Cholinergic system and cell proliferation.

The cholinergic system, comprising acetylcholine, the proteins responsible for acetylcholine synthesis and release, acetylcholine receptors and cholinesterases, is expressed by most human cell types. Acetylcholine is a neurotransmitter, but also a local signalling molecule which regulates basic cell functions, and cholinergic responses are involved in cell proliferation and apoptosis. So, activation of nicotinic and muscarinic receptors has a proliferative and anti-apoptotic effect in many cells. The content of choline acetyltransferase, acetylcholine receptors and cholinesterases is altered in many tumours, and cholinesterase content correlates with patient survival in some cancers. During apoptosis, acetylcholinesterase is induced and appears in the nuclei. Acetylcholinesterase participates in the regulation of cell proliferation and apoptosis through hydrolysis of acetylcholine and by other catalytic and non catalytic mechanisms, in a variant-specific manner. This review gathers information on the role of cholinergic system and specially acetylcholinesterase in cell proliferation and apoptosis.

The AChE membrane-binding tail PRiMA is down-regulated in muscle and nerve of mice with muscular dystrophy by merosin deficiency.

Since Duchenne muscular dystrophy was attributed to mutations in the dystrophin gene, more than 30 genes have been found to be causally related with muscular dystrophies, about half of them encoding proteins of the dystrophin-glycoprotein complex (DGC). Through laminin-2, the DGC bridges the muscle cytoskeleton and the extracellular matrix. Decreased levels of PRiMA-linked acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) have been observed in dystrophic muscle and nerve of dystrophin-deficient (mdx) and laminin-2 deficient (Lama2dy) mice. To help explain these observations, the relative content of AChE, BuChE and PRiMA mRNAs were compared in normal and Lama2dy mouse muscle and sciatic nerve. The 17-fold lower level of PRiMA mRNA in Lama2dy muscle explained the deficit in PRiMA-linked ChEs. This would increase acetylcholine availability and, eventually, the desensitization of nicotinic receptors. Abnormal development of the Schwann cells led to peripheral neuropathy in the Lama2dy mouse. Compared with normal nerve, dystrophic nerve displayed 4-fold less AChE-T mRNA, 3-fold more BuChE mRNA and 2.5-fold less PRiMA mRNA, which agreed with the lower AChE activity in dystrophic nerve, its increased BuChE activity and the specific drop in PRiMA-linked BuChE. The widely accepted role of glial cells as the source of BuChE, the observed dysmyelination of Lama2dy nerve and its increased BuChE activity support the idea that BuChE up-regulation is related with the aberrant differentiation of the Schwann cells.

Targeting of acetylcholinesterase to lipid rafts of muscle.

Despite the great progress made in setting the basis for the molecular diversity of acetylcholinesterase (AChE), an explanation for the existence of two types of amphiphilic subunits, with and without glicosylphosphatidylinositol (GPI) (Types I and II), has not been provided yet. In searching whether, as for the deficiency of dystrophin, that of merosin (laminin-alpha2 chain) alters the number of caveolae in muscle, a high increase in caveolin-3 (Cav3) was observed in the Triton X-100-resistant membranes (TRM) isolated from muscle of merosin-deficient dystrophic mice (Lama2dy). The rise in Cav3 was accompanied by that of non-caveolar lipid rafts, as showed by the greater ecto-5'-nucleotidase (eNT) activity, a marker of non-caveolar rafts, in TRM of dystrophic muscle. The observation of AChE activity in TRM, the increased levels of rafts and raft-bound AChE activity in merosin-deficient muscle and the presence of phospholipase C-sensitive AChE dimers in TRM supported targeting of glypiated AChE to rafts. This issue and the involvement of TRM in conveying nicotinic receptors to the neuromuscular junction and particular muscarinic receptors to cardiac sarcolemma strongly support a role for lipid rafts in targeting ACh receptors and glypiated AChE. Their nearby location in the surface membrane may provide cells with a fine tuning for regulating cholinergic responses.

