Acetylcholine Is Synthesized by and Acts As an Autocrine Growth Factor for Small Cell Lung Carcinoma1

Home , Vesamicol

[CANCER RESEARCH 63, 214–221, January 1, 2003] Acetylcholine Is Synthesized by and Acts as an Autocrine Growth Factor for Small Cell Lung Carcinoma1

Pingfang Song, Harmanjatinder S. Sekhon, Yibing Jia, Jennifer A. Keller, Jan Krzysztof Blusztajn, Gregory P. Mark, and Eliot R. Spindel2 Departments of Pathology [H. S. S.] and Behavioral Neuroscience [G. P. M.], Oregon Health and Science University, Portland, Oregon 97239; Department of Pathology and Laboratory Medicine, Boston University School of Medicine, Boston, Massachusetts 02118 [J. K. B.]; and Division of Neuroscience, Oregon National Primate Research Center, Beaverton, Oregon 97006 [H. S. S., P. S., Y. J., J. A. K., E. R. S.]

ABSTRACT need for further understanding of the biology of SCLC and development of new therapeutic approaches for SCLC treatment. It is well established that small cell lung carcinomas (SCLCs) express The neuroendocrine nature of SCLC is well established, as is the receptors for acetylcholine (ACh) and that stimulation of these receptors concept of growth regulation of SCLC by autocrine growth factors by nicotine or other cholinergic agonists stimulates cell growth via activation of nicotinic cholinergic receptors (nAChRs) and/or muscarinic such as gastrin-releasing peptide (4, 5). SCLC and some non-SCLCs cholinergic receptors (mAChRs). The aim of this study was to determine secrete a variety of neuropeptides, and many of these act as growth whether SCLC cells synthesize and secrete ACh and respond to endoge- factors (6, 7). The concept of autocrine growth factors has been nous ACh to create a functioning cholinergic autocrine loop. Reverse extended to the secretion of ligands for tyrosine- and threonine- transcription-PCR was used to screen a panel of SCLC cell lines for kinase-linked receptors such as basic fibroblast growth factor (bFGF, components of cholinergic signaling. Choline acetyltransferase (ChAT) FGF-2; Refs. 8), epidermal growth factor (EGF; Ref. 9) and trans- and the vesicular ACh transporter (VAChT), as well as ␣3, ␣5, ␣7, ␤2, forming growth factor ␤-1 (10). Therapeutic approaches derived from ␤ and 4, nAChR subunits and M3 and M5 mAChRs, were found to be this have included monoclonal antibodies against the epidermal present in most of the SCLC cell lines tested. Real-time PCR showed that growth factor receptors (11), broad spectrum neuropeptide antagonists mRNA levels for ChAT, VAChT, and ␣7 and ␤2 nAChR subunits varied significantly among different SCLC cell lines tested. The H82 cell line was (12), and inhibitors of tyrosine kinases and phosphatases (13). Thus, found to express the highest levels of ChAT, and that cell line was chosen autocrine growth factors can regulate SCLC growth and are potential for additional studies of ACh release and cell proliferation. ACh was easily therapeutic targets. detectable in H82 cell culture media, and levels of ACh were increased by Multiple reports have established that SCLC cells express nAChRs the acetylcholinesterase inhibitor neostigmine. Vesamicol, an inhibitor of and mAChRs (14–17) and that the activation of nAChR and/or VAChT, and hemicholinium-3, an inhibitor of choline transport, both mAChR with nicotine (18–20), ACh (19), and muscarine (19, 20) reduced H82 cell ACh basal release in a dose-dependent manner. In stimulates the growth of SCLC cells. Recent reports that a variety of parallel with the reductions of ACh release, vesamicol and hemicho- cell types in normal lung synthesize ACh (21–23) have led us to linium-3 also decreased H82 cell proliferation. H82 cell proliferation was hypothesize that lung cancers, similarly, may make ACh and that, also inhibited by the muscarinic and nicotinic antagonists atropine and mecamylamine, respectively, in dose- and time-dependent manners. Fi- therefore, SCLC growth may be regulated by a cholinergic autocrine nally, archival cases of SCLC were screened by immunohistochemistry for loop. expression of ChAT. Thirteen of 26 tumors screened were positive for In cholinergic neurons, the neurotransmitter ACh is synthesized ChAT. These findings demonstrate that SCLC can synthesize, secrete, and from choline and acetyl-CoA by ChAT (24) and is then translocated degrade ACh and that released ACh stimulates SCLC cell growth. Iden- into synaptic vesicles by the VAChT (25). In neurons, choline for tification of this new autocrine loop provides a potential new target for synthesis of ACh is transported by a specific high-affinity choline therapeutic intervention. transporter: CHT1 (26, 27). If SCLCs synthesize ACh, then ChAT must be present and the other components of neuronal cholinergic signaling may be present. If a cholinergic autocrine loop is present in INTRODUCTION SCLC, then interruption of ACh synthesis and signaling should mod- Global lung cancer deaths were over 1,000,000 for the year 2000 ify growth of SCLC. In this article, we show that SCLC tumors and and are projected to increase to 2,000,000 per annum by 2020–2030 cell lines express ChAT and that SCLC cell lines synthesize VAChT, (1). SCLC3 accounts for 20–25% of all new cases of primary lung CHT1, nAChR, and mAChR and synthesize ACh and that interruption cancer. In addition to morphological characteristics, SCLC