Human acetylcholinesterase. Immunochemical studies with monoclonal antibodies

Different glycosylation in acetylcholinesterases from mammalian brain and erythrocytes

Acetylcholinesterases (EC 3.1.1.7, AChE) have varying amounts of carbohydrates attached to the core protein. Sequence analysis of the known primary structures gives evidence for several asparagine-linked carbohydrates. From the differences in molecular mass determined on sodium dodecyl sulfate-polyacrylamide gel before and after deglycosylation with N-glycosidase F (EC 3.2.2.18), it is seen that dimeric AChE from red cell membranes is more heavily glycosylated than the tetrameric brain enzyme. Furthermore, dimeric and tetrameric forms of bovine AChE are more heavily glycosylated than the corresponding human enzymes. Monoclonal antibodies 2E6, 1H11, and 2G8 raised against detergent-soluble AChE from electric organs of Torpedo nacline timilei as well as Elec-39 raised against AChE from Electrophorus electricus cross-reacted with AChE from bovine and human brain but not with AChE from erythrocytes. Treatment of the enzyme with N-glycosidase F abolished binding of monoclonal antibodies, suggesting that the epitope, or part of it, consists of N-linked carbohydrates. Analysis of N-acetylglucosamine sugars revealed the presence of N-acetylglucosamine in all forms of cholinesterases investigated, giving evidence for N-linked glycosylation. On the other hand, N-acetylgalactosamine was not found in AChE from human and bovine brain or in butyrylcholinesterase (EC 3.1.1.8) from human serum, indicating that these forms of cholinesterase did not contain O-linked carbohydrates. Despite the notion that within one species, the different forms of AChE arise from one gene by different splicing, our present results show that dimeric erythrocyte and tetrameric brain AChE must undergo different postsynthetic modifications leading to differences in their glycosylation patterns.

Paroxysmal nocturnal hemoglobinuria: erythrocyte acetylcholinesterase deficit analyzed by immunoassay and fluorescence-activated sorting

An immunodisplacement assay based on a specific, solid-phase monoclonal antibody was designed to measure acetylcholinesterase in tissue extracts. Sample enzyme content was determined from the competitive reduction of binding of a purified acetylcholinesterase standard, with a detection limit of 5 ng or less. Washed erythrocyte membranes from six normal subjects averaged 1.8 units of acetylcholinesterase activity and 0.45 microgram of acetylcholinesterase content per milligram of total protein. Enzyme activity and content in samples from three patients with paroxysmal nocturnal hemoglobinuria (PNH) were reduced approximately in parallel, by as much as 70%. The residual cholinesterase had almost the same homospecific activity as the normal enzyme and was bound with equivalent affinity by six different antibodies. Therefore, the cholinesterase defect was dominated by enzyme loss rather than by structural abnormalities affecting enzyme function. Fluorescence-activated sorting of antibody-labeled erythrocytes revealed a bimodal population distribution. Up to 66% of the PNH cells lacked cholinesterase, and the rest had a near-normal enzyme content. Inasmuch as enzyme-deficient cells represent the complement-sensitive population, cell sorting may help in assessing clinical status and, perhaps, in developing new therapeutic modalities for PNH.

Monoclonal antibodies to rat brain acetylcholinesterase: comparative affinity for soluble and membrane-associated enzyme and for enzyme from different vertebrate species

Seven unique monoclonal antibodies were generated to rat brain acetylcholinesterase. Upon density gradient ultracentrifugation, immunoglobulin complexes with the monomeric enzyme appeared as single peaks of acetylcholinesterase activity with a sedimentation coefficient approximately 3S greater than that of the free enzyme. This behavior is consistent with the assumption of one binding site per enzyme molecule. Apparent dissociation constants of these antibodies for rat brain acetylcholinesterase calculated on the basis of this assumption ranged from about 10 nM to more than 1,000 nM. Some of the antibodies were less able to bind the membrane-associated enzyme that required detergent for solubilization than the naturally soluble acetylcholinesterase of detergent-free brain extracts. Species cross-reactivity was investigated with crude brain extracts from mammals (human, mouse, rabbit, guinea pig, cow, and cat) and from other vertebrates (chicken, frog, and electric eel). Three antibodies bound rat acetylcholinesterase exclusively; one had nearly the same affinity for all mammalian acetylcholinesterases investigated; the remaining three showed irregular binding patterns. None of the antibodies recognized frog and electric eel enzyme. Pooled antibody was found to be suitable for specific immunofluorescence staining of large neurons in the ventral horn of the rat spinal cord and smaller cells in the caudate nucleus. Other potential applications of these antibodies are discussed.