BuChE as, for example, alytic properties of the protein has assisted in AChE promotes acetylcholine hydrolysis by changes in levels and properties of AChE are our understanding of its inhibition by forming an acetyl-AChE intermediate with the associated with responses to numerous exter- and inhibitors release of choline, and the subsequent hydrolysis nal stimuli. Here, we discuss our current (BOX 2). However, several questions remain to of the intermediate to release acetate.

294 | APRIL 2001 | VOLUME 2 www.nature.com/reviews/neuro © 2001 Macmillan Magazines Ltd PERSPECTIVES

Crystallography and sequence analysis Box 1 | The synapse have identified a group of related and non-catalytic proteins. Some of these are In the presynaptic , choline- a transmembrane proteins with cytoplasmic acetyltransferase (ChAT) catalyses the Presynaptic neuron domains and extracellular AChE-homologous synthesis of acetylcholine (ACh) from choline ChAT Acetyl-CoA domains that share the unique topography of and acetyl-coenzyme A (panel a). ACh is vAChT ACh + choline AChE and its strong electric field (TABLE 1)18. packaged in synaptic vesicles via a vesicular On the basis of their structures, all of these ACh transporter (vAChT). Action potentials High-affinity (–) choline uptake are classified as α/β-fold proteins; on the trigger the release of ACh into the synaptic K+ basis of their electric fields, they are classified cleft, where ACh can bind to muscarinic as electrotactins19. receptors located on the pre- and postsynaptic AChE from different species are membrane. Muscarinic M2 receptors (M2) on the presynaptic membrane regulate ACh Choline organized and sequentially spliced in a man- + acetate release via a negative feedback response. At the ner associated with distinct domains in their postsynaptic site, M1 receptors transduce ACh protein products. They include sites for alter- signals through a pathway involving AChE-R native splicing of the pre-mRNA both at the 22 diacylglycerol (DAG), inositol-1,4,5- 5′ (REFS 20,21) and 3′ ends . Alternative splic- 2+ trisphosphate (Ins(1,4,5)P3) and a Ca - ing allows the production of three distinct dependent protein kinase (PKC). In the + 2+ AChE variants, each with a different carboxy- hippocampus, most of the postsynaptic K Ins(1,4,5)P3 Ca terminal sequence — the ‘synaptic’ (S),‘ery- receptors are of the M1 subtype; in the cortex cAMP DAG PKC throcytic’ (E) and ‘readthrough’ (R) AChE M2 receptors might also be located on the Postsynaptic site isoforms. The carboxy-terminal sequences postsynaptic membrane. Genomic disruption determine their homologous assembly into of the M1 receptor impairs the activation of ACh M1 M2 AChE-S vAChT AChE and their heterologous asso- several signal-transduction pathways and ciation with non-catalytic subunits that explains why muscarinic excitation is the direct the subcellular localization of the pro- primary cause of seizures106. ACh is hydrolysed b Cortex Hippocampus tein. In AChE-S, a cysteine located three in the synaptic cleft by AChE-S tetramers, residues from the carboxyl terminus of the which are indirectly attached to the human protein allows dimerization by disul- by a -like phide bridging. Two additional monomers tail5, or by another structural subunit to brain 107 can become associated by hydrophobic inter- . AChE-R monomers would remain actions23. These tetramers can attach cova- soluble within the synaptic cleft. A high- lently to a hydrophobic P subunit or to a col- affinity choline-uptake mechanism returns lagen-like protein known as the T subunit24. choline to the presynaptic neuron. Brain distribution of AChE includes both The collagen-like subunit has a polyproline acetylcholine-releasing and cholinoceptive sequence that can form a triple-helical struc- . Panel b shows a cranial section of a Amygdala ture that bundles together 4, 8 or 12 AChE-S brain stained for AChE activity (reproduced subunits23. In AChE-E, a glycyl bond near the with permission from REF. 108 © (1997) Harcourt). Note pronounced activity in the amygdala carboxyl terminus undergoes transamidation and caudate–putamen (striatum), and clearly detectable activity in the cortex and hippocampus. to attach a glycophosphatidylinositol group to the protein, which anchors the mature AChE-E to the outer surface of erythrocytes25. embryonic neurite extension and muscle however, have been confirmed and several AChE-R does not seem to have any feature development and before synaptogenesis28,29. research groups have established their mole- that allows for its attachment to other mole- In addition, the enzyme was also found in cular foundations. cules and it remains monomeric and soluble. adult non-cholinergic neurons, and in Last, it must be pointed out that another haematopoietic, osteogenic and various neo- Neuritogenesis. The first non-classical activity nomenclature labels the synaptic and ery- plastic cells. Early reports of a soluble, secret- of AChE that was clearly distinguished from thropoietic variants according to properties ed, monomeric form of AChE30 prompted its hydrolytic capacity was its role in neurito- of the proteins26 — they are termed T the idea that this enzyme could have non- genesis. Exogenous purified AChE promoted (tailed), and H (hydrophobic), respectively. enzymatic functions. Gradually, this encour- neurite growth from chick nerve cells in cul- The many guises in which AChE-T occurs aged a small but persistent group of investi- ture and, whereas several active-site inhibitors are also known as G for globular and A for gators to argue for the existence of failed to attenuate this effect34, an inhibitor of asymmetric. ‘non-classical’ activities for AChE31.Some of the peripheral site did block neuritogenesis35. the currently proposed functions are based Transfection with AChE or with antisense Multiple activities of AChE merely on correlations and circumstantial AChE cRNA-encoding vectors subsequently The idea that AChE has multiple, unrelated evidence, and might yet be disproved. For showed neuritogenic acitivity of the enzyme biological functions is not obvious and, example, the suggestion that AChE has in neuroblastoma cells36, in phaeochromocy- indeed, was not readily accepted. Cyto- intrinsic proteolytic activity was argued for toma (PC12) cells37 and in primary dorsal chemical data, however, attested to spatio- several years (for example, REF. 32), but even- root ganglion neurons38. Similarly, Xenopus temporally regulated expression of AChE tually was proved wrong by careful separa- motor neurons that expressed human AChE-S during very early embryogenesis27, during tion of the two activities33. Other activities, showed enhanced neurite growth rates (FIG. 2).

NATURE REVIEWS | NEUROSCIENCE VOLUME 2 | APRIL 2001 | 295 © 2001 Macmillan Magazines Ltd PERSPECTIVES

