Potential conflict of interest: Nothing to report.
Classical studies of cholinesterase activity during liver dysfunction have focused on butyrylcholinesterase (BuChE), whereas acetylcholinesterase (AChE) has not received much attention. In the current study, liver and plasma AChE levels were investigated in rats with cirrhosis induced after 3 weeks of bile duct ligation (BDL). BDL rats showed a pronounced decrease in liver AChE levels (∼50%) compared with sham-operated (non-ligated, NL) controls; whereas liver BuChE appeared unaffected. A selective loss of tetrameric (G4) AChE was detected in BDL rats, an effect also observed in rats with carbon tetrachloride-induced cirrhosis. In accordance, SDS-PAGE analysis showed that the major 55-kd immunoreactive AChE band was decreased in BDL as compared with NL. A 65-kd band, attributed in part to inactive AChE, was increased as became the most abundant AChE subunit in BDL liver. The overall decrease in AChE activity in BDL liver was not accompanied by a reduction of AChE transcripts. The loss of G4 was also reflected by changes observed in AChE glycosylation pattern attributable to different liver AChE forms being differentially glycosylated. BDL affects AChE levels in both hepatocytes and Kupffer cells; however, altered AChE expression was mainly reflected in an alteration in hepatocyte AChE pattern. Plasma from BDL rats had approximately 45% lower AChE activity than controls, displaying decreased G4 levels and altered lectin-binding patterns. In conclusion, the liver is an important source of serum AChE; altered AChE levels may be a useful biomarker for liver cirrhosis. (HEPATOLOGY 2006;43:444–453.)
The mammalian liver is rich in cholinesterases, enzymes that catalyze the hydrolysis of choline esters. The two main cholinesterases, acetylcholinesterase (AChE; EC 18.104.22.168), which terminates the action of acetylcholine post-synaptically, and butyrylcholinesterase (BuChE; EC 22.214.171.124), with no clear physiological function, differ in their substrate specificities and inhibition by selective inhibitors.1 In humans, reduced serum BuChE activity has been linked to chronic liver diseases, including cirrhosis.2 Because the level of AChE in human serum is much lower than the level of BuChE, most studies quantify serum BuChE levels as a liver function test, whereas AChE has not received much attention.
The rat is a good animal model to study the effect of liver pathology in AChE levels because BuChE activity in the rat is much lower than in human serum. Several authors have previously described the distribution of cholinesterases in the liver of rat,3–6 mouse,7 rabbit,8, 9 and chick.10 However, no significant conclusion has been reached about the potential physiological role of AChE in the liver and whether liver is an important source for serum AChE.
Ligature of the common bile duct in rats causes jaundice and impairment of liver function.11–13 In the rat, chronic bile duct ligation (BDL) leads to biliary cirrhosis within 3 to 4 weeks. BDL cholestatic liver injury occurs in conjunction with fibrosis, portal hypertension, portal-systemic shunting and immune system dysfunction.11, 13–15 In the current study, AChE expression and levels were investigated in rat liver and plasma with BDL-induced liver cirrhosis and in sham-operated controls. We also investigated whether the liver is an important source for serum AChE, and its potential use as a biomarker for cirrhosis.
AChE, acetylcholinesterase; BuChE, butyrylcholinesterase; BDL, bile duct ligation; NL, non-ligated; CCl4, carbon tetrachloride; RT-PCR, reverse transcripts polymerase chain reaction; cDNA, complementary DNA; Con A, Canavalia ensiformis lectin; LCA Lens culinaris agglutinin; RCA120, Ricinus communis agglutinin; WGA; Triticum vulgaris agglutinin.
Materials and Methods
Animals and Tissue Preparation.