The level of aryl acylamidase activity displayed by human butyrylcholinesterase depends on its molecular distribution.

Butyrylcholinesterase (BuChE) and acetylcholinesterase (AChE) display both esterase and aryl acylamidase (AAA) activities. Their AAA activity can be measured using o-nitroacetanilide (ONA). In human samples depleted of acetylcholinesterase, we noticed that the ratio of amidase to esterase activities varied depending on the source, despite both activities being due to BuChE. Searching for an explanation, we compared the activities of BuChE molecular forms in samples of human colon, kidney and serum, and observed that BuChE monomers (G(1)) hydrolyzed o-nitroacetanilide much faster than tetramers (G(4)). This fact suggested that association might cause differences in the AAA site between single and polymerized subunits. This and other post-translational modifications in BuChE subunits probably determine their level of AAA activity. The higher amidase activity of monomers could justify the presence of single BuChE subunits in cells as a way to preserve the AAA activity of BuChE, which could be lost by oligomerization.

Cholinesterases are down-expressed in human colorectal carcinoma.

The aberrations of cholinesterase (ChE) genes and the variation of ChE activity in cancerous tissues prompted us to investigate the expression of ChEs in colorectal carcinoma. The study of 55 paired specimens of healthy (HG) and cancerous gut (CG) showed that acetylcholinesterase (AChE) activity fell by 32% and butyrylcholinesterase (BuChE) activity by 58% in CG. Abundant AChE-H, fewer AChE-T, and even fewer AChE-R and BuChE mRNAs were observed in HG, and their content was greatly diminished in CG. The high level of the AChE-H mRNA explains the abundance of AChE-H subunits in HG, which as glycosylphosphatidylinositol (GPI)-anchored amphiphilic AChE dimers (G2(A)) and monomers (G1(A)) account for 69% of AChE activity. The identification of AChE-T and BuChE mRNAs justifies the occurrence in gut of A12, G4(H) and PRiMA-containing G4(A) AChE forms, besides G4(H), G4(A) and G1(H) BuChE. The down-regulation of ChEs might contribute to gut carcinogenesis by increasing acetylcholine availability and over-stimulating muscarinic receptors.

Muscular dystrophy with laminin deficiency decreases acetylcholinesterase activity in thymus of dystrophic Lama2dy mice.

We have studied the effect of muscular dystrophy by merosin deficiency on mouse thymus acetyl- (AChE) and butyrylcholinesterase (BuChE). The organ contains AChE and BuChE activities. Merosin deficiency causes an important decrease (46%) in AChE specific activity. Thymus produces dimers, monomers and tetramers of AChE, and the three kinds of AChE mRNAs. The drop in AChE activity in dystrophic animals could affect the amount of ACh reaching cholinergic receptors in cells of lymphoid organs.

Identification of inactive ecto-5'-nucleotidase in normal mouse muscle and its increased activity in dystrophic Lama2(dy) mice.

Ecto-5'-nucleotidase (eNT) activity and protein in normal (NM) and merosin-deficient dystrophic (DM) Lama2(dy) mice muscle were studied. eNT activity in DM was three- to four-fold that in NM. eNT in NM and DM displayed the same kinetic properties. Slot and Western blotting revealed that the immunoreactive protein was two to three times more abundant in control muscle, when NM and DM samples with the same eNT activity were compared, indicating that mouse muscle contains catalytically inactive eNT components. eNT activity and protein peaks coincided in sedimentation analyses, revealing that inactive eNT occurs as dimers. Most eNT activity and protein of NM bound to Lens culinaris (LCA) or Ricinus communis (RCA) agglutinins, but half of the activity and one-third of the protein bound to wheat germ agglutinin (WGA). Although WGA interaction did not permit full separation of inactive eNT, the results suggest that similar proportions of active species with and without WGA reactivity occur in mouse muscle, whereas a great fraction of the inactive eNT variants lack WGA reactivity. Because the level of eNT protein was little modified in DM, the higher eNT activity in dystrophic than in control muscle may result from misregulation in the synthesis of active and inactive eNT species or from conversion of inactive into active components.