This non-classical autocrine phenomenon Natural and man-made anti- Box 2 | remains to be independently confirmed by the Owing to its critical role in acetylcholine-mediated neurotransmission, AChE is a sensitive target use of other approaches, but it might prove to for both natural and synthetic cholinergic toxins. Among the natural anti-AChEs are plant- be of physiological relevance. derived and glycoalkaloid inhibitors109. A natural inhibitor of AChE was also found in a mollusc110. Blue- algae are equipped with anatoxins111, highly effective toxins that block Amyloid fibre assembly. AChE has been the active site. Green mamba venom includes the neurotoxic peptide fasciculin, which blocks reported to promote amyloid fibre assembly46. entrance to the active and the peripheral sites of AChE10. Use of anti-AChEs as defence and attack The same group found that this activity was weapons in nature, therefore, preceded their use by humans. blocked by the peripheral-site inhibitor pro- Synthetic anti-AChEs were first studied and manufactured as highly poisonous pidium, but not by the active-site AChE organophosphate and carbamate nerve gases and . In the clinic, controlled use of inhibitor , clearly identifying this AChE inhibitors has proved valuable for the treatment of diseases that involve compromised assembly-promoting activity as a non-classical acetylcholine-mediated neurotransmission (BOX 4). AChE function. AChE–amyloid-β complexes The phosphate or carboxylate group of the active-site inhibitor acts like the carboxyl-ester group showed AChE activity that was resistant to of a substrate to produce a stable phosphate or carboxylate ester, analogous to the acetyl-enzyme low pH, lacked the substrate inhibition that is intermediate of catalysis. In this form, the enzyme is unavailable for its physiological function. The characteristic of the enzyme and had reduced figure shows the reaction of AChE with the organophosphate, , a agent. sensitivity to anti-AChEs, properties that had previously been shown histochemically for O O AChE activity associated with Alzheimer’s OFP + Ser–AChE OOSerP –AChE + HF disease plaques3. The limitations of this study, HO CH3 CH3 however, include the high concentration of After inhibition of AChE by and carbamates, a slow spontaneous hydrolysis AChE that was needed to bind and promote of the blocked enzyme regenerates an active enzyme. However, with organophosphate poisons, amyloid fibre formation. In addition, only some of the acidic hydroxyl groups on the phosphate are esterified; slow spontaneous hydrolysis AChE binds amyloid fibrils in vitro, but both of these ester groups renders the phosphoryl-AChE permanently inactive. Antidotes for such AChE and BuChE bind it in vivo. poisoning have been developed to accelerate enzyme regeneration112. The slow reactivation of AChE was the basis of the use of during the Gulf War as a prophylactic in Haematopoiesis and thrombopoiesis. anticipation of the use of organophosphate poisons (see BOX 3). It was argued that the prior Haematopoietic and thrombopoietic activi- blocking of AChE with pyridostigmine (a derivative) prevents reaction with an ties were first proposed for AChE on the basis organophosphate, and the slow reactivation regenerates AChE activity after the threat of of its presence in blood-cell progenitors47,48. 64 poisoning has passed. The feedback response to anti-AChEs, including pyridostigmine , Recent analyses have shown transcriptional indicates that AChE over-production might function as an additional protective mechanism activation of ACHE during haemaglutinin- against organophosphate poisoning. induced lymphocyte activation. Inhibitor tests have suggested a mixed involvement of both catalytic and non-classical properties in Cell adhesion. The adhesive properties of the another non-catalytic transmembrane pro- this process49. Transient suppression and sub- AChE core domain with the variant-specific tein that has an AChE-homologous extracel- sequent overproduction of AChE after treat- carboxyl termini removed were studied in cul- lular domain and a carboxy-terminal cyto- ment with antisense oligodeoxynucleotides ture experiments using Drosophila cells. plasmic tail. When expressed in non-neuronal induced myeloid proliferation in mouse Chimeras were constructed in which the cells, induced synaptic vesicle clus- AChE-homologous domain of the cell-adhe- tering and presynaptic differentiation in adja- sion protein neurotactin39 was replaced with cent axons. This phenomenon was prevented

) 40 1 the homologous Drosophila or Torpedo AChE by the addition of the extracellular domain of – β 42 core. These chimeras retained the adhesive -, a binding partner of neuroligin , m h properties in a homotypic cell-aggregation which can presumably also bind AChE. µ 40 assay , whereas AChE alone showed no such Combined with the immunohistochemical 20 cell-adhesion property. However, AChE, unlike proof of the presence of neuroligin in adult 43 neurotactin, lacks a transmembrane domain. excitatory brain synapses , these findings rate ( Growth It was therefore argued that soluble AChE indicate a probable involvement of neuroligin (that is, AChE-R) might compete with its in synapse development and remodelling. 0 structural homologues for their binding part- Again, the actual role of AChE in this phe- Control S SIn R Core ners, and thus convey or interrupt morpho- nomenon, if any, remains to be tested in a AChE 37 genic signals into neurons (FIG. 3).However, more direct manner. Figure 2 | Neuritogenesis. The growth of neurites the participation of AChE in interactions of in cultured motor neurons from Xenopus tadpoles this type remains to be directly shown. Activation of neurons. Real-time was studied by time-lapse photography41. Rates of chemiluminescence has been used to visualize growth were observed in normal tadpoles and . A synaptogenic function of AChE release from dendrites of dopamine transgenics bearing coding sequences for human 44 AChE-S (S), AChE-R (R), an enzyme with no AChE that depended on its catalytic proper- neurons in mammalian substantia nigra . variant-specific carboxyl terminus (core), or an 41 ties has been observed in microinjected Purified recombinant AChE was later shown enzymatically inactive AChE-S into which a seven- Xenopus tadpoles. However, synaptogenic by the same group to enhance dopamine amino-acid stretch was inserted near the active activity has been also shown for neuroligin, release from midbrain dopamine neurons45. site (SIn). Comparison with control values is shown.

296 | APRIL 2001 | VOLUME 2 www.nature.com/reviews/neuro © 2001 Macmillan Magazines Ltd PERSPECTIVES

Table 1 | The α/β-fold superfamily* Protein Origin Triad Intra- Splicing Trans- Glyco- Assembly residues molecular variants membrane phospholipid potential disulphides domain linkage Cholinesterases Acetylcholinesterase Mammalian S E H 1 2 3 + – + + Acetylcholinesterase Torpedo S E H 1 2 3 + – + + Butyrylcholinesterase Mammalian S E H 1 2 3 – – – + Cholinesterase Insect S E H 1 2 3 – – + + Other Mammalian S E H 1 2 – – – – – Cholesterol esterase Mammalian S D H 1 2 – – – – – Esterases 6 and P Insect S D H 1 2 – – – – – Esterase B1 Culex S E H 1 – – – – – – Juvenile hormone esterase Heliotis S E H 1 – – – – – – D2 esterase Dictostelium S E H 1 2 3 – – – – Crystal protein Dictostelium S E H 1 2 3 – – – – Geothricum, Candida S E H 1 2 – – – – – Non-enzyme proteins Thyroglobulin Mammalian – – H 1 2 3 – – – – Neuroligin Mammalian – E H 1 2 – + + – – Neurotactin Drosophila – E H 1 2 3 – + – – Glutactin Drosophila – – – 1 2 – – – – – Gliotactin Drosophila – – – 1 2 – – + – – *Three related subfamilies belong to the α/β-fold superfamily on the basis of their protein-folding patterns: cholinesterases, other esterases and non-enzyme proteins. Sequence alignments show two prototypic ‘signatures’ in all of these, one near the serine at the enzyme active site (prosite accession no. PS00122) and the other close to the amino terminus, with a conserved cysteine involved in disulphide bridging (prosite accession no. PS00941). The non-enzyme proteins include a growing number with AChE-homologous extracellular domains and, in some cases, a single transmembrane region with a protruding cytoplasmic domain that can interact with intracellular signalling proteins. Note that the possibilities of glycophospholipid linkage and multisubunit assembly are unique to the cholinesterases, whereas the conserved positions of the three (presumed) intramolecular disulphide bonds have been evolutionarily conserved in many of the members of this superfamily. The triad is complete only in those members of the superfamily that are enzymes. These findings are summarized from REF. 18. bone-marrow-cell cultures50. This finding Regulation of ACHE expression that provides rapid responses to physiological might explain the increased risk of leukaemia Functional heterogeneity in AChE activity is stimuli (FIG. 4). among users of anti-AChE pesticides51,as regulated at the transcriptional, post-tran- Physiological cues that induce ACHE tran- anti-AChEs induce AChE overproduction52.A scriptional and post-translational levels, lead- scription include cell differentiation62, reduced related issue is whether AChE can itself be ing to complex expression patterns that AChE levels63 and acetylcholine-mediated tumorigenic. The ACHE gene is frequently reflect tissue and cell-type specificity, differen- excitation elicited by exposure to anti-AChE amplified, mutated or deleted in DNA from tiation state, physiological condition and agents and various traumatic insults64,65. leukaemic patients53,54, supporting the idea response to external stimuli. The concentra- AChE production increases during nervous that it is involved in myeloid proliferation. tion of AChE in a tissue, the distribution of its system development in avian28 and mam- However, no mechanism has yet been pro- alternative isoforms, its mode of oligomeric malian systems (for example, REF.66), and is posed to explain this effect, and it is not known assembly, and its subcellular disposition, gly- subject to regulation by the nerve growth fac- whether the mechanism is directly related to cosylation55 and proteolytic processing, are all tor receptor TrkA67. In situ hybridization and enzyme activity. subject to modulation through mechanisms cytochemical studies have shown transient The heterogeneity of non-classical func- that are not fully understood. Changes in AChE activation in thalamic neurons during tions of AChE is further complicated by the most of these steps have indeed been shown mammalian development68, and muscle dif- difficulty of distinguishing these functions to regulate the functional heterogeneity of ferentiation is associated with transcriptional from its classical regulatory role in terminat- AChE variants (for example, REFS 56–58). activation of the ACHE gene in subsynaptic ing acetylcholine-mediated neurotransmis- nuclei62. sion. For example, it is likely that AChE Transcription-regulated heterogeneity. Trans- The ACHE gene retains a certain level of would have both types of activity at acetyl- criptional control of AChE production plasticity in the adult. In neocortical and hip- choline neurons. However, inhibitor blockade depends on a principal and an alternative pocampal neurons, various external stimuli of its hydrolytic activity does not necessarily proximal promoter21,59, on a distal enhancer induce rapid, long-lasting activation of ACHE block its non-classical functions. To under- domain60, and on an internal enhancer posi- expression. In fact, psychological stress64,envi- stand the relationship between the forms and tioned within the first intron61. Sequential ronmental stimuli (anti-AChE intoxication)60 functions of AChE, a deeper insight is needed splicing of AChE pre-mRNA and regulation and head injury69 increase ACHE transcrip- into the molecular origins of the functional at the 5′ and 3′ ends yields tissue-, cell-type- tion (FIG. 4). It is presumed that the initial stim- heterogeneity of AChE. and developmental-state-specific regulation ulus induces excitation through acetylcholine