All animal experiments were performed in accordance with local guidelines. Male Sprague-Dawley rats (250-300 g) were used. BDL surgery was performed under intraperitoneal anesthesia with a mixture of diazepam and ketamine. After laparotomy, the common bile duct was doubly ligated with silk threads and excised between the ligatures to prevent regeneration. In sham non-ligated (NL) controls, the bile duct was identified, manipulated, and left in situ. Rats were sacrificed 21 days after surgery. Animals were halothane-anesthetized, and blood was drawn by cardiac puncture. Plasma was separated from blood cells by centrifugation. In a second experimental group, micronodular cirrhosis with ascites was induced in rats by inhalation of carbon tetrachloride (CCl4), twice weekly.13 Phenobarbital (0.3 g/L) was included in the drinking water for both rats administered CCl4 and control rats. Rats were sacrificed after 12 weeks. Cirrhosis was routinely confirmed by blinded examination of hematoxylin-eosin–stained sections, as previously described.13 The liver was rapidly removed and washed exhaustively in saline. Samples were stored at −80°C for later analysis.
Serum Biochemistry Analysis.
Serum protein, albumin, total bilirubin, direct bilirubin, and transaminases were measured using an auto-analyzer (Vitros 950 Chemistry System, Raritan, NJ).
Protein Extraction From Liver.
Small pieces of liver were homogenized (10% w/v) in ice-cold Tris-saline buffer (50 mmol/L Tris-HCl, 1 mol/L NaCl, and 50 mmol/L MgCl2, pH 7.4) containing 1% (w/v) Triton X-100 and supplemented with a cocktail of proteinase inhibitors.16 The suspension was then centrifuged at 100,000g at 4°C for 1 hour to recover a cholinesterase-rich fraction.
Enzymatic Assay and Protein Determination.
AChE and BuChE activity was determined by a modified microassay method from Ellman.16 One unit AChE or BuChE activity was defined as the amount of extract able to hydrolyze one nanomole acetylthiocholine or butyrylthiocholine per minute at 22°C. Total protein concentrations were determined by using the bicinchoninic assay (Pierce, Rockford, IL).
Molecular forms of AChE were separated according to its sedimentation coefficients by centrifugation on 5% to 20% (w/v)sucrose gradients containing 0.5% (w/v) Triton X-100.7, 16 Ultracentrifugation was performed at 150,000g in an SW 41Ti Beckman rotor for 18 hours at 4°C. Approximately 40 fractions were collected from the bottom of each tube and assayed for cholinesterase activities. We defined the ratio of AChE forms G4/(G1+G2) as the proportion of G4 molecules versus the sum of the light globular forms, G1 and G2. The sucrose fractions containing G4 and G1+G2 peaks were pooled, dialyzed against Tris buffer, and concentrated by ultrafiltration (Amicon Ultra 10,000 MWCO, Millipore Corporation, Bedford, MA). AChE species were then assayed by Western blotting and lectin-binding analysis.
Detection of AChE Variants by Western Blotting.
AChE variants were detected by inmunoblotting. Briefly, liver extracts were resolved by SDS-PAGE on 10% polyacrylamide slab gels. After electrophoresis, proteins were blotted onto nitrocellulose membranes, blocked with 5% non-fat milk, and incubated overnight with a goat anti-AChE polyclonal antibody (E-19, Santa Cruz Biotechnology, Santa Cruz, CA). The strips were incubated with a HRP-conjugated donkey anti-goat IgG antibody (Santa Cruz Biotechnology), and immunoreactive AChE was detected using the ECL-Plus kit (Amersham Life Science, Arlington Heights, IL) in a Luminescent Image Analyzer LAS-1000 Plus (FUJIFILM, Stamford, CT). Pre-stained molecular weight markers were used to determine protein size (Sigma-Aldrich Co, St. Louis, MO). For semiquantitative analysis, the intensity of AChE bands was measured with Science Lab Image Gauge v4.0 software provided by FUJIFILM.
RNA Isolation and Analysis of AChE Transcripts by Northern Blotting and RT-PCR.