is distin- of cholinergic signaling affects cell growth. guished from non-SCLC by its rapid tumor doubling time, high growth fraction, and early development of widespread metastasis (2). Overall, survival beyond 5 years occurs in only 3–8% of all patients MATERIALS AND METHODS with SCLC because relapse occurs within 2 years despite initial responses to chemotherapy and radiotherapy (3). Thus, there is clear Cell Culture and SCLC Tissue Samples. SCLC cell lines NCI H69, H82, H209, H345, H378, H417, H510, H592 were generously provided by Phelps et al. and Carney et al. (28, 29). H82 cells were grown in RPMI 1640 (Invitrogen, Received 6/27/02; accepted 10/31/02. Carlsbad, CA) supplemented with 5 ␮g/ml insulin, 5 ␮g/ml transferrin, 5 ng/ml The costs of publication of this article were defrayed in part by the payment of page ␮ charges. This article must therefore be hereby marked advertisement in accordance with sodium selenite, 100 units/ml penicillin, and 100 g/ml streptomycin. These 18 U.S.C. Section 1734 solely to indicate this fact. lines encompass both classic, neuroendocrine-differentiated SCLC cell lines 1 Supported by NIH Grants RR00163, CA69533, HD/HL37131, AG09525, DA11203, (H69, H209, H345, H378, H510, H592) and variant SCLC cell lines that lack and NS42793. neuroendocrine differentiation (H82, H417; Ref. 29). In addition, these lines 2 To whom requests for reprints should be addressed, at Division of Neuroscience, Oregon National Primate Research Center, Beaverton, OR 97006. Phone: (503) 690-5512; have previously been reported to express nicotinic and/or muscarinic receptors Fax: (503) 690-5384; E-mail: [email protected]. (17, 18, 30). Paraffin-embedded bronchoscopic biopsies of SCLC were ob- 3 The abbreviations used are: SCLC, small cell lung carcinoma; ACh, acetylcholine; tained from the Department of Pathology of the Oregon Health and Science ChAT, choline acetyltransferase; VAChT, vesicular ACh transporter; nAChR, nicotinic University. Five-␮m sections were cut, and 1 section was stained with H&E to ACh receptor; mAChR, muscarinic ACh receptor; RT, reverse transcription; HPLC, high-performance liquid chromatography; LEMS, Lambert-Eaton myasthenic syndrome; confirm the diagnosis; then other sections were processed for immunohisto- AEC, 3-amino-9-ethylcarbazole; MTS, methyltetrazolium. chemistry as described below. 214

Immunohistochemistry and Immunocytochemistry. Paraffin sections of GAAGTT, and TaqMan probe: TCATCTGCACCATCGGCATCCTGA; ␤2 SCLC were processed for immunohistochemistry as described previously (31). nAChR: 5Ј, GGTCCACGAACGGAACTTCA, 3Ј, CTGCCGCCTGCCATC- Antibodies used were mouse anti-ChAT (mAB 305; Chemicon International, TAC, and TaqMan probe GCACTTCCCATTTGACCAGCAGAACTG. 18S Inc.; 1:400), rabbit anti-VAChT (H-V005; Phoenix Pharmaceuticals), and rat RNA amplifications were conducted with the Pre-Developed TaqMan Assay anti-␣7 nAChR (mAB 319; 1:250; Ref. 32). Immunohistochemistry for SCLC Reagent (Applied Biosystems). All real-time PCR reactions were run in was performed using Vector ABC reagents and VIP as a chromogen (Vector triplicate. Laboratories, Burlingame, CA) for ChAT and AEC for ChAT. VIP is a Sequence Analysis of ChAT. Two ␮g of total RNA prepared from H82 proprietary peroxidase substrate from Vector Laboratories that yields a purple cells was reverse transcribed as described above. Primers were used to amplify color. Intensity of immunohistochemical staining was scored from 0 to 4ϩ by cDNAs spanning exons 5 to 11 and 11 to 18 to span the entire ChAT coding two independent readers (E.R.S. and H.S.S.) (where 0 ϭ no staining; region (24). Amplified cDNA bands were gel isolated and subcloned into 1ϩϭfocal weak staining; 2ϩϭfocal strong staining or diffuse weak pGEM-T (Promega Corp., Madison, WI) and sequenced. The 5Ј and 3Ј primers staining; 3ϩϭdiffuse medium staining; and 4ϩϭdiffuse strong staining). used to amplify from exon 5 to exon 11 of ChAT were TCCACACCTCTG- Fluorescent immunohistochemistry on H82 cells was performed using Texas- CATCCCTG and GACTTGTCGTACCAGCGATT; the 5Ј and 3Ј primers red conjugated second antibody. used to amplify from exon 11 to exon 18 of ChAT were ACCGGGACTCGCT- RT-PCR. RT-PCR and Southern hybridization were used to investigate the GGACATG and GGAGGTGAAACCTAGTGGCA. expression of nAChR, mAChR, ChAT, VAChT, and CHT1 genes in SCLC cell ACh Assay. For investigation of ACh release from H82 cells, 5 ϫ 106 cells lines. For nAChR, ␣3, ␣5, ␣7, ␤2, and ␤4 subunits were examined. For in 10 ml were plated in 60-cm2 culture dishes. After plating, the acetylcho- mAChR, M3 and M5 subtypes were examined. RNA was isolated from SCLC linesterase inhibitor neostigmine was added at a concentration of 5 ϫ 10-5 M cell lines with Tri Reagent (Molecular Research Center, Cincinnati, OH). (Sigma, St. Louis, MO) to inhibit ACh degradation. Drugs tested were the RT-PCR, gel electrophoresis and hybridization were performed as described VAChT inhibitor (Ϯ)-vesamicol (Sigma), the choline transport inhibitor, previously (31). Primers for PCR and internal primers for probes that were hemicholinium-3 (ICN Biomedicals Inc., Aurora, OH) for 24 h. After incuba- used are shown in Table 1. tion, cell suspensions were transferred to 15-ml centrifuge tubes and centri- Real-Time PCR Analysis. Real-time PCR was used to quantify ChAT, fuged