NATURE REVIEWS | NEUROSCIENCE VOLUME 2 | APRIL 2001 | 297 © 2001 Macmillan Magazines Ltd PERSPECTIVES

N tion 322 creates the Ytb blood group, indicat- ing that this residue might be part of an Neurotactin exposed epitope in AChE-E. Its frequency is higher in the Middle East as compared with C European populations81,82. Nevertheless, the β-neurexin? β-neurexin H322N mutation has no effect on the cat- N alytic properties of the enzyme. The negligi- ble polymorphism in the ACHE gene, in C AChE-S N comparison with the abundant mutations in N the homologous BCHE gene83, has been ? interpreted as a reflection of the absolute C necessity of a functional AChE. Chemical N hypersensitivity to anti-AChEs was found in Extracellular C human carriers of an autosomal-dominant transcription-activating in a distal 60 Intracellular ACHE enhancer domain . Parallel hypersen- C AChE- sitivity associated with progressive learning, neurotactin chimera Neuroligin-1 memory and neuromuscular impairments C was found in mice with a moderate overpro- N duction of transgenic human AChE-S84, indicating that the level of this protein must Figure 3 | Proposed mechanism of some non-classical AChE functions. AChE-S (blue) is shown by be kept within a narrow window. It was a folded line, as are the homologous regions of neurotactin (magenta) and neuroligin-1 (green). AChE-S therefore surprising that genomic disruption linked to the membrane through a structural subunit (orange) might be associated with β-neurexin (yellow) in a neighbouring molecule or one embedded in the same membrane (left). Alternatively, if they are of the mouse ACHE gene yielded unhealthy, 85 embedded in pre- and postsynaptic sites, neuroligin and β-neurexin might mediate cell–cell adhesion yet viable homozygotes .However,whereas (right). Neurotactin might adhere to another neurotactin molecule (not shown), or be linked to AChE AChE can be temporarily replaced by BuChE, through a yet-to-be identified bridging protein (?, purple). Core AChE, or similar proteins cannot form any its absence becomes lethal within three of these interactions by themselves. However, when the homologous region of AChE replaced a region of weeks8. One possibility to explain this finding neurotactin, the chimeric protein retained the cell–cell adhesion capacity of the original neurotactin40. It has 37 is that AChE, unlike BuChE, has three alter- therefore been proposed that AChE-R, which lacks the anchor to the membrane, might compete with native variants, with potentially distinct neurotactin, neuroligin-1 or AChE-S and modify cell–cell interactions and intracellular signalling. functions. It is therefore important to discuss the functional significance of alternative release and feedback overexpression of AChE AChE-overproducing transgenic mice, acetyl- splicing of the ACHE gene. acts to dampen excessive neurotransmission choline levels are reduced but acetylcholine- back towards normal levels64. This is impor- mediated neurotransmission is compensated Specific functions for splice variants? tant both for acetylcholine-mediated neuro- by increases in high-affinity choline transport Transcriptional activation of ACHE is often transmission and for other brain circuits and acetylcholine synthesis75. associated with a shift in its splicing pattern, modulated by acetylcholine, for example, glu- Last, transcriptional activation would also leading to accumulation of the rare AChE-R tamate-mediated transmission in the hip- impinge on the non-classical actions of AChE. variant. For example, AChE-R mRNA levels pocampus70. Excess AChE can also protect the The consequences associated with this effect increased considerably within 30 minutes of organism from anti-AChEs toxicity, which has are still unknown but they might be relevant confined swim stress, or after exposure to been shown in injection experiments71. to ongoing discussions on Gulf War syn- anti-AChEs or acetylcholine analogues. Transcriptional activation is common to drome76 (BOX 3) and the danger of terrorist Similar effects have been observed in mam- many genetically determined responses to attacks with nerve gas77. malian brain neurons in vivo after stress, in drugs and has been observed, for example, for cortical and hippocampal neurons in brain cytochrome P450 proteins72.However,in Mutation-related phenotypes. Congenital slices and in cultured HEK293 cells. This addition to transcriptional activation, AChE myasthenia has been traced to a mutation shift depends on neuronal activity, mus- mRNA transcripts in neuron, muscle and not in ACHE itself but in the COLQ gene, carinic signalling and intracellular Ca2+ levels. blood cells are subject to -con- which encodes the collagen-like structural In fact, AChE-R accumulation can be inhib- trolled, differentiation-induced stabiliza- subunit of neuromuscular AChE-S24,78. ited in brain slices by tetrodotoxin or by the tion56,73. Both these processes can increase the Natural amplification or mutation of the Ca2+ chelator BAPTA-AM64, and it is facilitat- amount of AChE when and where it is need- Drosophila ache gene results in ed by transfection with the M1 muscarinic ed, with AChE-R mRNA being significantly resistance79, which is an inherited response to receptor in HEK293 cells86. After head injury, less stable than AChE-S mRNA56. This fact environmental exposure. Amplification of neuronal AChE-R accumulation persists for should be of specific interest, as anti-AChEs the human BCHE gene, which encodes over two weeks69. are routinely used in the treatment of patients BuChE, has also been shown in the case of Stress-induced also with Alzheimer’s disease74. one family of farmers exposed to similar modifies K+ channels87 and is regulated by Chronically overproduced AChE can be insecticides80. Nucleotide polymorphisms in neuronal activity-dependent transcriptional expected to change acetylcholine balance, the human ACHE gene are rare, and have a changes in several splicing regulatory inducing activity-dependent, secondary feed- mild or no effect on the protein properties. proteins88 (FIG. 4).In ACHE , back responses in the nervous system. In Histidine substitution for asparagine at posi- alternative splicing changes the ratio of the

298 | APRIL 2001 | VOLUME 2 www.nature.com/reviews/neuro © 2001 Macmillan Magazines Ltd PERSPECTIVES