Muscular dystrophy alters the processing of light acetylcholinesterase but not butyrylcholinesterase forms in liver of Lama2(dy) mice.

In order to know whether the histopathological changes of liver, which accompany muscular dystrophy, affect the synthesis of cholinesterases, the distribution and glycosylation of acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) forms in normal (NL) and dystrophic Lama2(dy) mouse liver (DL) were investigated. About half of liver AChE, and 25% of BuChE were released with a saline buffer (fraction S(1)), and the rest with a saline-Brij 96 buffer (S(2)). Abundant light (G(2)(A) and G(1)(A)) AChE (87%) and BuChE (93%) forms, and a few G(4)(H) and G(4)(A) ChE species were identified in liver. The dystrophic syndrome had no effect on solubilization or composition of ChE forms. Most of the light AChE and BuChE species (>95%) were bound by octyl-Sepharose, while most light AChE forms (80%), but not BuChE isoforms (15%), were retained in phenyl-agarose. About half of the AChE dimers lost their amphiphilic anchor with phosphatidylinositol-specific phospholipase C (PIPLC), and the fraction of PIPLC-resistant species increased in DL. AChE T and R transcripts were detected by reverse transcriptase-polymerase chain reaction (RT-PCR) of liver RNA. ChE components of liver, erythrocyte, and plasma were distinguished by their amphiphilic properties and interaction with lectins. The dystrophic syndrome increased the liver content of the light AChE forms with Lens culinaris agglutinin (LCA) reactivity. The abundance of ChE tetramers in plasma and their small amount in liver suggest that after their assembly in liver they are rapidly secreted, while the light species remain associated to hepatic membranes.

The ecto-5'-nucleotidase subunits in dimers are not linked by disulfide bridges but by non-covalent bonds.

It has long been considered that ecto-5'-nucleotidase (eNT) dimers consist of subunits linked by disulfide bonds. Hydrophilic (6.7S) and amphiphilic (4.0S) dimers were separated by sedimentation analysis of eNT purified from bull seminal plasma. Hydrophilic (4. 2S) and amphiphilic (2.6S) eNT monomers were obtained after reduction of disulfide bonds in dimers. The amphiphilic eNT dimers or monomers were converted into their hydrophilic variants with phosphatidylinositol-specific phospholipase C. SDS-PAGE plus Western blot showed 68 kDa subunits, regardless of the addition of beta-mercaptoethanol to the SDS mixture. Active eNT monomers were obtained by addition of 1 M guanidinium chloride (Gdn) to dimers, and unfolded subunits by addition of 4 M Gdn. The results unambiguously demonstrate that the subunits in eNT dimers are not linked by disulfide bridges, but by non-covalent bonds, and that dissociation precedes inactivation and unfolding.

Increased butyrylcholinesterase levels in microsomal membranes of dystrophic Lama2dy mouse muscle.

The proportions and the glycosylation of butyrylcholinesterase (BuChE) forms in vesicles rich in sarcoplasmic reticulum from normal (NMV) and dystrophic (DMV) muscle were analyzed, using merosin-deficient dystrophic mice. BuChE activity in DMV was two- to threefold that in NMV. Globular amphiphilic G1A, G2A, and G4A and hydrophilic G4H BuChE forms were identified in NMV and DMV. The amount of G2A forms increased sevenfold in DMV, and the other forms increased about twofold. The higher BuChE level in DMV might reflect a maturational defect, with dystrophy preventing the down-regulation of BuChE with muscle development. About half of G1A, G2A, and G4H BuChE forms in NMV or DMV bound to Lens culinaris agglutinin (LCA), a higher fraction to wheat germ agglutinin (WGA), and little to Ricinus communis agglutinin (RCA). Most of the G4A forms in NMV or DMV bound to LCA or WGA; those from NMV failed to bind to RCA, whereas most of the variants in DMV bound to it, suggesting that the excess of tetramers in DMV is mainly RCA-reactive. The differential interaction of lectins with BuChE components from muscle microsomes, serum, and nerves confirmed that the microsomal BuChE was muscle-intrinsic. The results provide clues regarding the alterations that dystrophy produces in the biosynthesis of BuChE forms in muscle.