at 1000 rpm for 2 min at 4°C. Supernatants were aliquoted, rapidly VAChT, and ␣7 and ␤2 nAChR receptor subunit mRNA levels in SCLC cell frozen on dry ice, and stored at Ϫ80°C. For ACh assay, supernatants were lines H82, H69, H345, H378, H417, and H592. Procedures were as described thawed, and 20 ␮l of supernatant were injected directly into the HPLC for previously (33) and primers for 18S RNA were used for standardization. The enzyme-coupled electrochemical detection as described previously (34). Each primers and TaqMan probes were designed using Primer Express software sample was assayed at least in duplicate. The detection limit, defined as the (Applied Biosystems, Foster City, CA). Primers used were as follows: ChAT, amount of ACh needed to achieve a 3:1 signal:noise ratio, was 30 fmol per 5Ј primer, GGCTCAGAACAGCAGCATCA, 3Ј primer, TGAGACGGCG- 20-␮l injection. GAAATTAATGA, and TaqMan probe, CACGTCATCGTAGCCTGCTG- Cell Proliferation Assay. H82 cells were used to evaluate the effects of CAATCA; VAChT: 5Ј, CAGCGGCACGTCGTAGCT, 3Ј, TTGCTTCCAAG- endogenous ACh synthesis on cell proliferation because real-time PCR showed GCTATCCT, and TaqMan probe, AGCGGGCCCTTCATCCGACCG; ␣7 they had the highest levels of ChAT expression of the SCLC cell lines tested. nAChR: 5Ј, CATGGCCTTCTCGGTCTTCA, 3Ј, CACGGCCTCCAC- Drugs tested were the nAChR antagonist mecamylamine at 10Ϫ7,10Ϫ6, and

Table 1 Primers used for RT-PCR Gene Accession no. Primers Length (bp) ChAT AF-305906 5Ј-GGAGATGTTCTGCTGCTATG 280 3Ј-GGAGGTGAAACCTAGTGGCA Internal: TGGAAGAAAGCCTCATTGAC ChAT-N AF-305894 5Ј-CGTTGAGGCTTTGAGAAAGG 88 3Ј-ATTTCCAGAGGTCGGACTCT Internal: CATTCCGGCAGAGGAAGAAA ChAT-R AF-305893 5Ј-TCAACGCCGCTCTGGGGACG 105 3Ј-CTTCCTTCTGAGAGGGCCGC Internal: GCCCAGCGCACAGCCTGGGC ChAT-S AF305894 5Ј-TGAGGGAATTCACCCTACTC 70 3Ј-CTACTGTGCTCAGTGCTTCA Internal: CACCAGAGATGTGGCCGGAA VAChT XM࿝005733 5Ј-GACGTGAAGATCGGGGTGCTGTT 565 3Ј-CGAGGAAGGCGAGGGGAATGTTAC Internal: GCTGTTCGCGGCGCGCAGCC CHT1 AF276871 5Ј-TGGCCTACACTGATGTCGTT 670 3Ј-TACCAGAGCCCATACACAGT Internal: CACCTATGCTCAAGTGCTGT M3-AChR HSU 29589 5Ј-TGGAACAACAATGATGCTGC 432 3Ј-CCTTTTCCGCTTAGTGATCTG Internal: GCCGTGGACACAGCTAAGAC M5-AChR HUMACHRM 5Ј-ATCATGCCCTGCCCCTTCC 391 3Ј-GTAGCTTGCTGTTCCCCTGCC Internal: CCAGACACTGAGTGCCATTC ␣3-nAChR HSU62432 5Ј-GTGTTTCTCCTGGTGATCACTGAGACCATC 756 3Ј-CTTGATAACAGAAAGTACGTTCTTACTAGG Internal: GCTTCAGCCGCGCAGAGTCC ␣5-nAChR XM࿝007577 5Ј-AATGGGAGATTGTGAGTGCA 685 3Ј-CCCAAGAGATCCAACAATTG Internal: GGAGAGTATCTGGTATTTAC ␣7-nAChR XM࿝044096 5Ј-GCCGCAGGACGCTCTACTAT 432 3Ј-ACGGCACTCATCTCCACACT Internal: CCATGATCATCGTGGGCCTC ␤2-nAChR NM࿝000748 5Ј-TGTGCTCATCACCTCGCTAG 307 3Ј-GAGCAGCGCGGGCAGCTTCTC Internal: ATGGCGCCCTGGGTGAAGGTC ␤4-nAChR XM࿝007575 5Ј-GGCCATCCTCGTCTTCTACC 677 3Ј-CAGTGCCCAGGACGCACACA Internal: GCACTGACATTCTTCCTGCTGCTCATCTCC 215

based on the genomic structure of ChAT as described by Ohno et al. (24). As shown in Fig. 2A, primers from exon 5–11 amplified 2 bands. The predominant band corresponded exactly to the size predicted from the sequence of neuronal ChAT. A second smaller band was also amplified. Sequence analysis of the amplified bands showed that the sequence of the larger band was identical to that of neuronal ChAT, whereas the smaller band corresponded to neuronal ChAT missing exon 10. As shown in Fig. 2B, the primers from exon 11 to exon 18 amplified a single band that corresponded to the size predicted from neuronal ChAT. Sequence analysis of this band showed it to be identical to neuronal ChAT. Thus the ChAT protein expressed by H82 cells is identical in sequence to the ChAT expressed in neuronal tissue. Expression of mAChR and nAChR in SCLC Cell Lines. All of the SCLC cell lines except H378 expressed the ␣7, ␣5, and ␤2 nAChR subunit mRNAs (Fig. 3A). ␣3 nAChR was expressed in SCLC cell lines H69, H82, H345, H417, H592, was absent in H378, and was not tested in H209 and H510. ␤4 nAChR was expressed in SCLC cell lines H69, H82, H209, and H345, but not in H378, H417, Fig. 1. Expression of ChAT, VAChT, CHT1 mRNA in SCLC cell lines. RT-PCR and Southern hybridizations were performed on total RNA prepared from the indicated cell H510, and H592. The M3 mAChR and M5 mAChR were expressed lines. Primers and probes are described in “Materials and Methods.” A, ChAT, VAChT, in all of the SCLC cell lines tested (Fig. 3B). and CHT1 expression in SCLC cell lines H69, H82, H345, H378, H417, and H592. Quantitative Analysis of ChAT, VAChT, ␣7, and ␤2 nAChR Human basal forebrain RNA was used as a positive control (HBF343, HBF345). B, ChAT isoform gene expression. ChAT N, S, R isoform genes were expressed in all of the SCLC Subunit mRNA. To quantify levels of mRNA expression and to cell lines tested. ChAT M isoform was not expressed in any of the SCLC cell lines (data further confirm expression, real-time PCR was performed for selected not shown). cholinergic mRNAs. As shown in Table 2, real-time PCR with different primer sets confirmed the expression observed by conventional RT-PCR. The quantitation achievable with real-time PCR also Ϫ Ϫ Ϫ Ϫ 10 5 M; the mAChR antagonist atropine at 10 8,10 7, and 10 6 M; nicotine Ϫ Ϫ Ϫ Ϫ Ϫ showed that there was significant variation in levels of ChAT, at 10 8,10 7,10 6, and 10 5 M; carbachol, a stable ACh analogue, at 10 6 ␣ ␤ Ϫ Ϫ Ϫ Ϫ VAChT, 7 nAChR, and 2 nAChR mRNA among the SCLC cell and 10 5 M; the VAChT