AP-1 22465 biological function of the human protein in NFκB C/EBP E1 I1 E2 I2 E3 I3 E4I4 E5 E6 the evolutionarily remote Xenopus system. HFH-2 However, differences between AChE-S and GRE AChE-R were also observed in mammalian 5267 5484 Alternative splicing HNF3β glia. In rat C6 glioma, transfected AChE-S HNF5α/β induced process extension but AChE-R facili- Deletion tated the formation of small, round cells95.No molecular mechanism has yet been proposed E1 E2 E3 E4 E6 E1 E2 E3 E4 E5 E1E2 E3 E4 4′ E5 for this function. S E R Regulation of stress-induced disorders. Trans- Figure 4 | The human ACHE gene and its alternative messenger RNAs. The core of human AChE is genic mice expressing human AChE-R show encoded by three exons and parts of additional regions encode the variant-specific carboxy-terminal lower levels of stress-associated hallmarks of sequences. Transcription begins at E1, and E2 encodes a leader sequence that does not appear in any pathology (accumulation of heat-shock pro- mature protein. In addition to a proximal promoter (red line adjacent to E1), a distal enhancer region (more teins, presence of reactive glia and curled distal red line) is rich in potential regulatory sequences, some of which are shown as wedges. The transcriptional activation of ACHE by cortisol58 is probably due to the distal glucocorticoid response neurites), whereas transgenic with element (GRE). A deletion mutation in this region disrupts one of two HNF3 (hepatocyte nuclear factor 3) human AChE-S show accelerated stress- binding sites, a factor that also activates transcription60. Intron 1 (I1) contains an enhancer sequence56,57 related pathology96. AChE-S transgenic ani- indicated by a red dot. Nucleotide numbers are those of GeneBank cosmid AF002993. Normally, much mals, which fail to respond to external stimuli more AChE-S than AChE-R mRNA is produced, but under stress or inhibition of AChE, alternative splicing by AChE-R overproduction, are exceptionally produces much more of the AChE-R mRNA. sensitive to closed-head injury69 and to expo- Animated online sure to AChE inhibitors60. This indicates a probable physiological relevance for the multimeric AChE-S to the monomeric Induction of process extension. When ex- alternative splicing of AChE in response to AChE-R variant. These variants share a similar pressed in developing Xenopus motor neu- stressors, although it is not yet clear whether hydrolytic activity, but differ in their cellular rons, human AChE-S but not AChE-R or a the stress-induced changes in AChE expres- and subcellular distribution and might have carboxy-terminal truncated AChE, promoted sion are a cause or an effect. In addition, the distinct non-classical functions. In nematodes, neurite extension. This neuritogenic function stress-induced AChE-R protein seems to be for example, four different genes encode dis- was independent of the catalytic capacity, as a cleaved to yield its unique carboxy-terminal tinct AChE variants. They resemble the mam- catalytically inactive AChE-S variant retained peptide (which by itself has haematopoietic malian variants in their carboxyl termini89, this function41. This finding argues for a neu- growth-factor-like action), increasing DNA further emphasizing the likelihood of isoform- ritogenic activity of the carboxy-terminal replication and survival of CD34+ progeni- specific functions. Unequivocal evidence for sequence that is unique to AChE-S. A limita- tors in human blood cell cultures and in variant-specific differences in the non-classical tion of this study, however, is that it tested a mice subjected to confined swim stress58. functions of AChE is still sparse. Nevertheless, some information attesting to such functions now exists. Box 3 | Gulf War syndrome The heterogeneous cognitive and physiological impairments reported by veterans of the 1991 Variant-specific protein transport. The loca- Persian Gulf War are informally referred to as the Gulf War syndrome (GWS). During the war, tion of AChE-S in the neuromuscular junc- soldiers were exposed to harsh weather conditions, high levels of anti-AChE insecticides and tion (NMJ) has been studied in great numerous vaccinations. In addition, the peripherally acting carbamate anti-AChE, 90 detail . Enzyme clustering was found to be pyridostigmine bromide, was administered as a prophylactic in anticipation of chemical warfare. mediated by the heparan sulphate proteo- In short-term peacetime tests, pyridostigmine caused no incapacitation and elicited minimal glycan perlecan and to involve complex for- central nervous system effects113. However, factor analysis indicated delayed, but significant, mation with dystroglycan91. In microinject- impairment of cognition, ataxia and arthro-myo-neuropathy in some veterans. These symptoms ed Xenopus tadpoles, transgenic AChE-S were interpreted to reflect exposure to centrally acting anti-AChEs114. The idea that GWS is accumulated in the NMJ, whereas AChE-R related, at least in part, to the use of anti-AChEs in the war suggested a role for polymorphism in was found in epidermal cells92. AChE-S, but the ACHE, BCHE and genes (for example, REF. 60), or battlefield stress115 as factors not AChE-R, would adhere to the active promoting neurological symptoms and muscle weakness. It was argued that the psychological zone at the NMJ (see REF. 93 for a tomo- stress associated with war conditions impaired the integrity of the blood–brain barrier (BBB), graphic microscopic analysis of these junc- allowing penetrance of the otherwise peripheral-acting inhibitor pyridostigmine into the tions). Transgenic mice with a disrupted brain116, where it activated ACHE transcription64. Although stress-induced BBB leakage was 117,118 ColQ collagen-like AChE-S subunit have difficult to reproduce by some (perhaps because perfusion removed the over-produced congenital myasthenia94, demonstrating that secretory, soluble AChE-R), more recent studies confirmed the stress-promoted BBB disruption and attributed at least part of it to brain mast-cell activation119. Moreover, anti-AChEs were AChE-S is an essential component of the 120 NMJ. However, the secretory, soluble reported to facilitate viral invasion of the nervous system , and stress was found to increase lethality of low levels of an anti-AChE pesticide121, further strengthening the links between stress, AChE-R might also change acetylcholine AChE and BBB integrity. So, the possibility that exposure to anti-AChEs together with acute balance in the synaptic cleft52 and/or com- stress and a compromised immune system might have contributed to the cholinergic pete with AChE-homologous proteins (for consequences of GWS cannot be ruled out, a conclusion also of the Rand Corp report on the example, neuroligin) for interaction with pyridostigmine and GWS literature. their binding partners (BOX 1).

NATURE REVIEWS | NEUROSCIENCE VOLUME 2 | APRIL 2001 | 299 © 2001 Macmillan Magazines Ltd PERSPECTIVES

