Acetylcholinesterase, a key prognostic predictor for hepatocellular carcinoma, suppresses cell growth and induces chemosensitization

Abstract

Acetylcholinesterase (ACHE) plays important roles in the cholinergic system, and its dysregulation is involved in a variety of human diseases. However, the roles and implications of ACHE in hepatocellular carcinoma (HCC) remain elusive. Here we demonstrate that ACHE was significantly down-regulated in the cancerous tissues of 69.2% of HCC patients, and the low ACHE expression in HCC was correlated with tumor aggressiveness, an elevated risk of postoperative recurrence, and a low survival rate. Both the recombinant ACHE protein and the enhanced expression of ACHE significantly inhibited HCC cell growth in vitro and tumorigenicity in vivo. Further study showed that ACHE suppressed cell proliferation via its enzymatic activity of acetylcholine catalysis and degradation. Moreover, ACHE could inactivate mitogen-activated protein kinase and phosphatidyl inositol-3′-phosphate kinase/protein kinase B pathways in HCC cells and thereby increase the activation of glycogen synthase kinase 3β and lead to β-catenin degradation and cyclin D1 suppression. In addition, increased ACHE expression could remarkably sensitize HCC cells to chemotherapeutic drugs (i.e., adriamycin and etoposide). Conclusion: For the first time, we describe the function of ACHE as a tumor growth suppressor in regulating cell proliferation, the relevant signaling pathways, and the drug sensitivity of HCC cells. ACHE is a promising independent prognostic predictor for HCC recurrence and the survival of HCC patients. These findings provide new insights into potential strategies for drug discovery and improved HCC treatment. (HEPATOLOGY 2011;53:493-503)

Hepatocellular carcinoma (HCC) is one of the most prevalent human malignancies worldwide and especially in East Asia and South Africa.1 The risk factors for HCC, such as aflatoxin B exposure and hepatitis B virus or hepatitis C virus infection, are well documented. However, the molecular pathogenesis of HCC remains to be comprehensively elucidated. Making matters worse, the particularly high rate of postsurgical recurrence (50%-70% at 5 years) renders this disease a major challenge that is highly refractory to conventional chemotherapy and radiation.2

Our previous study, which was based on a large-scale complementary DNA transfection screening, showed that some neurotransmitter-related genes are involved in HCC cell proliferation or survival.3 Neurotransmitters have been for a long time considered to be confined to the nervous system, and reports about the presence of neurotransmitters in microorganisms, plants, and lower animals have emerged in recent years.4 The transmitter acetylcholine (Ach) might function in the regulation of cell fates such as proliferation, differentiation, and the establishment of cell-cell contacts, as reviewed by Schuller.5 However, the synthesis, secretion, and degradation processes of Ach and its biological role in the human liver remain unclear. Acetylcholinesterase (ACHE), a very important member of the cholinergic system, has been shown to promote neurite outgrowth and differentiation,6 as well as amyloid deposition,7 and this confirms the important role of ACHE in the development of Alzheimer's disease.8 The best characterized function of ACHE is the rapid hydrolysis of Ach in cholinergic synapses for terminating Ach-mediated neurotransmission.8 Interestingly, the expression of ACHE is not restricted to cholinergic nervous tissues and can be found even in various types of tumors.9 Moreover, ACHE may have some basic functions in nonneuronal tissues, such as cell differentiation and cell adhesion.10, 11 However, the function and molecular mechanism of this gene in hepatic carcinogenesis remain unknown.

In this study, we have found that ACHE is often down-regulated in HCC. Here, for the first time, we show that the down-regulated ACHE expression in HCC is significantly associated with tumor aggressiveness, an elevated risk for postoperative recurrence, and a poor survival rate. Further investigations have demonstrated that ACHE can inhibit HCC cell growth both in vitro and in vivo. Moreover, its enhancing effect on drug-induced apoptosis provides a potential combinatorial therapy for HCC patients.