inhibitor vesamicol at 10 7,10 6, and 10 5 M; and Ϫ Ϫ Ϫ Ϫ lines tested. Relative levels of ChAT mRNA in cell lines tested were the choline transport inhibitor hemicholinium-3 at 10 8,10 7,10 6, and 10 5 as follows: H82ϾH69ϾH417ϾH345ϾH378ϾH592. ChAT mRNA M. All of the experiments for each drug were performed at least twice with a minimum of 12 replicates per data point per experiment. All of the drugs were was highest in H82 cells and more than 70 times greater than from Sigma. H82 cells were plated with an 8-channel pipette at 5000 cells/well the lowest, H592, cells. Relative levels of VAChT were in 96-well plates. Drugs were added immediately after cell plating. The final H417ϾH345ϾH82ϾH592ϾH69ϾH378. ␣7 nAChR subunit mRNA medium volume of each well was 200 ␮l. Every 3 days, one-half of the volume levels were H82ϾH345ϾH592ϾH417ϾH69ϾH378, and ␤2 nAChR of the media and drugs were changed after centrifugation of the plates at 1000 subunit mRNA levels were H592ϾH345ϾH82ϾH417ϾH69ϾH378. rpm for 1 min. At 0, 6, 9, and 12 days of incubation, a MTS-based-assay was ACh Secretion by H82 Cells. Synthesis and secretion of ACh by used to measure cell growth. Twenty ␮l of MTS reagent (cellTiter 96 AQue- H82 cells was confirmed by HPLC with enzymatically coupled elec- ous; Promega Corporation, Madison, WI) were added per well, and absorbance at 490 nm was recorded 2 h later. Statistical Analysis. All of the data are presented as mean Ϯ SE. Data of cell growth were analyzed by ANOVA followed by Neuman-Keuls multiple- comparison test using NCSS 2001 (Kaysville, UT) statistical software. Stu- dent’s t test was used for data analysis of ACh release.

RESULTS Identification of ChAT, VAChT, and ChT1 Gene Expression in SCLC Cell Lines. If a cholinergic autocrine loop is functional in SCLC, then proteins to synthesize ACh must be present and cells must express receptors for ACh. To determine whether SCLC cell lines express the proteins needed to synthesize ACh, ChAT, VAChT, and CHT1 gene expression was examined by RT-PCR in SCLC cell lines H69, H82, H345, H378, H417, and H592. All of the tested SCLC cell lines expressed ChAT and VAChT although only H345 cells expressed CHT1 (Fig. 1A). In the brain, multiple forms of ChAT are expressed. As shown in Fig. 1B, mRNA coding for three ChAT isoforms, N, R, and S, was expressed in all of the SCLC cell lines. ChAT isoform M Fig. 2. PCR amplification of ChAT from H82 RNA. Upper panel, strategy used to was not expressed in any of the SCLC cell lines (data not shown). amplify the ChAT mRNA. The entire coding region was amplified in two overlapping Sequence Analysis of H82 Cell ChAT. To determine whether the segments, A and B: segment A, the amplification of exons 5–11; segment B, the amplification of exons 11–18. UT, untranslated. A, RT-PCR from exon 5 to exon 11. B, RT-PCR ChAT mRNA expressed in SCLC was similar to neuronal ChAT, from exon 11 to 18. In A, the 987-bp band is exactly the predicted size based on neuronal cDNAs spanning the entire coding region of ChAT were amplified, ChAT. The 810-bp band corresponds to authentic ChAT missing exon 10. In B, the 1151-bp band is exactly the predicted size based on neuronal ChAT. The identity of the subcloned, and sequenced from H82 mRNA. This was done in two amplified bands was confirmed by subcloning and sequence analysis as described in overlapping amplifications spanning exons 5–11 and 11–18 (Fig. 2) “Materials and Methods.” 216

cell growth only at 6 days (P Ͻ 0.05). As shown in Fig. 5B, the mAChR antagonist atropine also inhibited H82 cell growth in a concentration-dependent and time-dependent manner (P Ͻ 0.05). A Ϫ6 concentration of 10 M atropine significantly inhibited H82 cell growth at 6, 9, and 12 days (P Ͻ 0.05). As shown in Fig. 5C, the VAChT inhibitor vesamicol decreased H82 cell growth in a concentration-dependent and time-dependent manner (P Ͻ 0.05). Concen- Ϫ7 Ϫ6 Ϫ5 tration of 10 ,10 , and 10 M vesamicol significantly inhibited H82 cell growth at 9 and 12 days (P Ͻ 0.05). Similarly, the choline transport inhibitor hemicholinium-3 inhibited cell proliferation in a time- and dose-dependent manner (Fig. 5D). Showing that these were not just nonspecific toxic effects, the nicotinic agonist nicotine and the muscarinic agonist carbachol significantly stimulated cell growth in time- and dose-dependent manners (Fig. 5, E and F). ChAT Expression in SCLC Tumor Biopsies. Immunohisto- chemistry was performed to determine whether immunoreactive ChAT could be detected in paraffin sections of SCLC biopsies. As shown in Fig. 6, ChAT could be readily detected in SCLC. In 14 of 26 SCLC tumor biopsies tested, ChAT immunoreactivity was detected (Fig. 6, A–C) with average immunostaining intensity of 2ϩ.Asa positive control, ␣7 nAChR was tested and found to be expressed in 24 of 26 SCLC tested with average immunostaining intensity of 2ϩ. All of the tumors that expressed ChAT also expressed ␣7 nAChR, although there was no significant correlation between the intensity of Fig. 3. Expression of mAChR and nAChR mRNA in SCLC cell lines. RT-PCR and ChAT immunostaining and ␣7 immunostaining. ChAT immunoreac- Southern hybridizations were performed on total RNA prepared from the indicated cell lines. Primers and probes are described in “Materials and Methods.” A, nAChR expres- tivity was also detected in H82 cells (Fig. 