Characterization of acetylcholinesterase and butyrylcholinesterase forms in normal and dystrophic Lama2dy mouse heart.

In searching for possible differences in acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) forms of dystrophic heart, the properties of ChE species in normal (NH) and dystrophic Lama2dy mouse heart (DH) were investigated. BuChE predominated over AChE. Loosely- and tightly-bound ChEs were released with saline (extract S1) and saline-Triton X-100 buffers (S2). About 50% of AChE, and 25% of BuChE, in NH or DH was measured in S1, and the rest in S2. Asymmetric AChE forms A12 (15%) and A8 (11%), globular hydrophilic G(H)4 (8%), amphiphilic G(A)4 (15%), and G(A)2+G(A)1 (51%) AChE species, and BuChE forms G(H)4 (13%), G(A)4 (3%), and G(A)2+G(A)1 (84%) were identified in NH and DH. Most of the asymmetric and G(A)4 AChE species were bound to Triticum vulgaris (WGA) or Ricinus communis (RCA) agglutinins. About half of G(H)4 and G(A)2+G(A)1 AChE were bound to WGA, and less (10%) to RCA. Variable amounts of G(H)4+G(A)4 (60%), and G(A)2+G(A)1 (75%) BuChE bound to WGA, and 50 and 10% to RCA. The lack of structural differences between ChE species in NH and DH indicates that, in contrast to the ChE forms in mouse skeletal muscle, the biosynthesis of ChE components in heart is not disturbed by dystrophy.

Biochemical properties of 5'-nucleotidase from mouse skeletal muscle.

Ecto-5'-nucleotidase (eNT) from mouse muscle has been purified after extraction with detergent followed by chromatography on concanavalin A- and AMP-Sepharose. Three fractions were recovered: UF was NT non-retained in immobilised AMP; F-I was bound enzyme eluted with beta-glycerophosphate, and F-II was bound NT released with AMP. eNT was 80000-fold purified in F-II, this fraction showing proteins of 74, 68 and 51 kDa after immunoblotting. NT in UF migrated at 6.7S after centrifugation in sucrose gradients with Triton X-100, the peak being split into two of 6.7S and 4.4S in gradients with Brij 96. Ecto-NT in F-I or F-II migrated at 5.8S in Triton X-100-, or 4.4S in Brij 96-containing gradients. The hydrodynamic behaviour, concentration in Triton X-114, binding to phenyl-agarose, and sensitivity to phosphatidylinositol-specific phospholipase C revealed that enzyme forms in F-I or F-II were amphiphilic dimers with linked phosphatidylinositol residues, whilst most of NT forms in UF were hydrophilic dimers. A zinc/protein molar ratio of 2.2 was determined for eNT in F-II. NT activity was decreased in assays made in imidazole buffer, and was partly restored with 10 microM Zn2+ or 100 microM Mn2+. In assays with Tris buffer, NT showed a Km for AMP of 12 microM, and was competitively inhibited by ATP or ADP.

Glycosylation of cholinesterase forms in brain from normal and dystrophic Lama2dy mice.