Abbreviations: Ach, acetylcholine; ACHE, acetylcholinesterase; AChRα7, acetylcholine receptor α7; ADR, adriamycin; Akt, protein kinase B; CCK8, Cell Counting Kit 8; CHAT, choline acetyltransferase; CNL, corresponding noncancerous liver; DFS, disease-free survival; EGF, epidermal growth factor; ERK, extracellular signal-regulated kinase; GSK3β, glycogen synthase kinase 3β; HCC, hepatocellular carcinoma; IC₅₀, median inhibitory concentration; mAChR, muscarinic-acetylcholine receptor; MAPK, mitogen-activated protein kinase; MEC, mecamylamine; mRNA, messenger RNA; NA, not applicable; nAChR, nicotinyl-acetylcholine receptor; Neo, neostigmine; OD_450nm, optical density at 450 nm; OS, overall survival; PI3K, phosphatidyl inositol-3′-phosphate kinase; shRNA, short hairpin RNA; TNM, tumor-node-metastasis; VAChT, vesicular acetylcholine transporter; VP-16, etoposide; Wm, wortmannin.

Materials and Methods

Cell Culture and Reagents.

Human HCC cell lines, including SMMC-7721, SMMC-7721-GFP (established by our laboratory with stable integration of the green fluorescent protein gene), Huh-7, SK-Hep1, and PLC/PRF/5, and the human hepatoblastoma cell line HepG2 were cultured in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (Invitrogen), 100 IU/mL penicillin G, and 100 mg/mL streptomycin sulfate (Sigma-Aldrich, St. Louis, MO). The chemical reagents and drugs that were used (listed in Supporting Table 1) were diluted to the desired concentrations in dimethyl sulfoxide (Sigma-Aldrich) or sterilized water according to the manufacturer's protocol and were applied at the concentrations described in the Results section.

HCC Samples and Clinical Features.

In this study, three sets of HCC samples were used. The first set was used for the identification of the cholinergic system and the pilot study of ACHE expression in HCC. The second set was used for the analysis of ACHE expression by immunohistochemical staining and its correlation with clinicopathological features. The third set was used for the analysis of the Ach concentration by liquid chromatography/high-performance liquid chromatography and its correlation with ACHE expression. Brief details are reported in the supporting information. All human materials were obtained with informed consent, and protocols were approved by the ethical review committee of the World Health Organization Collaborating Center for Research in Human Production (authorized by the Shanghai Municipal Government).

Reverse-Transcription Polymerase Chain Reaction and Real-Time Polymerase Chain Reaction.

Total RNA was isolated from cells or tissues with the TRIzol reagent (Invitrogen) according to the manufacturer's protocol. Real-time polymerase chain reaction was performed with the ABI-Prism 7300 sequence detection system (Applied Biosystems, Foster City, CA). The primers for the SYBR Green assay (Takara, Dalian, China) and reverse-transcription polymerase chain reaction are shown in Supporting Table 2. All experiments were performed in triplicate unless stated otherwise.

Immunohistochemical Staining and Western Blotting.

Paraffin sections of HCC tissues and sections of normal liver tissues were processed for immunohistochemical staining as described previously.12 Western blotting was performed as described.12 The antibodies are listed in Supporting Table 3. Brief details are reported in the supporting information.

Cell Proliferation Assay, Foci Formation Assay, and In Vivo Tumor Formation Assay.

Cell proliferation was monitored with the Cell Counting Kit 8 (CCK8) assay (Dojindo Laboratories, Kumamoto, Japan) according to the manufacturer's instructions. Megascopic cell colonies were stained with crystal violet (Sigma-Aldrich). For in vivo tumor formation, 2.5 × 10⁶ SMMC-7721-GFP cells that stably expressed ACHE or ACHE-short hairpin RNA (shACHE) were suspended in 100 μL of serum-free Dulbecco's modified Eagle's medium and subcutaneously injected into one flank of each mouse (eight female BALB/c-nu/nu in each group for 8 weeks); the vector control cells were injected into the other flank. After 6 weeks, the mice were sacrificed, and the parameters were measured. Brief details are reported in the supporting information.

ACHE Activity Assay.

The ACHE activity was analyzed with Fluoro-AChE (Cell Technology, Mountain View, CA) according to the manufacturer's guide. Brief details are reported in the supporting information.

Plasmid Constructs, Transfection, Lentivirus Production, and Transduction.

The plasmids are listed in Supporting Table 4. Brief details of the plasmid construction are reported in the supporting information. The lentivirus was harvested 48 hours after the transfection of target plasmids into HEK-293T cells with the Lipofectamine 2000 transfection reagent (Invitrogen). The target cells were infected with the filtered lentivirus and 6 μg/mL polybrene (Sigma-Aldrich) for 24 hours.

