- Top of page
- Experimental procedures
Acetylcholinesterase (AChE) exists in various molecular forms, depending on alternative splicing of its transcripts and association with structural proteins. Tetramers of the ‘tailed’ variant (AChET), which are anchored in the cell membrane of neurons by the PRiMA (Proline Rich Membrane Anchor) protein, constitute the main form of AChE in the mammalian brain. In the mouse brain, stress and anticholinesterase inhibitors have been reported to induce expression of the unspliced ‘readthrough’ variant (AChER) mRNA which produces a monomeric form. To generalize this observation, we attempted to quantify AChER and AChET after organophosphate intoxication in the mouse brain and compared the observed effects with those of stress induced by swimming or immobilization; we also analyzed the effects of heat shock and AChE inhibition on neuroblastoma cells. Active AChE molecular forms were characterized by sedimentation and non-denaturing electrophoresis, and AChE transcripts were quantified by real-time PCR. We observed a moderate increase of the AChER transcript in some cases, both in the mouse brain and in neuroblastoma cultures, but we did not detect any increase of the corresponding active enzyme.
In mammals, the acetylcholine hydrolyzing enzyme, acetylcholinesterase (AChE) presents various molecular forms. A single gene leads to a large variety of transcripts depending on promoter usage (Meshorer et al. 2004), combined with alternative splicing in the 3′ region (Li et al. 1993; Massoulié 2002). Alternative acceptor sites of splicing in the 3′ region leads to three different variants (R, H, T), which possess the same catalytic domain but distinct C-terminal domains (Fig. 1). The AChER (‘readthrough’) variant results from the absence of splicing after the last catalytic exon and produces a non-amphiphilic monomeric () enzyme, which normally represents a minor AChE species. The AChEH (‘hydrophobic’) variant produces glycophosphatidylinositol-anchored dimers, mostly found in blood cells. The AChET (‘tailed’) variant terminates with a 40-residues peptide (t), which is highly conserved among vertebrates; AChET subunits produce amphiphilic monomers (T1), homomeric dimers (T2) and tetramers (T4), as well as hetero-oligomers with collagen ColQ and with the transmembrane PRiMA (Proline Rich Membrane Anchor) protein.
Figure 1. Acetylcholinesterase (AChE) 3′-splice variants. Schematic structure of the 3′ end of the AChE gene. The exons 2, 3 and 4 encoding the catalytic domain are represented by grey boxes. An alternative choice of splice acceptor sites can generate three splice variants: the ‘readthrough’ variant (AChER) is the result of the lack of splicing at the 3′ end after the last catalytic exon 4, the ‘hydrophobic’ variant (AChEH) is produced by splicing from exon 4 to the H region (also called exon 5) or, in the adult muscles and brain, splicing generally occurs between exon 4 and the ‘tailed’ (or ‘synaptic’) variant (AChET) encoding region (also called exon 6) (Li et al. 1991). Specific C-terminal peptidic sequences (r, h and t peptides) for the different protein variants are shown in bold and aligned for human and mouse. For the h peptide, glypiation site is indicated (*) and cleaved region is in italics. The t peptide, which is important for the quaternary association of AChE subunits, particularly for their assembly with anchoring proteins into physiologically active forms, presents a remarkable conservation between mouse and human, among mammals and even throughout vertebrates in general. Positions of the AChER and AChE primers used for real-time PCR are represented on the genomic sequence.
Download figure to PowerPoint
These diverse molecular forms have specific localizations and probably specific functions. In the muscle, the size of the T4-PRiMA pool depends on the activity of the muscle and may exert a specific adaptive function, distinct from that of collagen-tailed forms (Jasmin and Gisiger 1990; Jasmin et al. 1991). In the brain, during early development or in the neurodegenerative Alzheimer's disease, the ratio between monomers and tetramers is higher than in normal adult brain (Atack et al. 1983; Muller et al. 1985) where T4-PRiMA is the main form in the different brain structures, cholinergic or not (Perrier et al. 2003). The t peptide also induces a partial intracellular degradation of unassembled subunits (Belbeoc'h et al. 2003; Falasca et al. 2005).
Acute stress and anticholinesterases have been reported to induce a long-term overexpression of the readthrough variant of AChE in the central nervous system (Kaufer et al. 1998; Meshorer et al. 2002). Neuronal AChER accumulation would be accompanied by long-lasting modifications such as hyperexcitation of glutamatergic activity (Cohen et al. 2002). Whether these modifications are due to cholinergic or non-cholinergic action of accumulated ‘readthrough’ is still unclear.
As heat shock is known to affect splicing (Shen et al. 1993; Takechi et al. 1994), we analyzed its effects on AChE transcripts in murine neuroblastoma cells. We also analyzed the effects of irreversible inhibition by an organophosphate on AChE expression and recovery in neuroblastoma cells and in vivo, in different structures of the mouse brain. We also analyzed the effects of immobilization and swimming stresses. AChE transcripts were quantified by real-time PCR, and AChE molecular forms were characterized by non-denaturing electrophoresis and by sedimentation in sucrose gradients.
- Top of page
- Experimental procedures
In this study, we examined the recovery of acetylcholinesterase (AChE) after partial inhibition, in several regions of the mouse brain. We considered the effect of such inhibition, as well as of other perturbations, on the production of the R (‘readthrough’) and T (‘tailed’) mRNA variants, which differ by the splicing of the 3′ region following the last exon encoding the catalytic domain. In addition to the mouse brain, we examined these transcripts in a neuroblastoma cell line (N18TG2) and in a neuroblastoma × glioma hybrid cell line (NG108CC15), under control conditions, after AChE inhibition, and after heat shock.