Liver RNA was extracted using TRIzol Reagent (Invitrogen Life Technologies, Grand Island, NY) according to the manufacturer's instructions. Aliquots (30 μg/sample) of total RNA were separated in denaturing 1% agarose gels and transferred to nylon membrane by the capillary method. A fragment of the AChE coding sequence was used as a probe using the DIG High Prime DNA Labelling and Detection Starter Kit II (Roche, Barcelona, Spain). Hybridization, washing, and detection of hybridized bands were performed according to the manufacturer's instructions.
Liver AChE mRNAs also were identified by reverse transcriptase polymerase chain reaction (RT-PCR) with selected primers. First-strand complementary DNAs (cDNAs) were obtained by reverse transcription using SuperScript III Reverse Transcriptase (Invitrogen Life Technologies) according to the manufacturer's instructions, using 1 μg total RNA and oligo (dT)12-18. Oligonucleotide primers used for PCR analysis of the transcripts are indicated in Fig. 4. For semiquantitative RT-PCR, aliquots of the cDNA samples were incubated with Taq DNA Polymerase (1 U) in the presence of primers (0.2 μmol/L each) and a dNTP Mix (0.2 mmol/L). The number of cycles (25-35) was chosen to ensure that quantification was performed in the linear amplification phase. All cDNAs were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH).
Lectin-Binding Analysis of AChE.
Aliquots of liver extracts and plasma were mixed with immobilized lectins [Canavalia ensiformis (Con A), Lens culinaris agglutinin (LCA), Ricinus communis agglutinin (RCA120), and Triticum vulgaris agglutinin (wheat germ, WGA); Sigma-Aldrich]. After overnight incubation at 4°C, AChE–lectin complexes were separated from free AChE by centrifugation. The unbound AChE activity in the supernatant fraction was used to compare differences in lectin binding among groups.
AChE Histochemistry and Immunocytochemistry.
Animals were perfused through the heart with 0.1 mol/L phosphate buffer, pH 7.4, followed by 4% paraformaldehyde in phosphate buffer containing 1% CaCl2. The liver was then removed and immersed in the same fixative overnight. Sections of paraffin-embedded tissues were stained with hemotoxylin-eosin and examined by light microscopy for necrosis and other structural changes. For AChE histochemistry, the tissue was cryoprotected in 30% sucrose and 40-μm-thick frozen sections were cut on a sliding microtome. Frozen sections were rinsed and treated for AChE according to the procedure of Karnovsky and Roots.17 AChE staining was also examined by immunohistochemical analysis with the anti-AChE antibody E-19 (1:100 dilution). Similar sections were also incubated with the polyclonal anti-Kupffer cell antibody CD68 (1:100 dilution; Santa Cruz Biotechnology).
Isolation of Hepatocytes and Kupffer Cells.
Liver cells were isolated by sequential digestion with collagenase, as previously described.18 Briefly, the liver suspension was centrifuged for 5 minutes at 50g, and the pellet containing the hepatocytes was resuspended in Hank's balanced salt solution and centrifuged at 400g (10 minutes, 4°C), over a 60% Percoll solution (Amersham Pharmacia Biotech) to obtain purified hepatocytes. The supernatant containing the non-parenchymal cell fraction was washed with Hank's balanced salt solution and centrifuged on a 25%/50% Percoll gradient at 800g (20 minutes, 4°C). The fraction enriched in Kuppfer cells and sinusoidal endothelial cells was harvested from the interface of the gradient. Kupffer cells were isolated from non-adherent cells by selective attachment onto tissue culture plates.
Measurements are expressed as means ± SEM. Data were analyzed by using a Student t test with SigmaStat v2.03 software (SPSS, Chicago, IL).
Rats killed 21 days after BDL showed weight loss, jaundice, dark urine, ascites, and evident signs of portal hypertension. These symptoms were not seen in NL animals. Conventional serum markers of liver function also revealed liver damage in BDL rats (Fig. 1).
Liver AChE and BuChE Activities in NL and BDL Rats.