Fluorescence-Activated Cell Sorting Analysis.

At least 20,000 cells were analyzed with fluorescence-activated cell sorting cytometry (Epics Altra, Beckman Coulter, Miami, FL) and MultiCycle AV for Windows 5.0 (Phoenix Flow Systems, San Diego, CA). Brief details are reported in the supporting information.

Statistical Analysis.

The data are expressed as means and standard errors. Statistical analysis was performed with the Student t test (two-tailed; P < 0.05 was considered significant) unless otherwise specified (χ2 test, Pearson's correlation, and linear regression).

Results

ACHE Is Often Down-Regulated in HCC and Is Reversely Associated With Tumor Aggressiveness and Patient Prognosis.

The expression of the cholinergic system in nonneuronal organs and in the preneuronal stage is a widespread phenomenon,13 but its existence and function in HCC remain unclear. Here we examined the expression of its main components in HCC tissues, corresponding noncancerous liver (CNL) tissues, normal liver tissues, and human liver cancer cell lines. The results showed that all the molecular equipment of the system, including choline acetyltransferase (CHAT), ACHE, vesicular acetylcholine transporter (VAChT), nicotinyl-acetylcholine receptor (nAChR), and muscarinic-acetylcholine receptor (mAChR), was expressed in the detected samples (Supporting Fig. 1A,B). During detailed scanning of the key cholinergic components, we found that the messenger RNA (mRNA) expression level of ACHE was significantly lower in HCC tissues versus CNL tissues or normal liver tissues (Fig. 1A). Furthermore, the immunohistochemical staining assay also showed that the expression of ACHE protein was down-regulated in 69.2% (166/240) of HCC patients (Fig. 1B,C and Supporting Table 5).

Figure 1. ACHE is often down-regulated in HCC and is reversely associated with patient prognosis. (A) Real-time polymerase chain reaction analysis of CHAT, ACHE, VAChT, and acetylcholine receptor α7 (AChRα7) expression in HCC, matched noncancerous, and normal (N) liver tissues. The values are expressed as means and standard errors of the mean. (B) Immunohistochemical staining of ACHE in HCC and matched noncancerous liver tissues (original magnification ×400). (C) The expression of ACHE protein was down-regulated in 69.2% (166/240) of the HCC patients. (D) OS for the expression of ACHE. (E) DFS for the expression of ACHE. The statistical significance was labeled with the χ² test.

To further explore the clinical impact of down-regulated ACHE in HCC tissues, we analyzed the ACHE expression status with respect to various parameters in 240 HCC patients. Our results showed that a high expression level of ACHE was associated with improved overall survival (OS; P < 0.001) and disease-free survival (DFS; P = 0.002; Fig. 1D,E and Table 1). Patients were divided into three groups: (I) the expression of ACHE was low and was scored as 1 (n = 46), (II) the expression of ACHE was moderate and was scored as 2 (n = 125), and (III) the expression of ACHE was high and was scored as 3 (n = 69). Significant differences in recurrence and survival were found between groups I and II and between groups I and III. The median DFS and OS were 7.0 and 35.5 months for group I and 15.0 and 61.0 months for group II, respectively, and they were not reached for group III (Supporting Tables 6 and 7). To estimate the association of ACHE with tumor biological behaviors, comparisons of the clinicopathological features with ACHE expression were made. Patients with low ACHE expression were more likely to exhibit aggressive clinicopathological features; high ACHE expression correlated with single tumor nodules versus multiple tumor nodules (P = 0.002), small size (P = 0.023), and complete encapsulation (P = 0.009), whereas ACHE-low patients reached an advanced tumor-node-metastasis (TNM) stage (P = 0.009; Table 2). Thus, our data strongly imply that the ACHE expression level in HCC tissues is an independent indicator of HCC aggressiveness and prognosis.