In the mouse brain, the level of AChE activity is about 10-fold higher in the striatum than in the somatosensory cortex or the cerebellum. All brain regions mostly contained a minor monomeric species and a major species which consists of an AChE tetramer associated with the membrane anchoring protein PRiMA (T4-PRiMA) (Perrier et al. 2002). Following a non-lethal treatment with the irreversible organophosphorous inhibitor soman, the monomers had already recovered their control level after 24 h. The relative activity of tetramers was decreased, after 24–48 h, in the order cortex > cerebellum > striatum; it recovered its original level after 1 week in the cerebellum and the striatum, but not in the cortex where it remained lower than control level after 2 weeks. The time required for assembly, transport and membrane insertion of the heteromeric T4-PRiMA form certainly contributes to its slower, recovery, but cannot explain the observed differences between brain regions. These differences probably arise from the fact that AChE tetramers are mostly produced locally by interneurons in the striatum, but at least partly imported by axonal flow from basal nuclei and from the brain stem to the cortex and cerebellum. In contrast, monomers mostly represent locally synthesized, rapidly renewed precursors of the tetramers. A predominantly intrinsic origin of the monomeric species is consistent with the fact that they are not transported by the fast axonal flow (Couraud and Di Giamberardino 1980; Brimijoin 1983) contrary to tetramers. Their relative amounts show a good correlation with total AChE mRNA expression in the cortex, striatum and cerebellum and therefore reflect the biosynthesis of AChE in these regions (Perrier et al. 2003). It has also been shown in cultured neuroblastoma cells that the half life of monomers is very short (2–3 h) compared to that of tetramers (about 50 h) (Lazar et al. 1984).
The levels of AChE transcripts and of PRiMA transcripts were not significantly changed, 24 h or 48 h after treatment with soman, in the three mouse brain regions examined, or in neuroblastoma cells after partial or complete inhibition. This shows that AChE inhibition did not trigger an up-regulation of the genes encoding the catalytic subunits or their anchoring protein, PRiMA. In the striatum, the level of the ‘readthrough’ transcript (R), appeared transiently increased (threefold at 24 h) after soman poisoning. It is not surprizing that this was not correlated with any change in total AChE transcripts, as splicing towards either T or R variants occurs post-transcriptionally. In addition the R transcripts represented at most 2% of the total AChE transcripts. In soman-treated neuroblastoma cells, the R transcripts did not increase, indicating that the effect observed in the mouse brain probably arises from the perturbation experienced by the nervous system, rather than from a cell-autonomous response to AChE inhibition.
To study the effect of stress, we followed as closely as possible the protocols used previously by the group of H. Soreq for forced swimming and for an immobilization stress, and we used the same strain of mice (Kaufer et al. 1998; Meshorer et al. 2002). We observed no significant changes in c-fos or AChE transcripts after a 4-day swimming protocol. However, immobilization did induce a two to threefold increase of c-fos transcripts, which may be considered as an index of stress response; it did not significantly change total AChE transcripts but increased the proportion of R transcripts about threefold, reaching about 1.5% of the total AChE mRNA. This increase in ‘readthrough’ mRNA levels did not significantly change total mRNA levels, in contrast with previous reports. Our estimations of R transcripts are consistent with previous observations that they represent a very low proportion of total AChE transcripts in the brain (Li et al. 1991) and in other tissues (Legay et al. 1995). In addition, we did not find any significant contribution of the AChER species to AChE activity in the brain, in agreement with the fact that the proportion of R transcripts remained very low.
Because heat shock can affect splicing (Shen et al. 1993; Takechi et al. 1994), we examined its effect on AChE transcripts in N18TG2 and NG108CC15 cells. The two cell types differ markedly in their expression of cholinergic traits: in particular the level of AChE activity is higher in the neuroblastoma × glioma hybrid cell line NG108CC15 (Melone et al. 1987). Both cell lines contain a higher proportion of R transcripts than in the mouse brain, compared to total AChE transcripts, and this may represent a character of incomplete differentiation, as it is also observed in embryonic tissues (Legay et al. 1995). As expected, HSP70 transcripts were highly increased after heat shock in the two cell lines; PRiMA transcripts were unaffected. In NG108CC15 cells, which express a higher level of AChE activity than N18TG2 cells (Biagioni et al. 1995), we found no change in total AChE transcripts or in the AChER variant; in contrast, we observed a modest increase of AChE transcripts in N18TG2, and a sixfold increase of AChER, which reached approximately the same level as in NG108CC15 cells. This suggests that the proportion of spliced (T) and unspliced (R) transcripts largely depends on the level of transcription of the AChE gene, which can be increased by heat shock in N18TG2, but not in NG108CC15, possibly because it is already maximal, corresponding to a more advanced cholinergic differentiation of these neuroblastoma × glioma hybrid cells.
The present results show that heat shock may affect splicing of the 3′ region of AChE transcripts, in a cell autonomous manner. In the mouse brain, AChER transcripts were increased under some stress conditions and after poisoning with an anti-AChE inhibitor, probably through multicellular stimulation effects rather than by cell autonomous mechanisms; their level remained low, compared to those of total AChE transcripts, which were unchanged; thus, the AChER transcript does not represent a sensitive indicator of stress. We did not detect active AChER enzyme in the mouse brain. It seems very unlikely that stress-induced suppression of splicing in the 3′ region of AChE transcripts would result in any physiologically significant change of AChE activity and of cholinergic activation in the brain.