different splice variants. Whereas previous Box 4 | Reassessing the role of AChE in neurological disease theories on the involvement of AChE in neu- Pharmacological inhibitors of AChE are important in controlling diseases that involve impaired rophysiology were largely limited to acetyl- acetylcholine-mediated neurotransmission. For example, Alzheimer’s disease involves selective choline-mediated neurotransmission, acetyl- loss of cholinergic neurons in the brain122. In , auto-antibodies reduce the choline is known to modulate additional number of nicotinic acetylcholine receptors at the neuromuscular junction123. AChE inhibition circuits, for example, glutamate-mediated increases the synaptic concentration of acetylcholine and allows a higher occupancy rate and hippocampal activity70. It would be very longer duration at its receptor124. Nevertheless, anti-AChE therapeutics do not address the interesting to know, for example, if the solu- aetiology of the diseases for which they are used. ble AChE-R monomers secreted under stress Neurodegeneration serve to modulate glutamate-mediated neuro- Neuritic plaques from Alzheimer’s patients include catalytically active AChE3. Transgenic mice transmission or affect the stress-induced overexpressing AChE show progressive cognitive deficits and neuropathological markers (for changes in long-term potentiation104 or long- example, cortical dendrite and spine loss) reminiscent of Alzheimer’s disease125.Also, term depression105. AChE–amyloid-β complexes were more neurotoxic than amyloid-β alone126, suggesting Together, data obtained by geneticists, neu- that AChE imbalances might be involved in neurodeterioration. Last, increased AChE levels were robiologists and biochemists during the past reported in the brains of transgenic mice that express the carboxy-terminal fragment of the few years forcefully argue that the AChE story amyloid-β protein127. Reconsideration of the rationale and expectations of conventional anti- continues. Once again, the infinite complexity AChE therapeutics is therefore required, as anti-AChEs induce AChE accumulation in of nature is shining through, demanding a cerebrospinal fluid65. new look at an old friend. Neuromotor dysfunction Hermona Soreq and Shlomo Seidman are In chickens, muscle dystrophy is associated with changes in AChE properties128,129.In humans, at the Department of Biological Chemistry, congenital myasthenia might be caused by ColQ mutations5,24, and standard treatment of Institute of Life Sciences, The Hebrew University myastenia gravis includes life-long administration of AChE inhibitors. However, AChE inhibitors of Jerusalem, Israel 91904. promote overexpression of AChE-R in muscle, and exert dire effects on muscle integrity and Correspondence to H.S. e-mail: [email protected] neuromuscular junction structure in healthy mice52. This suggests that AChE-R might participate in the aetiology of myopathic syndromes and questions the potential long-term effects of anti- Links AChE drugs on disease progression. The unexplained muscle weakness reported by some of the DATABASE LINKS Gulf War veterans who were exposed to various anti-AChEs might also be relevant here (see BOX 3). AChE | BuChE | COLQ FURTHER INFORMATION The ESTHER Antisense suppression makes sense database | GWS literature In vivo antisense destruction of AChE mRNA using chemically protected oligonucleotides ENCYCLOPEDIA OF LIFE SCIENCES directed against AChE messenger RNA prevents AChE overproduction induced by Acetylcholine organophosphate AChE inhibitors in muscle52, arrests the haematopoietic process elicited by this 58 overproduction under acute psychological stress and improves survival and recovery from 1. Dale, H. The action of certain esters and ethers of closed-head injury69. Antisense agents, therefore, emerge as a promising alternative to anti-AChE choline, and their relation to . J. Pharmacol. Exp. Therap. 6, 147–190 (1914). therapeutics, especially as this is the only approach that promises a degree of selectivity among 2. Loewi, O. & Navratil, E. Uber humorale Ubertragbarkeit the several variants of AChE58. der Herznervenwirkung. X Mitteilung. Pfluger’s Arch. 214, 678–688 (1926). 3. Wright, C. I., Geula, C. & Mesulam, M. M. Neurological Prospects ulated ACHE gene expression permit behav- cholinesterases in the normal brain and in Alzheimer’s disease: relationship to plaques, tangles, and patterns of There are many challenges left regarding the ioural, electrophysiological and developmen- selective vulnerability. Ann. Neurol. 34, 373–384 (1993). non-classical functions of AChE. First, it is tal studies. Mating AChE-overexpressing mice 4. Polinsky, R. J., Holmes, K. V., Brown, R. T. & Weise, V. CSF acetylcholinesterase levels are reduced in multiple necessary to substantiate the nature of the with transgenic animals that overexpress the system atrophy with autonomic failure. Neurology 39, protein regions and the amino-acid residues human amyloid-β might prove the involve- 40–44 (1989). 5. Ohno, K. et al. The spectrum of mutations causing that mediate non-classical activities. Second, ment of AChE in the aetiology of Alzheimer’s end-plate acetylcholinesterase deficiency. Ann. Neurol. additional AChE binding partners that trans- disease (BOX 4). Moreover, the morphogenic 47, 162–170 (2000). 6. Silver, A. A histochemical investigation of cholinesterases fer morphogenic signals into the cell must be capacities that are peculiar to AChE indicate at neuromuscular junctions in mammalian and avian identified. Third, the intracellular signalling that it might modify the differentiation of muscle. J. Physiol. (Lond.) 169, 386–393 (1963). 7. Augustinsson, K. B. & Nachmansohn, D. Distinction 101 cascades that translate AChE-mediated events embryonic stem cells . Finally, positron between acetylcholinesterase and other choline ester- into altered cellular properties should be emission tomography of AChE activity might splitting enzymes. Science 110, 98–99 (1949). 8. Li, B. et al. Abundant tissue butyrylcholinesterase and its delineated and the physiological significance be developed to the level of real-time imaging possible function in the acetylcholinesterase knockout of such cascades should be tested. To this end, of this enzyme in the human brain. Tracers mouse. J. Neurochem. 75, 1320–1331 (2000). 97 98 9. Sussman, J. L. et al. Atomic structure of two-hybrid and/or phage-display screens that are hydrolysed by AChE already permit acetylcholinesterase from Torpedo californica: a can guide us to new protein partners. measurement and imaging of AChE activity prototypic acetylcholine-binding protein. Science 253, 872–879 (1991). Mutagenesis and co-crystallization of these in humans and have shown a reduction of 10. Bourne, Y., Taylor, P. & Marchot, P. Acetylcholinesterase protein pairs can reveal the nature of their cerebral AChE activity in Alzheimer’s102 and inhibition by fasciculin: crystal structure of the complex. 103 Cell 83, 503–512 (1995). interactions. Conditional genetic disruption Parkinson’s diseases . Further development 11. Harel, M. et al. Three-dimensional structures of of AChE variants and their protein partners99 of this approach can make AChE a sensitive, Drosophila melanogaster acetylcholinesterase and of its 100 complexes with two potent inhibitors. Protein Sci. 9, combined with DNA-microarray analyses reliable biosensor of changes in acetylcholine- 1063–1072 (2000). can reveal the biological pathway(s) in which mediated neurotransmission, perhaps even 12. Kryger, G. et al. Structures of recombinant native and E202Q mutant human acetylcholinesterase complexed these interactions are involved. under stressful conditions. with the -venom toxin fasciculin-II. Acta Crystallogr. D Biol. Crystallogr. 56, 1385–1394 (2000). At the level of the whole organism, the One of the key challenges in this field is to 13. Shafferman, A. et al. Mutagenesis of human various models with up- or downreg- search for the physiological functions of the acetylcholinesterase. Identification of residues involved in

300 | APRIL 2001 | VOLUME 2 www.nature.com/reviews/neuro © 2001 Macmillan Magazines Ltd PERSPECTIVES