Table 1. Univariate Analysis of the Prognosis Factors Associated With Survival

Variable	DFS			OS
	Hazard Ratio	95% Confidence Interval	P	Hazard Ratio	95% Confidence Interval	P
For the univariate analysis, the Cox proportional hazards regression model was used. Abbreviation: NA, not applicable.
Overall	NA	NA	0.001	NA	NA	<0.001
I versus II	1.91	1.23-2.95	0.004	2.17	1.41-3.34	0.0004
I versus III	2.33	1.40-3.88	0.001	2.72	1.63-4.55	0.0001
II versus III	1.22	0.78-1.90	0.380	1.25	0.79-1.98	0.338

Table 2. Correlation of the Clinicopathological Findings With Tumor ACHE Expression

Variable	ACHE (n)
	Low	Moderate	High	P
Pearson's χ^2 test was used. The bolding stands for the P-values with significant difference. * Cytokine-based immunotherapy, such as interferon-α and interleukin-2. † Radiofrequency ablation and percutaneous ethanol injection.
Age
≤52 years	21	67	31	0.391
>52 years	20	58	38
Gender
Female	9	19	8	0.501
Male	37	16	61
Hepatitis history
Yes	44	116	64	0.782
No	2	9	5
Alpha-fetoprotein
≤20 ng/mL	12	39	31	0.068
>20 ng/mL	34	86	38
Gamma-glutamyltransferase
≤54 (U/L)	19	52	33	0.671
>54 (U/L)	27	73	36
Liver cirrhosis
Yes	39	111	62	0.689
No	7	14	7
Tumor differentiation
I-II	24	66	45	0.205
III-IV	22	59	24
Tumor size
≤5 cm	18	56	43	0.023
>5 cm	28	69	26
Tumor multiplicity
Single	27	97	60	0.002
Multiple	19	28	9
Tumor encapsulation
None	26	71	24	0.009
Complete	20	54	45
Vascular invasion
Yes	22	69	40	0.553
No	24	56	29
TNM stage
I	15	55	36	0.009
II	11	42	23
III	20	28	10
Adjuvant therapy
None	18	38	28	0.215
Transarterial chemoembolization	24	75	30
Immunotherapy*	4	12	11
Postrecurrent therapy
None	10	15	4	0.230
Transarterial chemoembolization	15	35	13
Regional†	1	1	3
Resection	2	9	5

ACHE Inhibits Cell Proliferation In Vitro and Suppresses the Tumorigenicity of HCC Cells In Vivo.

To elucidate the role of ACHE in liver cancer, we first examined the mRNA and protein expression of ACHE in hepatic cancer cell lines (Supporting Fig. 2A,B), and we performed cell proliferation assays with the recombinant ACHE protein treatment in SMMC-7721 and other HCC cells. Interestingly, the results showed that the recombinant ACHE protein could decrease the cell proliferation rate, whereas the ACHE inhibitor neostigmine (Neo) exhibited the opposite effects (Fig. 2A and Supporting Figs. 3A and 4A,B; the efficiency of the recombinant ACHE protein or Neo was detected as shown in Supporting Fig. 3B,C). To further determine the impact of ACHE on the growth of liver cancer cells, we established stable cell lines transduced by the lentivirus carrying the ACHE gene or ACHE–short hairpin RNA (shRNA), which were designated Lenti-ACHE and Lenti-shACHE, respectively, in SMMC-7721 or other HCC cells (Supporting Fig. 5). The results demonstrated that high expression of ACHE could decrease cell proliferation; on the contrary, knockdown of ACHE could enhance the proliferation of HCC cells (Fig. 2B and Supporting Figs. 4C,D and 6). All these data indicated that ACHE could inhibit HCC cell proliferation in vitro. To further explore the effect of ACHE on tumorigenicity in vivo, Lenti-ACHE or Lenti-shACHE cells based on SMMC-7721-GFP were transplanted into nude mice subcutaneously, with the control cells placed in the other flank. Interestingly, the size and weight of the tumors formed by Lenti-ACHE cells were significantly decreased in comparison with the tumors formed by Lenti-vector cells (Fig. 2C,D and Supporting Fig. 7A,B,E), whereas the parameters of the tumors that developed from Lenti-shACHE cells were significantly increased (Fig. 2E,F and Supporting Fig. 7C-E). Taken together, these observations suggest that ACHE suppresses tumor formation in HCC.

Figure 2. ACHE inhibits HCC cell proliferation in vitro and tumorigenicity in vivo. (A) The recombinant ACHE protein inhibited and the ACHE inhibitor Neo enhanced the cell proliferation rate of SMMC-7721 cells. (B) ACHE overexpression inhibited and ACHE-shRNA enhanced the cell proliferation rate of SMMC-7721 cells. (C) ACHE overexpression decreased the volume of tumors in nude mice. (D) ACHE overexpression decreased the weight of tumors in nude mice. (E) ACHE-shRNA increased the volume of tumors in nude mice. (F) ACHE-shRNA increased the weight of tumors in nude mice. The values are expressed as means and standard errors of the mean. Abbreviation: OD_450nm, optical density at 450 nm.