We first attempted to determine whether AChE and BuChE levels were altered in rat BDL liver (Fig. 2A). BDL liver samples had lower AChE activity (51% decrease; P < .001) compared with NL controls. The level of liver AChE was also decreased in CCl4–cirrhosis-induced rats (28% decrease; P = .002) compared with controls (Fig. 2B). Despite the marked decrease in AChE in the liver with cirrhosis, no significant change in BuChE levels was observed in the experimental groups.
AChE is expressed as several molecular forms with a cell type-dependent pattern that is altered in many pathological processes.19 These molecular forms can be distinguished by their molecular weights and hydrodynamic properties. To determine whether the decrease in AChE activity was attributable to a decrease of a specific form, liver supernatants were fractionated on sucrose density gradients to separate the major AChE species. Abundant 10.2 ± 0.1S (Svedberg units) and 4.7 ± 0.1S species were identified in liver homogenates (Fig. 2C-D). These molecular forms corresponded to AChE tetramers (G4) and monomers (G1). On some of the gradients, a dimeric (G2; 5.5 ± 0.2S) form of AChE was distinguished. We obtained similar profiles when animals were perfused with saline at sacrifice, indicating that most of the liver AChE activity assayed did not come from blood. In the BDL liver, the G1+G2 peak decreased slightly, whereas the tetramer was almost undetectable. Consequently, a strong decrease in the G4/(G1+G2) ratio occurred in BDL rats as compared with NL controls (Fig. 2C). Similarly, in rats administered CCl4, the G4 peak, and subsequently the G4/(G1+G2) ratio, was significantly reduced (Fig. 2D).
No significant differences were observed in the level of predominant light forms and tetramers of BuChE in both the model animals and their respective controls (data not shown).
Immunodetection and Quantification of Liver AChE.
Because an inactive pool of AChE subunits has been demonstrated,20–22 we also analyzed rat liver extracts by sodium dodecyl sulfate polyacrylamide gel electrophoresis under fully reducing conditions, followed by Western blotting using the E-19 anti-AChE antibody. This polyclonal antibody was raised against a peptide mapping to the amino terminus of AChE, common to all AChE forms and thus presumably detects all variants, including inactive subunits. Antibodies similar to E-19, directed against the N-terminal domain common to all AChE variants, have shown several migrating AChE bands.23, 24 In accordance, E-19 detected three major bands of approximately 70, 65, and 55 kd in liver homogenates (Fig. 3A-B). The immunoreactivity of the major 55-kd band was decreased in BDL compared with NL (61% decrease, P < .001). The 65-kd subunit that represents the second most intense band in NL, was significantly increased (35%, P = .03) in BDL to become the most abundant AChE species. The level of the minor immunoreactive 70-kd band was not affected in BDL rats. Because of the lack of correlation between the increase in immunoreactivity of the 65-kd band and the decline of all AChE species, we assume that this band corresponds, at least in part, to inactive subunits of the protein.
In an attempt to assign immunopositive AChE bands to specific AChE species, immunoblots were performed for the G4 and G1+G2 peaks from sucrose density gradients (Fig. 3C, E). In NL, most of the G4 peak was the 55 kd, whereas both the 55- and 70-kd bands were observed for the G1+G2 peak. In BDL rats, the increase in immunoreactivity of the 65-kd band corresponds to lighter AChE forms, in accordance with the identity of this band as inactive AChE. Inactive AChE molecules have only been described in G1+G2 peak, whereas the G4 peak appears to contain only active AChE.20–22
Because developing tissue is rich in light catalytic and non-catalytic AChE species,19 molecular form and banding patterns also were studied in embryonic rat liver (Fig. 3D). Abundant G1 and G2 forms were identified in developing liver, with a predominant 65-kd band detected by immunoblotting.
Analysis of AChE Transcripts.