6D). H82 cells and tumor sion. Expression of ␣3, ␣5, ␣7, ␤2, and ␤4 nAChR subunit mRNA as shown. B, mAChR samples also expressed VAChT immunoreactivity (data not shown). expression. M3 mAChR and M5 mAChR were expressed in all of the SCLC cell lines Thus, these data demonstrate that SCLC tumors as well as SCLC cell tested. ND, not done. lines express ChAT. trochemical detection. As shown in Fig. 4, A and B, ACh was easily DISCUSSION detected in supernatants of H82 cell cultures. No ACh was detected in Ϫ5 the media prior to cell culture. Addition of 5 ϫ 10 M neostigmine This study presents data that SCLC express a cholinergic autocrine significantly increased media ACh levels (P Ͻ 0.001). This confirms loop that can regulate cell growth. The data presented here demon- the synthesis of ACh by H82 cells and also suggests the presence of strate that: 1, genes for all components of an ACh autocrine loop acetylcholinesterase activity in the H82-conditioned culture medium. including ChAT, VAChT, CHT1, nAChR and mAChR are expressed ACh was also readily detectable in supernatants of H345 cell cultures, in SCLC cells; 2, ChAT immunostaining is present in biopsies of thus demonstrating that ACh is secreted by both classic and variant SCLC and in SCLC cell lines; 3, SCLC cells are able to synthesize, SCLC cell lines. Next, vesamicol, an inhibitor of VAChT, was used to secrete, and degrade ACh; and 4, SCLC cell growth is modulated by determine the role of VAChT in ACh release from H82 cells. As endogenous ACh synthesis. To our knowledge, this is the first study shown in Fig. 2C, vesamicol reduced H82 cell ACh basal release in a to demonstrate that SCLC cells have a cholinergic phenotype and that Ϫ6 Ϫ5 concentration-dependent manner. Vesamicol (10 and 10 M) sig- ACh exerts as an autocrine growth factor in human lung tumors. nificantly decreased H82 cell ACh release (P ϭ 0.012 and P ϭ 0.013, For SCLC to express a cholinergic autocrine loop, ACh must be respectively). suggesting the presence of a vesamicol-sensitive synthesized and secreted. The enzyme ChAT is required for ACh VAChT activity in H82 cells. However, the ACh decrease caused by synthesis and, therefore, the expression of ChAT in SCLC must be Ϫ5 10 M vesamicol was only 40%, suggesting that a vesamicol-inde- definitively demonstrated to prove the existence of the autocrine loop. pendent pathway for H82 ACh release may also exist in SCLC cells. Because of its importance, we have, therefore, demonstrated the Next, the sensitivity of ACh secretion to hemicholinium-3 was deter- expression of ChAT in SCLC by multiple techniques. The presence of mined. As shown in Fig. 4D, hemicholinium-3 reduced H82 cell basal ChAT mRNA has been demonstrated by both conventional and real- ACh release in a concentration-dependent manner. Concentrations of time PCR with multiple independent primer sets. SCLC ChAT has Ϫ7 Ϫ6 Ϫ5 10 ,10 , and 10 M hemicholinium-3 resulted in the significant been amplified and sequenced to show that the mRNA is authentic decrease of H82-cell basal ACh release in the presence of 5 ϫ 10Ϫ5 ChAT. The presence of ChAT protein has been demonstrated by M neostigmine (P Ͻ 0.010, P Ͻ 0.047, and P Ͻ 0.037, respectively), suggesting the presence of a hemicholinium-3-sensitive, choline trans- Table 2 ChAT, VACHT, ␣7- and ␤2-nAChR subunit mRNA levels in SCLC cell lines porter in H82 cells. RNAs levels in the cell lines shown were quantified by real-time PCR as described in Regulation of H82 Cell Growth. The role of endogenously syn- “Materials and Methods.” Levels were standardized to 18S RNA and are expressed as thesized ACh in regulating H82 cell growth was determined with mean Ϯ SE. Comparisons are valid within RNAs but not between RNAs. inhibitors of ACh synthesis and with cholinergic receptor antagonists. SCLC cell line ChAT VAChT ␣7-nAChR ␤2-nAChR As shown in Fig. 5A, the nAChR antagonist mecamylamine inhibited H69 4.26 Ϯ 1.44 3.81 Ϯ 0.19 0.03 Ϯ 0.01 11.87 Ϯ 6.53 H82 cell growth in a concentration-dependent and time-dependent H82 7.15 Ϯ 0.81 8.82 Ϯ 0.67 1.86 Ϯ 0.40 35.81 Ϯ 25.61 Ϯ Ϯ Ϯ Ϯ Ͻ Ϫ5 H345 2.71 0.97 18.50 2.61 1.51 0.20 73.88 22.56 manner (P 0.05). A concentration of 10 M mecamylamine sig- H378 0.22 Ϯ 0.00 0.79 Ϯ 0.53 0.002 Ϯ 0.000 2.17 Ϯ 1.23 nificantly inhibited H82 cell growth at 6, 9, and 12 days (P Ͻ 0.05). H417 3.82 Ϯ 1.81 34.43 Ϯ 2.16 0.48 Ϯ 0.12 22.59 Ϯ 4.48 Ϫ6 H592 0.10 Ϯ 0.07 8.45 Ϯ 1.01 0.73 Ϯ 0.20 93.07 Ϯ 48.34 A concentration of 10 M mecamylamine significantly inhibited H82 217

Fig. 4. ACh release from H82 cells. H82 cells were plated as described in “Materials and Meth- ods.” Drugs were added immediately after plating and cells incubated for 24 h; then media were removed for HPLC assay of ACh. A, HPLC chromatogram from cells cultured in the absence of neostigmine. B, HPLC chromatogram from cells Ϫ5 cultured in the presence of 5 ϫ 10 M neostigmine. C, effect of vesamicol on ACh levels in medium from cultured H82 cells. Drugs were added immediately after plating at the concentration shown and media was removed for assay 24 h later. Neostigmine (Neo) concentra- Ϫ5 tion ϭ 5 ϫ 10 M.(n ϭ 6). D, effect of hemicholinium-3 on ACh levels in medium from cultured H82 cells. Drugs were added immediately after plating at the concentration shown and medium was removed for assay 24 h later. Neostigmine Ϫ5 (Neo) concentration ϭ 5 ϫ 10 M (n ϭ 4). Data are represented as mean Ϯ SE. ✝, P Ͻ 0.05 for no neostigmine compared with 5 ϫ 10Ϫ5 neostigmine. P Ͻ 0.05 compared with no vesamicol (Ves)or ,ء hemicholinium-3 (Hem-3) but with neostigmine.