Ach Promotes HCC Cell Proliferation and Activates Mitogen-Activated Protein Kinase (MAPK) and Phosphatidyl Inositol-3′-Phosphate Kinase (PI3K)/Protein Kinase B (Akt) Pathways.

Because Ach is the substrate of ACHE, we first determined the impact of Ach on liver cancer cells. The results showed that Ach could promote the proliferation of SMMC-7721 cells in a dose-dependent and time-dependent manner (Fig. 3A and Supporting Fig. 4E). In addition, Ach could trigger the activation of MAPK/extracellular signal-regulated kinase (ERK) and PI3K/Akt pathways and induce the phosphorylation of ERK1/2 and Akt, respectively; the triggering signaling pathway could be restricted by the nAChR antagonist mecamylamine (MEC) and the mAChR antagonist atropine (Fig. 3B; the densitometry analysis is shown in Supporting Fig. 8). Moreover, the inhibitors of Ach synthesis (hemicholinium-3, the choline transport inhibitor), Ach transport (vesamicol, the VAChT inhibitor), nAChR (MEC), and mAChR (atropine) could obviously inhibit SMMC-7721 cell proliferation (Fig. 3C,D). These results suggest that Ach is an effective regulator of HCC cell proliferation.

Figure 3. The disruption of Ach signaling significantly influences the proliferation of SMMC-7721 cells. (A) Ach enhanced the cell proliferation rate of SMMC-7721 cells. (B) Ach induced the phosphorylation of ERK1/2 and Akt, and the specific acetylcholine receptor antagonists repressed these effects after 30 minutes of treatment (representative data). MEC was used as the nAChR antagonist, atropine was used as the mAChR antagonist, and EGF was used as the positive control. The amount of ERK, Akt, and β-actin was detected as the loading control. (C) Hemicholinium-3 (the choline transport inhibitor) and vesamicol (the VAChT inhibitor) inhibited the cell proliferation rate. (D) MEC and atropine inhibited the cell proliferation rate of SMMC-7721 cells. The values are expressed as means and standard errors of the mean. Abbreviation: OD_450nm, optical density at 450 nm.

ACHE Suppresses Cell Proliferation by Reducing the Extracellular Ach Content.

To better understand the possible mechanism by which ACHE inhibits liver cancer cell viability, we first confirmed that ACHE also removes Ach from liver cancer cells. We found that the amount of Ach in human HCC tissues and xenograft models was inversely correlated with their ACHE expression levels (Fig. 4A and Supporting Fig. 7F). The Ach contents of different hepatic cancer cell lines were inversely correlated with their endogenous ACHE expression at both mRNA and protein levels (Supporting Fig. 9A,B); similar results were observed in the stable cell lines (Supporting Fig. 10). Besides, the Ach contents of HCC cell lines were inversely correlated with their ACHE activity (Fig. 4B). The addition of Neo, an inhibitor of ACHE activity, could significantly increase the Ach contents in the culture media of HCC cells (Fig. 4C). Besides Ach, some other cholinergic agonists resistant to degradation by ACHE, such as carbachol (also known as carbamylcholine), also had a promoting effect on SMMC-7721 cell proliferation (Supporting Fig. 9C,D). Furthermore, Ach enhancement of cell proliferation was abrogated in Lenti-ACHE cells, whereas carbachol enhancement functioned in both Lenti-ACHE and Lenti-vector cells (Fig. 4D,E). Taken together, these results indicate that the growth inhibition of HCC cells mediated by ACHE is associated with its degradation function for Ach.

Figure 4. The extracellular Ach content contributes to the growth suppression induced by ACHE in SMMC-7721 cells. (A) Inverse correlation between the ACHE mRNA level and the Ach content in human HCC tissues. Pearson's correlation was used. (B) Inverse correlation between the ACHE activity and the Ach content in HCC cell lines. Pearson's correlation was used. (C) Neo increased the Ach content in the cultured medium of the indicated cells. (D) Both Ach and carbachol enhanced the proliferation of Lenti-vector cells. (E) Carbachol enhanced the proliferation rate of Lenti-ACHE cells, but Ach did not. The values are expressed as means and standard errors of the mean. Abbreviation: OD_450nm, optical density at 450 nm.