catalytic activity and in polypeptide folding. J. Biol. Chem. 36. Koenigsberger, C., Chiappa, S. & Brimijoin, S. Neurite Taylor, P., Quinn, D. M., Rotundo, R. L. & Gentry, M. K.) 267, 17640–17648 (1992). differentiation is modulated in neuroblastoma cells 51–55 (Plenum, New York, 1998). 14. Nair, H. K., Seravalli, J., Arbuckle, T. & Quinn, D. M. engineered for altered acetylcholinesterase expression. 58. Grisaru, D. et al. ARP, a peptide derived from the stress- Molecular recognition in acetylcholinesterase catalysis: J. Neurochem. 69, 1389–1397 (1997). associated acetylcholinesterase variant has free-energy correlations for substrate turnover and 37. Grifman, M., Galyam, N., Seidman, S. & Soreq, H. hematopoietic growth promoting activities. Mol. Med. inhibition by trifluoro ketone transition-state analogs. Functional redundancy of acetylcholinesterase and (in the press). Biochemistry 33, 8566–8576 (1994). neuroligin in mammalian neuritogenesis. Proc. Natl Acad. 59. Li, Y., Camp, S., Rachinsky, T. L., Bongiorno, C. & 15. Ripoll, D. R., Faerman, C. H., Axelsen, P. H., Silman, I. & Sci. USA 95, 13935–13940 (1998). Taylor, P. Promoter elements and transcriptional control of Sussman, J. L. An electrostatic mechanism for substrate 38. Bigbee, J. W., Sharma, K. V., Chan, E. L. & Bogler, O. the mouse acetylcholinesterase gene. J. Biol. Chem. guidance down the aromatic gorge of Evidence for the direct role of acetylcholinesterase in 268, 3563–3572 (1993). acetylcholinesterase. Proc. Natl Acad. Sci. USA 90, neurite outgrowth in primary dorsal root ganglion 60. Shapira, M. et al. A transcription-activating polymorphism 5128–5132 (1993). neurons. Brain Res. 861, 354–362 (2000). in the ACHE promoter associated with acute sensitivity to 16. Shafferman, A. et al. Electrostatic attraction by surface 39. de la Escalera, S., Bockamp, E. O., Moya, F., Piovant, M. anti-. Hum. Mol. Genet. 9, charge does not contribute to the catalytic efficiency of & Jimenez, F. Characterization and gene cloning of 1273–1281 (2000). acetylcholinesterase. EMBO J. 13, 3448–3455 (1994). neurotactin, a Drosophila transmembrane protein related 61. Chan, R. Y., Boudreau-Lariviere, C., Angus, L. M., 17. Radic, Z., Kirchhoff, P. D., Quinn, D. M., McCammon, to cholinesterases. EMBO J. 9, 3593–3601 (1990). Mankal, F. A. & Jasmin, B. J. An intronic enhancer J. A. & Taylor, P. Electrostatic influence on the kinetics of 40. Darboux, I., Barthalay, Y., Piovant, M. & Hipeau containing an N-box motif is required for synapse- and ligand binding to acetylcholinesterase. Distinctions Jacquotte, R. The structure–function relationships in tissue-specific expression of the acetylcholinesterase between active center ligands and fasciculin. J. Biol. Drosophila neurotactin show that cholinesterasic gene in skeletal muscle fibers. Proc. Natl Acad. Sci. USA Chem. 272, 23265–23277 (1997). domains may have adhesive properties. EMBO J. 15, 96, 4627–4632 (1999). 18. Taylor, P., Luo, Z. D. & Camp, S. in Cholinesterases and 4835–4843 (1996). 62. Rotundo, R. L. Nucleus-specific translation and assembly Cholinesterase Inhibitors (ed. Giacobini, E.) 63–79 (Martin 41. Sternfeld, M. et al. Acetylcholinesterase enhances neurite of acetylcholinesterase in multinucleated muscle cells. Dunitz, London, 2000). growth and synapse development through alternative J. Cell Biol. 110, 715–719 (1990). 19. Botti, S. A., Felder, C. E., Sussman, J. L. & Silman, I. contributions of its hydrolytic capacity, core protein, and 63. Galyam, N. et al. Complex host cell responses to Electrotactins: a class of adhesion proteins with variable C termini. J. Neurosci. 18, 1240–1249 (1998). antisense suppression of ACHE gene expression. conserved electrostatic and structural motifs. Protein 42. Scheiffele, P., Fan, J., Choih, J., Fetter, R. & Serafini, T. Antisense Nucl. Acid Drug Dev. 11, 51–57 (2001). Eng. 11, 415–420 (1998). Neuroligin expressed in nonneuronal cells triggers 64. Kaufer, D., Friedman, A., Seidman, S. & Soreq, H. Acute 20. Luo, Z. D., Camp, S., Mutero, A. & Taylor, P. Splicing of 5′ presynaptic development in contacting axons. Cell 101, stress facilitates long-lasting changes in cholinergic gene introns dictates alternative splice selection of 657–669 (2000). expression. Nature 393, 373–377 (1998). acetylcholinesterase pre-mRNA and specific expression 43. Song, J. Y., Ichtchenko, K., Sudhof, T. C. & Brose, N. 65. Nordberg, A., Hellstrom-Lindahl, E., Almkvist, O. & during myogenesis. J. Biol. Chem. 273, 28486–28495 Neuroligin 1 is a postsynaptic cell-adhesion molecule of Meurling, L. Activity of acetylcholinesterase in CSF (1998). excitatory synapses. Proc. Natl Acad. Sci. USA 96, increases in Alzheimer’s patients after treatment with 21. Atanasova, E., Chiappa, S., Wieben, E. & Brimijoin, S. 1100–1105 (1999). . Alzheimer’s Reports 2, 347–352 (1999). Novel messenger RNA and alternative promoter for 44. Llinas, R. R. & Greenfield, S. A. On-line visualization of 66. Grisaru, D. et al. Human osteogenesis involves murine acetylcholinesterase. J. Biol. Chem. 274, dendritic release of acetylcholinesterase from mammalian differentiation-dependent increases in the 21078–21084 (1999). substantia nigra neurons. Proc. Natl Acad. Sci. USA 84, morphogenically active 3′ alternative splicing variant of 22. Li, Y., Camp, S., Rachinsky, T. L., Getman, D. & Taylor, P. 3047–3050 (1987). acetylcholinesterase. Mol. Cell. Biol. 19, 788–795 (1999). Gene structure of mammalian acetylcholinesterase. 45. Holmes, C., Jones, S. A., Budd, T. C. & Greenfield, S. A. 67. Schober, A. et al. Reduced acetylcholinesterase (AChE) Alternative exons dictate tissue-specific expression. Non-cholinergic, trophic action of recombinant activity in adrenal medulla and loss of sympathetic J. Biol. Chem. 266, 23083–23090 (1991). acetylcholinesterase on mid-brain dopaminergic neurons. preganglionic neurons in TrkA-deficient, but not TrkB- 23. Bon, S., Coussen, F. & Massoulie, J. Quaternary J. Neurosci. Res. 49, 207–218 (1997). deficient, mice. J. Neurosci. 17, 891–903 (1997). associations of acetylcholinesterase. II. The polyproline 46. Inestrosa, N. C. et al. Acetylcholinesterase accelerates 68. Robertson, R. T. et al. Do subplate neurons comprise a attachment domain of the collagen tail. J. Biol. Chem. assembly of amyloid-β peptides into Alzheimer’s fibrils: transient population of cells in developing neocortex of 272, 3016–3021 (1997). possible role of the peripheral site of the enzyme. Neuron rats? J. Comp. Neurol. 