immunohistochemistry in both cell lines and tumors, and, finally, the of the expression of ACh, acetylcholinesterase, and cholinergic re- activity of ChAT has been proven by measuring ACh secretion in the ceptors in the placenta (reviewed in Ref. 38 by Sastry). The role of medium. To further confirm this novel finding, ACh was measured placental ACh remains unknown, however. ChAT has also been independently in both the Mark laboratory (Oregon Health and Sci- reported in glia (39) and in WBCs (40). Reinheimer et al. (22) and ence University, Portland, OR) and the Blusztajn laboratory (Boston Klapproth et al. (41) have described ACh synthesis by bronchial University School of Medicine, Boston, MA). Thus, multiple assays epithelial cells, describing the presence of ChAT and measuring ACh confirm the synthesis of ACh by SCLC. in bronchial epithelial cells. The exact function of nonneuronal en- The ChAT gene includes 18 exons (24), and the coding region of dogenous ACh is unclear, although our data suggest a role in regu- ChAT spans exons 5–18. PCR fragments spanning these exons were lating growth. A key question is which elements of ACh synthesis in amplified from H82 RNA sequenced and found to be identical to neurons are also used by cancers. neuronal ChAT; thus, SCLCs express the same ChAT protein as do The initial step for ACh synthesis is transport of choline into the neurons. In neural tissues, multiple isoforms of ChAT are expressed. cell. In neurons, the high-affinity sodium-dependent choline trans- These forms differ primarily in the 5Ј-untranslated region and express porter CHT1 is needed for ACh synthesis (26, 27). However, by similar ChAT proteins. Using exon-specific probes, we found that all RT-PCR, most SCLC cell lines including H82 cells lack CHT1; thus, of the SCLC cell lines tested expressed ChAT isoforms N, R, and S. CHT1 is clearly not required for ACh synthesis by SCLC. Friedrich et ChAT isoform M was not observed. Real-time PCR revealed that al. (42) have described a sodium-independent choline transporter that ChAT mRNA levels differed considerably among the SCLC cell lines is expressed in endothelial cells and that is inhibited by hemicho- ϳ ␮ investigated. ChAT mRNA levels in H82 cells were highest among linium-3 with a Ki of 50 M. The data presented in Fig. 4C suggest the SCLC cell lines examined, nearly 70 times higher than in H592 that a choline transporter with a similar affinity for hemicholinium-3 cells. ChAT mRNA was detected in both variant (H82) and classic might be responsible for providing the choline for ACh synthesis in (H345) SCLC cell lines and, therefore, at first analysis, does not SCLC. Such a sodium-independent choline transporter has also been appear to correlate with neuroendocrine phenotype and expression of described in pulmonary type II cells in which the transport of choline neuropeptides (29, 35). Consistent with this, ACh was measured in the is also key for surfactant synthesis (43). That H345 cells express medium of both classic and variant cell lines, although further analysis CHT1 suggests that there will be diversity as to which choline of additional cell lines is needed to determine whether the expression transporter is expressed by SCLC. of the cholinergic autocrine loop is indeed independent of neuroen- In neurons, VAChT packages ACh into synaptic vesicles, a neces- docrine differentiation. Interestingly, there was also evidence for an sary step for action potential-regulated secretion (44). In this study all alternate ChAT transcript lacking exon 10 that would in all likelihood of the SCLC cell lines examined expressed VAChT mRNA, although produce an inactive protein. This alternate form has also been previ- mRNA levels varied greatly between lines. The VAChT gene is ously observed in neural tissue (36). located within the ChAT gene, and these two genes comprise the The concept of nonneuronal ACh synthesis is longstanding. One of cholinergic gene locus (25). This raises the possibility that cells that the earliest descriptions of nonneuronal ACh synthesis was by Morris express ChAT may coexpress VAChT even if VAChT is not needed who in 1966 reported that ACh was synthesized in the placenta (37). for ACh secretion. However, given that vesamicol, a specific inhibitor Subsequent to the report by Morris, there have been multiple reports of VAChT (45) decreased H82 ACh release in a concentration- 218

Fig. 5. Effect of modifying autocrine cholinergic signaling on H82 cell growth. H82 cells were plated in 96-well culture plates and cell proliferation measured after specified drug treatments. Cell numbers were measured at 0, 6, 9, and 12 days with using the MTS assay as described in “Materials and Methods.” A, mecamylamine, a nicotinic antagonist; B, atropine, a muscarinic antagonist; C, vesamicol, a VAChT inhibitor; D, hemicholinium-3, a choline-uptake inhibitor; E, nicotine, a nicotinic agonist; F, carbachol, a muscarinic agonist. Ⅺ, Ϫ8 Ϫ7 Ϫ6 control; o,10 M; ,10 M; ,10 M; f, Ϫ5 P Ͻ 0.05 by Neuman-Keuls test after ,ء .M 10 ANOVA. All of the data are expressed as the mean Ϯ SE of 12 replicates.