Differences in the oligosaccharides attached to acetyl- (AChE) and butyrylcholinesterase (BuChE) forms in brain from control and merosin-deficient Lama2dy dystrophic mice were investigated by means of their interaction with agarose-immobilized lectins. Asymmetric AChE, hydrophilic and amphiphilic AChE and BuChE tetramers, and amphiphilic AChE and BuChE monomers were identified in brain. All ChE forms were strongly adsorbed to the lectins concanavalin A (Con A), Lens culinaris (LCA) and Triticum vulgaris (WGA), and poorly so to Ricinus communis agglutinin (RCA), suggesting that the oligosaccharides in AChE or BuChE subunits are similarly processed regardless of their state of polymerization. The lack of differences in the interaction of lectins with homologous AChE and BuChE forms in normal and dystrophic tissue indicates that, in contrast to ChEs forms in skeletal muscle, the dystrophic condition does not disturb the processing of the oligosaccarides of brain enzyme forms.

MeCP2 is a transcriptional repressor with abundant binding sites in genomic chromatin.

MeCP2 is an abundant mammalian protein that binds to methylated CpG. We have found that native and recombinant MeCP2 repress transcription in vitro from methylated promoters but do not repress nonmethylated promoters. Repression is nonlinearly dependent on the local density of methylation, becoming significant at the density found in bulk vertebrate genomic DNA. Transient transfection using fusions with the GAL4 DNA binding domain identified a region of MeCP2 that is capable of long-range repression in vivo. Moreover, MeCP2 is able to displace histone H1 from preassembled chromatin that contains methyl-CpG. These properties, together with the abundance of MeCP2 and the high frequency of its 2 bp binding site, suggest a role as a global transcriptional repressor in vertebrate genomes.

Glycosylation of acetylcholinesterase forms in microsomal membranes from normal and dystrophic Lama2dy mouse muscle.

The distribution and glycosylation of acetylcholinesterase (AChE) forms in vesicles derived from sarcoplasmic reticulum of normal muscle (NMV) were investigated and compared with those from dystrophic muscle vesicles (DMV). AChE activity was similar in NMV and DMV. Most of the AChE in NMV and half in DMV were released with Triton X-100. Asymmetric (A12) and globular hydrophilic and amphiphilic (G4H, G4A, G2A, and G1A) AChE species occurred in NMV and DMV, the lighter forms being predominant. The percentage of G4H and G4A decreased in DMV. A fraction of the AChE that could not be extracted with detergent was detached with collagenase. Most of the detergent-released A12 AChE from NMV and nearly half in DMV failed to bind to Ricinus communis agglutinin (RCA-I). Conversely, the collagenase-detached isoforms bound to RCA, revealing that asymmetric AChE associated with internal membranes or basal lamina differed in glycosylation. Moreover, nearly half of G4A AChE in DMV and a few in NMV bound to RCA. Most of the RCA-unreactive G4A forms in NMV come from sarcolemma. The results indicate that dystrophy induces minor changes in the distribution and glycosylation of AChE forms in internal membranes of muscle.

Differential interaction of the monoclonal antibody AE-1 with acetylcholinesterase oligomers and monomers from rabbit muscle microsomes, human brain and fetal bovine serum.

The monoclonal antibody AE-1 raised against acetylcholinesterase (AChE) from human erythrocytes (HE) is shown to react with active asymmetric and tetrameric AChE components from rabbit muscle microsomes (RMM), and with tetrameric forms from human brain (HB) or fetal bovine serum (FBS). However, it failed to bind to AChE monomers from RMM or HB. The results of Western blot revealed that the determinant for AE-1 consisted of a conformational domain, not a primary sequence region, in the AChE subunit. The antibody recognized HE monomers and FBS dimers, but not FBS monomers. The formation of labile immunocomplexes between AE-1 and AChE subunits may explain the lack of interaction between the antibody and the monomers from non-erythrocyte sources.

Binding of histone H1 to DNA is indifferent to methylation at CpG sequences.