ACHE Inactivates MAPK and PI3K/Akt Pathways and Activates Glycogen Synthase Kinase 3β (GSK3β) in HCC Cells.

Because ACHE could significantly inhibit liver cancer cell proliferation in vitro and in vivo, we examined its intracellular signaling pathways involved in cell growth. Surprisingly, the activation patterns of ERK1/2 and Akt induced by growth stimuli were quite different between Lenti-ACHE cells and Lenti-vector cells. The phosphorylation levels of ERK1/2 and Akt induced by epidermal growth factor (EGF) and Ach were lower in Lenti-ACHE cells. In addition, Neo treatment remarkably increased the Ach-induced phosphorylation of ERK1/2 and Akt in both Lenti-ACHE and Lenti-vector cells (Fig. 5A; the densitometry analysis is shown in Supporting Fig. 11).

Figure 5. ACHE inactivates MAPK and PI3K/Akt pathways and activates GSK3β in SMMC-7721 cells. (A) Phosphorylation patterns of ERK1/2 and Akt in Lenti-ACHE and Lenti-vector cells after 30 minutes of the indicated treatment (representative data). Wm was used as the PI3K inhibitor, Neo was used as the ACHE inhibitor, and U0126, used as the MEK1/2 inhibibtor. The amount of ERK, Akt, and β-actin was detected as the loading control. (B) ERK1/2 and Akt activation status, GSK3β activation status, and β-catenin and cyclin D1 expression levels in the cells treated as indicated for 24 hours (representative data). The amount of ERK, Akt, β-actin, and α-tubulin served as the internal control. (C) ACHE overexpression affected the cell cycle distribution. (D) GSK3β inhibitors (ARA014418 and TDZD8) rescued ACHE-induced HCC cell growth suppression. The cells were treated as indicated for 3 days before the analysis. The values are expressed as means and standard errors of the mean. Abbreviation: Wm, wortmannin.

Because GSK3β is well known to be the downstream target of Akt and ERK, we analyzed whether ACHE affects the activation status of GSK3β. With the inactivation of ERK1/2 and Akt in Lenti-ACHE cells, the phosphorylation level of GSK3β was reduced, and this resulted in more GSK3β activation as well as β-catenin degradation; this was accompanied by the down-regulation of cyclin D1 expression (Fig. 5B; the densitometry analysis is shown in Supporting Fig. 12) and thus led to the cell cycle redistribution of Lenti-ACHE cells (Fig. 5C). In addition, the effects of ACHE overexpression on the activation of ERK and Akt and the inactivation of GSK3β could be reversed by the addition of carbachol (Fig. 5B). Furthermore, GSK3β inhibitors could effectively rescue ACHE-mediated HCC cell growth suppression (Fig. 5D). These results suggest that ACHE might inhibit cell proliferation via the activation of the GSK3β pathway in HCC cells.

ACHE Enhances Cell Apoptosis Induced by Chemotherapeutic Drugs and Results in Increased Drug Sensitivity in HCC Cells.

Because GSK3β is also involved in apoptosis, we attempted to estimate the effects of ACHE on GSK3β activation induced by chemotherapeutic drugs. When SMMC-7721 cells were incubated with either adriamycin (ADR) or etoposide (VP-16), drug-induced GSK3β activation could be enhanced by the recombinant ACHE protein (Fig. 6A; the densitometry analysis is shown in Supporting Fig. 13A). The endogenous ACHE overexpression also led to the enhancement of drug-induced GSK3β activation (Fig. 6B; the densitometry analysis is shown in Supporting Fig. 13B). Moreover, the apoptotic level represented by caspase-3 and caspase-7 activity showed similar results (Supporting Fig. 14). Furthermore, we found that the recombinant ACHE protein could sensitize SMMC-7721 cells to ADR or VP-16 and potentiate the therapeutic effects with an almost 50% reduction of the median inhibitory concentration (IC₅₀; Fig. 6C,D). The IC₅₀ values of these chemicals in Lenti-ACHE cells were also reduced in comparison with the Lenti-vector control (Fig. 6E,F). These results indicate that ACHE can increase apoptotic cell death induced by chemotherapeutic drugs and enhance the drug sensitivity of liver cancer cells.