Some of the molecular heterogeneity of AChE derives from alternative RNA splicing, generating three polypeptide encoding transcripts with the same catalytic domain, but distinct C-terminal peptides.19, 25, 26 The “readthrough” or R-transcripts encode monomeric soluble subunits. The “hydrophobic” or H-transcripts generate glycosylphosphatidylinositol-anchored dimers and monomers. Finally, the “tailed” or T-transcripts encode subunits that produce a wide variety of monomeric and oligomeric forms, including tetramers. Therefore, almost all the loss of AChE activity in BDL liver should correspond to subunits encoded by T-transcripts. To test this, we performed Northern blot analysis of AChE mRNA (Fig. 4A). In accordance with previous reports,9, 27 we detect a weak band of approximately 2.4 kb, with an additional band of approximately 3.8 kb that appeared only after overexposure. The 2.4-kb band paralleled the major transcript in rat brain, the T-transcript, and was unaltered in liver extract from BDL rats compared with NL rats. Because the content of AChE messengers in rat liver was very low, we performed semi-quantitative RT-PCR analysis of the AChE mRNA. Figure 4B shows the position and orientation of PCR primers. We detect the T transcript using primers 1bs/Ta and the R-transcript with 1bs/Ha (Fig. 4C). The overall decrease in AChE activity detected in BDL liver was not accompanied by a reduction in the T-transcript, whereas BDL causes a 50% increase of the R-transcript (Fig. 4D). The alternatively spliced transcript R-AChE is induced under psychological, chemical, or physical stress.28, 29 We did not detect the H-transcript, which has been previously described to be scarce in adult rat liver.7, 27 Phosphatidylinositol-specific phospholipase C treatment demonstrated that the level of G1 and G2 glycosylphosphatidylinositol-anchored species was only slightly decreased in the BDL liver (not shown).
Lectin Binding of Liver AChE.
To further examine whether the glycosylation of AChE is altered in BDL rats, we determined the ability of AChE to bind to several immobilized lectins, proteins that avidly bind to specific sugar moieties of glycoproteins.30 The liver AChE from control rats bound strongly to Con A, moderately to RCA and WGA, and weakly to LCA (Fig. 5A). A significant change occurred in the lectin-binding ability of AChE assayed in BDL compared with NL liver extracts. No significant change in BuChE lectin binding was observed (Fig. 5B). Because carbohydrate moieties vary between the different AChE forms,31 we further investigated glycosylation in individual AChE forms. The G4 and G1+G2 peaks were separated by sucrose gradient centrifugation, incubated separately with immobilized lectins, and the percentage of enzymatic activity unbound was calculated for each form (Fig. 5C). The fraction of G2 and G1 recognized by each lectin was similar in NL and BDL extracts, whereas the G4 form binds more efficiently to all 4 lectins, particularly to RCA, than G2 and G1.
Morphological Observations and AChE Staining.
Histological examination confirmed the development of biliary cirrhosis in the BDL liver (Fig. 6A-B). The injury caused cholestasis, necrosis, periductular inflammation, and fibrosis. The distribution of AChE-positive cells in the rat liver is illustrated by immunocytochemistry using the E-19 antibody in Fig. 6C-D. Similar to previous reports, we found that AChE was present in hepatocytes5, 32 and in the resident liver macrophages, the Kupffer cells.4, 8 Histochemical analysis of AChE activity confirmed both the pattern and the intensity of AChE staining (not shown). In BDL animals AChE staining reflected the severe alterations in the liver and, in agreement with our biochemical determinations, we detected a strong loss in AChE labelling in hepatocytes and Kupffer cells. Specific Kupffer cell staining with the CD68 antibody showed an apparent loss of this macrophage cell type in the BDL liver (Fig. 6E-F).
G4 AChE Loss in BDL Liver Occurs in Hepatocytes.
Hepatocytes and Kupffer cells were isolated and AChE activity extracted. In NL controls, the level of AChE-specific activity was 1.65-fold higher in Kupffer cells than in hepatocytes (Fig. 6G), confirming our histological observations. In the BDL group, AChE levels were decreased by 47% (P = .01) and 65% (P < .001) for hepatocytes and Kupffer cells respectively, compared with cells from NL controls. To determine whether the G4 loss in BDL liver is attributable to a depletion of this AChE form in hepatocytes, cell homogenates were fractionated on sucrose density gradients. Control hepatocytes were rich in G4 and light forms similarly to that observed for control livers (Fig. 6H). The loss of tetramers in the BDL condition demonstrates that changes in the liver are a consequence of altered hepatocyte AChE pattern.