dependent manner, VAChT appears to be active in SCLC cells. This Our identification of nAChR in SCLC cell lines and tumors is is further supported by the immunohistochemical detection of VAChT consistent with other reports of nAChR in SCLCs. The initial obser- Ϫ5 in SCLCs (not shown). On the other hand, in the H82 cells, 10 M vation of the effects of nicotine on SCLCs dates to Lauweryns et al. vesamicol decreased ACh release by only 40% as opposed to the (48), who noted that nicotine caused degranulation of rabbit pulmo- larger suppressions seen in the neuronal cells (44). This suggests that nary neuroendocrine cells. Subsequently Cunningham et al. (16) SCLC cell lines may also use a second, vesamicol-insensitive, perhaps showed the presence of both nicotinic and muscarinic receptors on nonvesicular, mechanism for ACh secretion as has been reported for SCLCs. Chini et al. (30) and Tarroni et al. (15) did the first detailed placenta (46). description of subtypes of nAChRs in SCLC and reported the pres- For a cholinergic autocrine loop to exist, SCLC must also express ence of ␣3, ␣5, ␤4, and ␣7 nicotinic subunits; and Quik et al. (18), receptors for ACh. There are two classes of ACh receptors, nAChR similarly, reported the presence of ␣7 nAChR in SCLC. Thus the and mAChR. mAChR are G protein-coupled receptors. nAChR be- nAChR subtypes that we have detected are consistent with previous long to the family of ligand-gated ion channels. nAChR consists of reports. Similarly, our observation of M3 receptors in SCLC cell lines five homologous or heterologous subunits. To date, cDNA sequences is also consistent with other reports (14, 20, 49). for 17 kinds of nAChR subunits have been determined (47). These Although the components for a cholinergic autocrine loop are include ␣4–10, ␤1–4, ␥, ␦, and ⑀ (47). In the central nervous system, clearly present in SCLC, to prove an autocrine loop exists, modulation ␣ ␤ the most abundant heteromeric form is ( 4)2( 2)3, the most abundant of growth must be demonstrated. To prove this, we have demonstrated ␣ homomeric form is ( 7)5. Autonomic ganglia express a complex autocrine stimulation at multiple levels in the cholinergic loop. The mixture of ␣3, ␣5, ␣6, ␤2, and ␤4 receptors. As shown in Fig. 3, inhibition of choline transport with hemicholinium-3 decreased ACh SCLCs express ␣3, ␣5, ␣7, ␤2, and ␤4 nAChR. This pattern is synthesis and significantly decreased cell proliferation (Fig. 5). The ␣ consistent with the expression of ( 7)5 homomers and ganglionic inhibition of VAChT with vesamicol decreased ACh synthesis and heteromeric forms of nAChR in SCLCs. Also, as shown in Fig. 3, all significantly decreased cell proliferation. Blockade of the receptors of the SCLC cell lines tested expressed the M3 and M5 mAChR. for endogenously synthesized ACh with muscarinic and nicotinic 219

The extent to which these variations in ChAT and cholinergic receptor expression is related to tumor growth, progression, or prognosis also requires additional studies. It is also clear that only a subset of SCLC will express this cholinergic autocrine loop and that some SCLC will not synthesize or respond to ACh and that some SCLC may only degrade ACh. Additional studies are required to determine how widespread the expression of the cholinergic autocrine loop by SCLC will be. The expression of a cholinergic phenotype by SCLC may help explain the SCLC paraneoplastic syndrome, LEMS, which is charac- terized by antineuronal antibodies leading to myasthenia (56). It is possible that the presentation of cholinergic antigens such as ChAT, VAChT, and CHT1 results in production of antibodies targeted against tumor cells that react with central and peripheral cholinergic nerves and lead to LEMS. The expression of splice variants of ChAT, as shown in Fig. 2, may also promote autoantibody production. Given that antibodies against neuronal nAChR have been reported in SCLC patients with paraneoplastic neurological syndromes (57), it is likely that antibodies to these other cholinergic factors can be produced. Fig. 6. Immunohistochemistry of ChAT and ␣7 nAChR expression in SCLC tumor and cell lines. A, ChAT immunostaining in bronchoscopic biopsy of SCLC (ϫ100; chromo- Interestingly, in a recent report, SCLC patients with LEMS had a gen ϭ VIP). B, higher power view of A (ϫ400). C, ␣7 nAChR immunostaining in significantly longer survival compared with SCLC patients without bronchoscopic biopsy of SCLC (ϫ400; chromogen ϭ AEC). D, immunofluorescent staining of ChAT in H82 cells. ChAT immunostaining is red (Texas-red-labeled second LEMS (58), which suggests that antibodies to cholinergic proteins antibody). Nuclei are counterstained with Blue 4Ј,6-diamidine-2-phenylindole (DAPI). may impair tumor growth. ϫ630. In summary, our data demonstrate the expression of a cholinergic autocrine loop in