426, 632–650 (2000). 24. Donger, C. et al. Mutation in the human 16, 881–891 (1996). 69. Shohami, E. et al. Antisense prevention of neuronal acetylcholinesterase-associated collagen gene, COLQ, is 47. Paoletti, F., Mocali, A. & Vannucchi, A. M. damages following head injury in mice. J. Mol. Med. 78, responsible for congenital myasthenic syndrome with Acetylcholinesterase in murine erythroleukemia (Friend) 228–236 (2000). end-plate acetylcholinesterase deficiency (Type Ic). Am. cells: evidence for megakaryocyte-like expression and 70. Gray, R., Rajan, A. S., Radcliffe, K. A., Yakehiro, M. & J. Hum. Genet. 63, 967–975 (1998). potential growth-regulatory role of enzyme activity. Blood Dani, J. A. Hippocampal synaptic transmission enhanced 25. Futerman, A. H., Low, M. G., Ackermann, K. E., 79, 2873–2879 (1992). by low concentrations of nicotine. Nature 383, 713–716 Sherman, W. R. & Silman, I. Identification of covalently 48. Lev-Lehman, E., Deutsch, V., Eldor, A. & Soreq, H. (1996). bound inositol in the hydrophobic membrane-anchoring Immature human megakaryocytes produce nuclear- 71. Ashani, Y. et al. Butyrylcholinesterase and domain of Torpedo acetylcholinesterase. Biochem. associated acetylcholinesterase. Blood 89, 3644–3653 acetylcholinesterase prophylaxis against Biophys. Res. Commun. 129, 312–317 (1985). (1997). poisoning in mice. Biochem. Pharmacol. 41, 37–41 26. Massoulie, J. et al. The polymorphism of 49. Kawashima, K. & Fujii, T. Extraneuronal cholinergic (1991). acetylcholinesterase: post-translational processing, system in lymphocytes. Pharmacol. Ther. 86, 29–48 72. Evans, W. E. & Relling, M. V. Pharmacogenomics: quaternary associations and localization. Chem. Biol. (2000). translating functional genomics into rational therapeutics. Interact 119–120, 29–42 (1999). 50. Soreq, H. et al. Antisense oligonucleotide inhibition of Science 286, 487–491 (1999). 27. Fitzpatrick-McElligott, S. & Stent, G. S. Appearance and acetylcholinesterase gene expression induces progenitor 73. Luo, Z. D. et al. Calcineurin enhances localization of acetylcholinesterase in embryos of the cell expansion and suppresses hematopoietic apoptosis acetylcholinesterase mRNA stability during C2-C12 leech Helobdella triserialis. J. Neurosci. 1, 901–907 ex vivo. Proc. Natl Acad. Sci. USA 91, 7907–7911 muscle cell differentiation. Mol. Pharmacol. 56, 886–894 (1981). (1994). (1999). 28. Betz, H., Bourgeois, J. P. & Changeux, J. P. Evolution of 51. Brown, L. M. et al. exposure and other 74. Giacobini, E. in Cholinesterases and Cholinesterase cholinergic proteins in developing slow and fast skeletal agricultural risk factors for leukemia among men in Inhibitors (ed. Giacobini, E.) 181–226 (Martin Dunitz, muscles in chick embryo. J. Physiol. 302, 197–218 Iowa and Minnesota. Cancer Res. 50, 6585–6591 London, 2000). (1980). (1990). 75. Erb, C. et al. Compensatory mechanisms facilitate 29. Layer, P. G. Cholinesterases preceding major tracts in 52. Lev-Lehman, E. et al. Synaptogenesis and myopathy hippocampal acetylcholine release in transgenic mice vertebrate neurogenesis. Bioessays 12, 415–420 (1990). under acetylcholinestrase overexpression. J. Mol. expressing human acetylcholinesterase. J. Neurochem. 30. Kreutzberg, G. W. Neuronal dynamics and axonal flow. IV. Neurosci. 14, 93–105 (2000). (in the press). Blockage of intra-axonal enzyme transport by colchicine. 53. Lapidot-Lifson, Y. et al. Coamplification of human 76. Pollet, C. et al. Medical evaluation of Persian Gulf Proc. Natl Acad. Sci. USA 62, 722–728 (1969). acetylcholinesterase and butyrylcholinesterase genes in veterans with fatigue and/or chemical sensitivity. J. Med. 31. Appleyard, M. E. Secreted acetylcholinesterase: non- blood cells: correlation with various leukemias and 29, 101–113 (1998). classical aspects of a classical enzyme. Trends Neurosci. abnormal megakaryocytopoiesis. Proc. Natl Acad. Sci. 77. Okumura, T. et al. Report on 640 victims of the Tokyo 15, 485–490 (1992). USA 86, 4715–4719 (1989). subway sarin attack. Ann. Emerg. Med. 28, 129–135 32. Small, D. H. Non-cholinergic actions of 54. Stephenson, J., Czepulkowski, B., Hirst, W. & Mufti, G. (1996). acetylcholinesterases: proteases regulating cell growth Deletion of the acetylcholinesterase locus at 7q22 78. Ohno, K., Brengman, J., Tsujino, A. & Engel, A. G. and development? Trends Biochem. Sci. 15, 213–216 associated with myelodysplastic syndromes (MDS) and Human endplate acetylcholinesterase deficiency caused (1990). acute myeloid leukaemia (AML). Leuk. Res. 20, 235–241 by mutations in the collagen-like tail subunit (ColQ) of the 33. Checler, F., Grassi, J. & Vincent, J. P. Cholinesterases (1996). asymmetric enzyme. Proc. Natl Acad. Sci. USA 95, display genuine arylacylamidase activity but are totally 55. Velan, B. et al. N-glycosylation of human 9654–9659 (1998). devoid of intrinsic peptidase activities. J. Neurochem. 62, acetylcholinesterase: effects on activity, stability and 79. Fournier, D. et al. Drosophila melanogaster 756–763 (1994). biosynthesis. Biochem. J. 296, 649–656 (1993). acetylcholinesterase gene. Structure, evolution and 34. Layer, P. G., Weikert, T. & Alber, R. Cholinesterases 56. Chan, R. Y., Adatia, F. A., Krupa, A. M. & Jasmin, B. J. mutations. J. Mol. Biol. 210, 15–22 (1989). regulate neurite growth of chick nerve cells in vitro by Increased expression of acetylcholinesterase T and R 80. Prody, C. A., Dreyfus, P., Zamir, R., Zakut, H. & Soreq, H. means of a non-enzymatic mechanism. Cell Tissue Res. transcripts during hematopoietic differentiation is De novo amplification within a ‘silent’ human 273, 219–226 (1993). accompanied by parallel elevations in the levels of their cholinesterase gene in a family subjected to prolonged 35. Small, D. H., Reed, G., Whitefield, B. & Nurcombe, V. respective molecular forms. J. Biol. Chem. 273, exposure to organophosphorous insecticides. Proc. Natl Cholinergic regulation of neurite outgrowth from isolated 9727–9733 (1998). Acad. Sci. USA 86, 690–694 (1989). chick sympathetic neurons in culture. J. Neurosci. 15, 57. Camp, S. & Taylor, P. in Structure and Function of 81. Bartels, C. F., Zelinski, T. & Lockridge, O. Mutation at 144–151 (1995). Cholinesterases and Related Proteins (eds Doctor, B. P., codon 322 in the human acetylcholinesterase (ACHE)