Because the 65-kd band increased in BDL liver, we next compared the AChE banding pattern for both cell types by immunoblotting with the E-19 antibody. The relative abundance of AChE bands differed in hepatocytes and Kupffer cells (Fig. 6I). In both cases, the 55 kd was the most abundant band. In hepatocytes, this was followed by the 70-kd and 65-kd bands, which were undetectable in Kupffer extracts. In hepatocytes from BDL rats, only traces of the 55-kd band were observed, the 70-kd band slightly decreased, whereas the 65-kd was slightly increased, indicating again that changes in liver AChE are mainly derived from hepatocytes.
AChE in Plasma.
Serum cholinesterases presumably originate in liver cells; however, other organs also can contribute to the pool of these enzymes in plasma. To determine whether a change in plasma AChE is similar to that occurring in liver after BDL, we examined its levels, pattern of molecular forms, and glycosylation. Analysis of AChE activity in plasma indicated that the plasma from BDL rats had lower AChE activity than that of control animals (44% decrease; P < .001; Fig. 7A). Sedimentation analysis of AChE in control rat plasma showed abundant tetramers and trace amounts of light forms (Fig. 7B). As expected, we found that the decreased levels corresponded to the loss of G4 forms in plasma, resulting in significant differences (P < .001) in the G4/(G1+G2) ratio between BDL and NL rats (Fig. 7C). We further characterized this alteration by lectin binding. The decrease of plasma G4 AChE as a result of liver dysfunction in BDL animals, was also reflected by a decrease in the percentage of plasma AChE unbound to the lectin RCA compared with controls (P < .001; Fig. 7D). Analysis of plasma AChE enabled us to fully discriminate between BDL and NL subjects. We found similar levels and binding properties of plasma BuChE to lectins in both NL and BDL rats (not shown), indicating that circulating BuChE was not affected.
In this study, we found a marked decrease in the AChE levels in BDL rat livers with cirrhosis. A similar decrease was observed in rats with CCl4-induced cirrhosis. This change was specific for AChE, because liver BuChE, whose decrease in human serum has been linked to chronic liver disease,2 was apparently unaffected. The two cholinesterases are regulated by separate mechanisms.33, 34 In humans, the amount of BuChE in plasma is much higher than that of AChE,35–37 whereas similar amounts of both enzymes are found in liver.38 This study shows that liver disease has an effect on the expression and levels of AChE in both liver and plasma.
The physiological significance of cholinesterase activity in liver and serum has thus far not been elucidated. AChE may function in hepatocytes and other epithelial cells to inactivate blood-circulating acetylcholine.7 However, AChE also may be involved in some intercellular and intracellular regulatory mechanisms.4 In addition to enzymatic activity, a possible structural change of AChE may contribute to cellular alterations in liver cirrhosis, because roles for AChE other than its acetylcholine hydrolytic activity have been proposed, such as regulation of cell growth and cell adhesion.19, 39, 40 The different molecular forms of AChE thus may reflect specific physiological functions of the enzyme with different regulatory requirements. We found a depletion of tetramers in liver of BDL rats with cirrhosis, whereas G1 and G2 species were only slightly affected. A decrease in G4 but sparing of G1 AChE has been noted in pathological states such as Alzheimer disease.16, 41 The prevalence of lighter AChE forms in pathological conditions, including cirrhosis, resembles an embryonic pattern of expression.19, 39–41 The inactive AChE fraction, which is presumably increased in the liver of BDL rats, is proportionally more abundant in embryonic than in adult tissues.19 Thus, the loss in oligomeric species may reflect a pathologically induced loss of cellular differentiation control or a change in expression as part of a response activated by the insult.