The possibility that histone H1 binds preferentially to DNA containing 5-methylcytosine in the dinucleotide CpG is appealing, as it could help to explain the repressive effects of methylation on gene activity. In this study, the affinity of purified H1 for methylated and non-methylated DNA sequences has been tested using both naked DNA and chromatin. Based on a variety of assays (bandshifts, filter-binding assays, Southwestern blots, and nuclease sensitivity assays), we conclude that H1 has no significant preference for binding to naked methylated DNA. Similarly, H1 showed the same affinities for methylated and non-methylated DNA when assembled into chromatin in a Xenopus oocyte extract. Thus potential cooperative interaction of H1 with polynucleosomal complexes is not enhanced by the presence of DNA methylation.

Amphiphilic properties of molecular forms of acetylcholinesterase in normal and dystrophic muscle.

Acetylcholinesterase (AChE) molecular forms were studied in normal (NM) and in dystrophic (DM) 129B6F1/J mouse muscle. Successive extractions of the tissue with saline and saline-Triton X-100 buffers yielded two soluble fractions, S1 and S2. Forty percent of the AChE in NM was measured in S1 and 60% in S2, and 65% and 35%, respectively, in extracts from DM. A12, A8, G4, G2, and G1 forms of AChE were found in S1 and S2 from NM and DM. A similar content of asymmetric molecules was noticed between NM and DM. G4 AChE was a minor species in DM, and G1 and G2 AChE were more abundant in DM than in NM. The amphiphilic properties of the several molecules were assessed by Triton X-114 phase-partitioning and hydrophobic chromatography. Thirty and 70% of the enzyme in a mixture of S1 and S2 partitioned in the detergent-rich and in the detergent-poor phases, respectively, whether the extracts were obtained from NM or DM. Asymmetric and G4 AChE predominated in the aqueous phase and G1 and G2 in the detergent phase. Ten and 25% of the enzyme in S1 from NM or DM, respectively, was adsorbed to the phenyl-agarose. Elution of the retained enzyme followed by sedimentation analysis revealed that a certain amount of asymmetric and most of the G1 and G2 forms were associated with the matrix. The content of amphiphilic asymmetric and light globular forms was notably higher in DM than in NM. The results suggest that dystrophic muscle produces a specific pattern of molecular forms of AChE.

G4 forms of acetylcholinesterase and butyrylcholinesterase in normal and dystrophic mouse muscle differ in their interaction with Ricinus communis agglutinin.

Differences in glycosylation between molecular forms of acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) in muscle and serum of normal and dystrophic mice have been studied by means of their adsorption to immobilized lectins. Application of a two-step extraction procedure, first with saline buffer, and second with saline buffer and Triton X-100, brought into solution most of the muscle AChE and BuChE activities. The AChE activity was five times greater than that of BuChE in normal (NM) and dystrophic muscle (DM). The AChE activity in the serum of dystrophic mice was twice that measured in control animals, but the BuChE activity remained almost unchanged. Both AChE and BuChE in muscle and serum bound completely to concanavalin A (Con A) and Lens culinaris agglutinin (LCA). A12, A8 and G4 AChE, but not the light G2 and G1 AChE forms, in NM and DM were completely adsorbed to wheat germ agglutinin (WGA). Similarly, G4 BuChE, but not the G2 and G1 forms, were associated to WGA. A high proportion of G4 and G1 AChE and G4 BuChE forms in mouse serum were fixed to WGA. Asymmetric AChE in NM and DM reacted with Ricinus communis agglutinin (RCA) but the light AChE and BuChE forms in muscle and serum did not bind to the lectin. G4 AChE and G4 BuChE in NM were not recognized by RCA, but the isoforms in DM bound fully to the lectin. Serum G4 AChE from control or dystrophic mice did not react with RCA, but G4 BuChE was fixed to the lectin. Since RCA is specific for galactose, the results suggest that in dystrophic muscle galactose is incorporated early in G4 AChE and this affects the level of the functional tetramers destined for insertion in the plasma membrane.