Figure 6. ACHE enhances the chemosensitization of SMMC-7721 cells. (A) The recombinant ACHE protein promoted the GSK3β activation induced by the indicated drugs. (B) ACHE overexpression promoted the GSK3β activation induced by the indicated drugs. (C) Potentiation of the cytotoxicity of ADR by the recombinant ACHE protein. (D) Potentiation of the cytotoxicity of VP-16 by the recombinant ACHE protein. (E) Potentiation of the cytotoxicity of ADR in Lenti-ACHE cells. (F) Potentiation of the cytotoxicity of VP-16 in Lenti-ACHE cells. The values are expressed as means and standard errors of the mean.

Discussion

The presence of neurotransmitters has been proved not to be a privilege of the nervous system. At a later evolutionary stage, the transmitters are derived from both the nervous system and nonneuronal tissues or organs.14 In the present study, we have demonstrated that all the molecular equipment of the cholinergic system is expressed in both the human liver and HCC. Among the examined molecules, only ACHE, the Ach hydrolytic enzyme, was obviously down-regulated in almost 70% of detected cancer tissues from 240 HCC patients. Despite much evidence suggesting the abnormal expression of the ACHE gene in various tumor types,9 no causal relationship has been established between the ACHE expression level and tumor development, except in the case of astrocytoma.15 An ACHE-derived peptide displayed growth factor activity in hematopoiesis,16 and a role of ACHE as a myeloid tumor suppressor gene has been postulated.17 Besides, the long-term treatment of cholinesterase inhibitors increased the rate of mammary carcinoma formation in rats.18 In our study, we found that HCC patients with low ACHE expression in their tumors had a significantly high risk of cancer recurrence and poor median OS and DFS. The correlation with clinicopathological features further strengthens the idea that the ACHE expression level is an independent prognostic factor for a significant reverse correlation with well-established factors such as tumor size, multiplicity, encapsulation, and TNM stage. Therefore, we report that dampened ACHE expression in HCC is significantly associated with tumor aggressiveness and an enhanced risk of a poor prognosis. Although some reports have also evaluated ACHE as a prognostic factor for cirrhosis,19 a discrepancy remains in our results, and there was no significant correlation between cirrhosis and ACHE expression in this study.

Moreover, we found ACHE to be a growth suppressor in HCC. Superfluous ACHE can remarkably inhibit HCC cell proliferation both in vitro and in vivo. We also confirmed that extracellular ACHE can effectively inhibit HCC cell growth by reducing the Ach content in the culture medium. Although there was no statistical difference in the amount of nAChR or mAChR between HCC and CNL tissues, because of the important role of nAChR in other tumor types,5 the activated pentamer should be further analyzed in HCC cells to determine its function.20, 21

We have further demonstrated that ACHE overexpression can effectively inactivate MAPK and PI3K/Akt signaling pathways in HCC cells. Because MAPK and Akt are pivotal growth-promoting signaling modulators,22, 23 we postulate that the inhibitory effects of ACHE on cell proliferation might be mediated by the inactivation of MAPK and Akt. Moreover, it has been proposed that the activation of GSK3β can be inhibited by the PI3K/Akt pathway,24 and Erk serves as an adaptor inducing GSK3β degradation25; therefore, it is not surprising to find GSK3β activation in Lenti-ACHE cells. In addition, increasing evidence shows that GSK3β is an important apoptotic regulator that contributes to chemotherapy-induced cytotoxicity.26 Because ACHE can decrease the phosphorylation levels of ERK and Akt and result in an increase in the activation status of GSK3β in HCC cells, we postulate that GSK3β might also be involved in ACHE-enhanced apoptotic cell death and drug sensitivity induced by chemotherapeutic drugs.

In conclusion, the present study has for the first time demonstrated that ACHE acts as the key prognosis predictor for HCC patients, regulates HCC cell proliferation in vitro and tumorigenicity in vivo, and enhances drug-induced cell apoptosis and thereby increases the drug sensitivity of HCC cells. All these data can explain why a loss of ACHE leads to clinical features of high aggressiveness and poor prognosis. Our study provides new insight into the application of ACHE as a predictor of clinical outcomes and as a novel strategy in drug design for improving HCC treatment.

Acknowledgements

The authors are most grateful to Dr. T. Didier for the gift of the pWPXL, psPAX2, and pMD2.G lentivirus plasmids.