Another interesting finding in our study is that the decreased level of G4 AChE activity in BDL liver was not accompanied by a reduction in AChE T transcripts. AChE levels are regulated at transcriptional, post-transcriptional and posttranslational levels, leading to complex expression patterns that are cell-type specific, and modulated by differentiation state, physiological, and pathological conditions through mechanisms that are not fully understood. Moreover, Bon and co-workers have demonstrated that the C-terminal t peptide of AChE, which characterizes type T subunits, determines its oligomeric association, secretion, and degradation.42, 43 BDL may therefore affect AChE oligomerization and degradation by influencing the t peptide. However, BuChE, which contains a common tetramerization domain as AChE, remains unaffected by BDL.
Regarding the moderate increase of the AChE R-transcript in BDL, in some cases an increase in this specific R-transcript is not accompanied by any increase of the corresponding active enzyme.29 Whether the increase of 65-kd subunits, assigned in part to inactive AChE, is related to increasing R-transcript expression in the BDL liver warrants further study.
Many pathological states, including cirrhosis,44, 45 cause characteristic changes in the glycosylation pattern of particular glycoproteins. The percentage of fucosylated serum cholinesterases is altered in hepatocellular carcinomas and cirrhosis.46 Nevertheless, changes in AChE glycosylation in BDL liver appear to reflect changes in the contents of differentially glycosylated liver AChE forms. However, the levels and molecular pattern of BuChE remain unaffected by BDL and display a normal pattern of glycosylation.
Because abnormal glycosylation can increase protein degradation, affecting AChE stability and secretion,47, 48 loss of tetramers may reflect a perturbed glycosylation. However, the glycosylation of lighter AChE species appear to be unaffected in the liver of BDL rats. Interestingly, the loss in serum tetramers in BDL rats correlated with an increased proportion of AChE binding to the lectin RCA, indicating that depleted G4 forms are poorly recognized by the lectin, whereas the G4 form from the liver binds strongly to RCA. This difference suggests that the liver produces two pools of AChE tetramers, a membrane-bound pool and a secreted pool, possibly characterized by distinct glycosylation patterns. Most of the AChE in rat plasma exists as tetramers, which may be secreted from various tissues, including liver. Different cell types add different carbohydrate moieties onto cholinesterases.31 Thus, the unusual glycosylation of AChE in BDL plasma could be attributable to a decrease in the proportion of tetramers that originate in the liver, with differing glycosylation patterns from isoforms of different cellular origin. The altered AChE glycosylation pattern observed in BDL liver and plasma is probably a consequence of the depletion of both cellular and secreted pools of G4 AChE and not attributable to perturbed glycosylation.
The depletion of G4 may also be the consequence of degeneration of a particular liver cell type. Kupffer cells do not function normally when biliary obstruction occurs. In agreement with previous reports for obstructive jaundice and chronic CCl4 administration,49, 50 we found that Kupffer cells were depleted on induction of liver cirrhosis. Although these cells have higher AChE levels than hepatocytes, it is unlikely that Kupffer cells contribute much to liver AChE activity, because they represent less than 15% of total liver cells. Thus, the decrease in Kupffer cell numbers in the BDL liver is not sufficient to explain the depletion in the tetrameric AChE, which represents approximately 50% of total liver AChE activity. Indeed, we show a selective G4 depletion in hepatocytes isolated from BDL livers.
Our studies show that a decrease of serum AChE levels in BDL rats correlate with the overall loss of AChE activity observed in liver. This provides clear evidence that a fraction of plasma AChE is synthesized in liver. In addition, the current study indicates that the decreased levels of serum AChE and biochemical features such as molecular or glycosylation pattern may be reliable markers of liver cirrhosis.
The authors thank Dr. J. Massoulié (Laboratoire de Neurobiologie Cellulaire et Moléculaire, CNRS, Paris) France, for comments on the manuscript and the generous gift of the cDNA encoding T-subunit of rat brain AChE. We also thank to D. Seguí, E. Ausó, M.-J. Molina, and M. Ródenas for technical assistance.