NATURE REVIEWS | NEUROSCIENCE VOLUME 2 | APRIL 2001 | 301 © 2001 Macmillan Magazines Ltd PERSPECTIVES

gene accounts for YT blood group polymorphism. Am. J. intensifies neurodeterioration correlates. Proc. Natl Acad. 114. Haley, R. W., Kurt, T. L. & Hom, J. Is there a Gulf Hum. Genet. 52, 928–936 (1993). Sci. USA 97, 8647–8652 (2000). War Syndrome? Searching for syndromes by factor 82. Ehrlich, G. et al. Population diversity and distinct 97. Chien, C. T., Bartel, P. L., Sternglanz, R. & Fields, S. The analysis of symptoms. J. Am. Med. Assoc. 277, haplotype frequencies associated with ACHE and BCHE two-hybrid system: a method to identify and clone genes 215–222 (1997). genes of Israeli Jews from trans-Caucasian Georgia and for proteins that interact with a protein of interest. Proc. 115. Sapolsky, R. M. The stress of Gulf War syndrome. Nature from Europe. Genomics 22, 288–295 (1994). Natl Acad. Sci. USA 88, 9578–9582 (1991). Med. 393, 308–309 (1998). 83. La Du, B. N. et al. Phenotypic and molecular biological 98. Nissim, A. et al. Antibody fragments from a ‘single pot’ 116. Friedman, A. et al. Pyridostigmine brain penetration analysis of human butyrylcholinesterase variants. Clin. phage display library as immunochemical reagents. under stress enhances neuronal excitability and induces Biochem. 23, 423–431 (1990). EMBO J. 13, 692–698 (1994). early immediate transcriptional response. Nature Med. 2, 84. Beeri, R. et al. Transgenic expression of human 99. Mayford, M., Abel, T. & Kandel, E. R. Transgenic 1382–1385 (1996). acetylcholinesterase induces progressive cognitive approaches to cognition. Curr. Opin. Neurobiol. 5, 117. Grauer, E., Alkalai, D., Kapon, J., Cohen, G. & Raveh, L. deterioration in mice. Curr. Biol. 5, 1063–1071 (1995). 141–148 (1995). Stress does not enable pyridostigmine to inhibit brain 85. Xie, W. et al. Postnatal developmental delay and 100. Young, R. A. Biomedical discovery with DNA arrays. Cell cholinesterase after parenteral administration. Toxicol. supersensitivity to organophosphate in gene-targeted 102, 9–15 (2000). Appl. Pharmacol. 164, 301–304 (2000). mice lacking acetylcholinesterase. J. Pharmacol. Exp. 101. Mezey, E., Chandross, K. J., Harta, G., Maki, R. A. & 118. Lallement, G. et al. Heat stress, even extreme, does not Ther. 293, 896–902 (2000). McKercher, S. R. Turning blood into brain: cells bearing induce penetration of pyridostigmine into the brain of 86. von der Kammer, H. et al. Muscarinic acetylcholine neuronal generated in vivo from bone marrow. pigs. Neurotoxicology 19, 759–766 (1998). receptors activate expression of the EGR gene family of Science 290, 1779–1782 (2000). 119. Esposito, P. et al. Acute stress increases permeability of transcription factors. J. Biol. Chem. 273, 14538–14544 102. Kuhl, D. E. et al. In vivo mapping of cerebral the blood–brain-barrier through activation of brain mast (1998). acetylcholinesterase activity in aging and Alzheimer’s cells. Brain Res. 888, 117–127 (2001). 87. Xie, J. & McCobb, D. P. Control of alternative splicing of disease. Neurology 52, 691–699 (1999). 120. Grauer, E. et al. Viral neuroinvasion as a marker for BBB potassium channels by stress hormones. Science 280, 103. Shinotoh, H. et al. Positron emission tomographic integrity following exposure to cholinesterase inhibitors. 443–446 (1998). measurement of acetylcholinesterase activity reveals Life Sci. 68, 985–990 (2001). 88. Daoud, R., Da Penha Berzaghi, M., Siedler, F., Hubener, differential loss of ascending cholinergic systems in 121. Relyea, R. A. & Mills, N. Predator-induced stress makes M. & Stamm, S. Activity-dependent regulation of Parkinson’s disease and progressive supranuclear palsy. the pesticide more deadly to gray treefrog alternative splicing patterns in the rat brain. Eur. J. Ann. Neurol. 46, 62–69 (1999). tadpoles (Hyla versicolor). Proc. Natl Acad. Sci. USA 98, Neurosci. 11, 788–802 (1999). 104. Vereker, E., O’Donnell, E. & Lynch, M. A. The inhibitory 2491–2496 (2001). 89. Combes, D., Fedon, Y., Grauso, M., Toutant, J. P. & effect of interleukin-1β on long-term potentiation is 122. Coyle, J. T., Price, D. L. & DeLong, M. R. Alzheimer’s Arpagaus, M. Four genes encode acetylcholinesterases coupled with increased activity of stress-activated protein disease: a disorder of cortical cholinergic innervation. in the nematodes Caenorhabditis elegans and kinases. J. Neurosci. 20, 6811–6819 (2000). Science 219, 1184–1190 (1983). Caenorhabditis briggsae. cDNA sequences, genomic 105. Xu, L., Anwyl, R. & Rowan, M. J. Behavioural stress 123. Patrick, J. & Lindstrom, J. Autoimmune response to structures, mutations and in vivo expression. J. Mol. Biol. facilitates the induction of long-term depression in the . Science 180, 871–872 (1973). 300, 727–742 (2000). hippocampus. Nature 387, 497–500 (1997). 124. Keesey, J. C. Contemporary opinions about Mary 90. Anglister, L., Stiles, J. R. & Salpeter, M. M. 106. Hamilton, S. E. et al. Disruption of the m1 receptor gene Walker: a shy pioneer of therapeutic neurology. Acetylcholinesterase density and turnover number at frog ablates muscarinic receptor-dependent M current Neurology 51, 1433–1439 (1998). neuromuscular junctions, with modeling of their role in regulation and seizure activity in mice. Proc. Natl Acad. 125. Beeri, R. et al. Enhanced hemicholinium binding and synaptic function. Neuron 12, 783–794 (1994). Sci. USA 94, 13311–13316 (1997). attenuated dendrite branching in cognitively impaired 91. Peng, H. B., Xie, H., Rossi, S. G. & Rotundo, R. L. 107. Gennari, K., Brunner, J. & Brodbeck, U. Tetrameric acetylcholinesterase-transgenic mice. J. Neurochem. Acetylcholinesterase clustering at the neuromuscular detergent-soluble acetylcholinesterase from human 69, 2441–2451 (1997). junction involves perlecan and dystroglycan. J. Cell Biol. caudate nucleus: subunit composition and number of 126. Alvarez, A. et al. Stable complexes involving 145, 911–921 (1999). active sites. J. Neurochem. 49, 12–18 (1987). acetylcholinesterase and amyloid-β peptide change the 92. Seidman, S. et al. Synaptic and epidermal accumulations 108. Franklin, R. B. J. & Paxinos, G. The Mouse Brain in biochemical properties of the enzyme and increase the of human acetylcholinesterase are encoded by alternative Stereotaxic Coordinates (Academic, San Diego, 1997). of Alzheimer’s fibrils. J. Neurosci. 18, 3′-terminal exons. Mol. Cell. Biol. 15, 2993–3002 (1995). 109. Holmstedt, B. in Cholinesterases and Cholinesterase 3213–3223 (1998). 93. Harlow, M. L., Ress, D., Stoschek, A., Marshall, R. M. & Inhibitors: Basic, Preclinical and Clinical Aspects (ed. 127. Sberna, G. et al. Acetylcholinesterase is increased in the McMahan, U. J. The architecture of material Giacobini, E.) 1–8 (Martin Dunitz, London, 2000). brains of transgenic mice expressing the C-terminal at the frog’s neuromuscular junction. Nature 409, 110. Abramson, S. N., Radic, Z., Manker, D., Faulkner, D. J. & fragment (CT100) of the β-amyloid protein precursor 479–484 (2001). Taylor, P. Onchidal: a naturally occurring irreversible of Alzheimer’s disease. J. Neurochem. 71, 723–731 94. Feng, G. et al. Genetic analysis of collagen Q: roles in inhibitor of acetylcholinesterase with a novel mechanism (1998). acetylcholinesterase and butyrylcholinesterase assembly of action. Mol. Pharmacol. 36, 349–354 (1989). 128. Wilson, B. W. & Viola, G. A. Multiple forms of and in synaptic structure and function. J. Cell Biol. 144, 111. Carmichael, W. The toxins of cyanobacteria. Sci. Am. acetylcholinesterase in nutritional and inherited muscular 1349–1360 (1999). 270, 78–86 (1994). dystrophy of the chicken. J. Neurol. Sci. 16, 183–192 95. Karpel, R. et al. Overexpression of alternative human 112. Wilson, I. B. Molecular complementarity and antidotes for (1972). acetylcholinesterase forms modulates process alkylphosphate poisoning. Fed. Proc. 18, 752–758 129. Silman, I., di Giamberardino, L., Lyles, L., Couraud, J. Y. extensions in cultured glioma cells. J. Neurochem. 66, (1959). & Barnard, E. A. Parallel regulation of 114–123 (1996). 113. Keeler, J. R., Hurst, C. G. & Dunn, M. A. Pyridostigmine acetylcholinesterase and pseudocholinesterase in 96. Sternfeld, M. et al. Excess ‘readthrough’ used as a pretreatment under wartime normal, denervated and dystrophic chicken skeletal acetylcholinesterase attenuates but the ‘synaptic’ variant conditions. J. Am. Med. Assoc. 266, 693–695 (1991). muscle. Nature 280, 160–162 (1979).

We welcome correspondence

Has something in the journal caught your attention?

If so, please write to us about it by sending an email to: [email protected] and flag it for the attention of the Nature Reviews Neuroscience editors.

Correspondence to the journal will be selected by the editors for publication on the Nature Reviews Neuroscience Website at http://www.nature.com/reviews/neuroscience/ where it will be linked to the relevant article.

302 | APRIL 2001 | VOLUME 2 www.nature.com/reviews/neuro © 2001 Macmillan Magazines Ltd