SCLC and provide a number of new targets for modifying tumor growth. These findings also provide a theoretical antagonists, slowed cell growth and, conversely, muscarinic and nic- basis for understanding the basis by which nicotine and related com- otinic agonists stimulated cell growth. Given that muscarinic antago- pounds modulate lung cancer growth. nists, such as atropine, show some antagonism of nicotinic receptors (50), the development and use of broad spectrum anticholinergic agents could be particularly effective in slowing tumor growth. ACKNOWLEDGMENTS Our identification of a cholinergic autocrine loop by SCLC now We thank Jon Lindstrom for supplying antibodies to the ␣7 nAChR, Kalama provides a framework and rationale for the many reports in the Taylor for technical assistance with cell culture, Greg Disson (Oregon National literature that nicotine and related compounds stimulate SCLC Primate Research Center, Beaverton, OR) for providing the control human growth. Nylen et al. (51) and Schuller et al. (19) have demonstrated brain RNA, and Christopher Corless (Oregon Health and Science University, that nicotine stimulates growth of lung cancer cell lines as well as of Portland, OR) for providing the SCLC archival specimens. cultured pulmonary neuroendocrine cells and that similar effects are obtained with ACh and muscarine. Maneckjee and Minna (17) REFERENCES showed that nicotine blocked opiate-induced inhibition of lung cancer cell line growth. Cattaneo et al. (52) suggested that the effect of 1. Proctor, R. N. Tobacco and the global lung cancer epidemic. Nat. Rev. Cancer, 1: 82–86, 2002. nicotine was mediated by the release of serotonin. Alternately, Nylen 2. Cotran, R. S., Kumar, V., and Collins, T. Robins Patholologic Basis of Disease, Ed. et al. (53) and Novak et al. (54) have suggested that the effects of 6, pp. 697–755. Philadelphia, PA: W. B. Saunders Company, 1999. nicotine are mediated by the release of gastrin-releasing peptide from 3. Simon, G., Ginsberg, R. J., and Ruckdeschel, J. C. Small-cell lung cancer. Chest Surg. Clin. North Am., 11: 165–188, 2001. SCLC. Our results suggest that nicotine modulates the endogenous 4. Cuttitta, F., Carney, D. N., Mulshine, J., Moody, T. W., Fedorko, J., Fischler, A., and cholinergic autocrine loop, which may, in turn, be linked to other Minna, J. D. Bombesin-like peptides can function as autocrine growth factors in human small cell lung cancer. Nature, 316: 823–826, 1985. growth factor pathways. 5. Cuttitta, F., Carney, D. N., Mulshine, J., Moody, T. W., Fedorko, J., Fischler, A., and The exact nAChR subtype mediating the effects of nicotine on Minna, J. D. Autocrine growth factors in human small cell lung cancer. Cancer Surv., SCLC growth remains unclear because the effects of nicotine on 4: 707–727, 1985. ␣ 6. Staley, J., Fiskum, G., Davis, T. P., and Moody, T. W. Neurotensin elevates cytosolic growth have been reported to be blocked both by -bungarotoxin calcium in small cell lung cancer cells. Peptides, 10: 1217–1221, 1989. (␣BGT; Refs. 18, 55), a specific blocker of ␣7 nAChR, and by 7. Giaccone, G., Battey, J., Gazdar, A. F., Oie, H., Draoui, M., and Moody, T. W. mecamylamine (52), a relatively nonspecific nAChR antagonist that Neuromedin B is present in lung cancer cell lines. Cancer Res., 52 (Suppl.): 2732s– 2736s, 1992. would block the effects of ␣7 homomeric nAChR as well as more 8. Berger, W., Setinek, U., Mohr, T., Kindas-Mugge, I., Vetterlein, M., Dekan, G., complex heteromeric nAChRs containing ␣3, ␣5, ␤2, and ␤4 sub- Eckersberger, F., Caldas, C., and Micksche, M. Evidence for a role of FGF-2 and FGF receptors in the proliferation of non-small cell lung cancer cells. Int. J. Cancer, 83: units. These results are consistent with the receptors that we have 415–423, 1999. identified on SCLC and with our observation that mecamylamine 9. Masui, H., Kawamoto, T., Sato, J. D., Wolf, B., Sato, G., and Mendelsohn, J. Growth inhibits H82 proliferation. Given the multiple subtypes of nAChR inhibition of human tumor cells in athymic mice by anti-epidermal growth factor receptor monoclonal antibodies. Cancer Res., 44: 1002–1007, 1984. expressed by SCLC, it is likely that multiple nAChR subtypes can 10. Kang, Y., Prentice, M. A., Mariano, J. M., Davarya, S., Linnoila, R. I., Moody, T. W., modulate growth. Characterization of which nAChRs are most im- Wakefield, L. M., and Jakowlew, S. B. Transforming growth factor-␤1 and its portant in the cholinergic autocrine loop will require additional stud- receptors in human lung cancer and mouse lung carcinogenesis. Exp. Lung Res., 26: 685–707, 2000. ies. Indeed, given the variation in levels of nAChR subunit expression 11. Bunn, P. A., Jr., Soriano, A., Johnson, G., and Heasley, L. New therapeutic strategies for among the different SCLC cell lines, it is likely that different SCLC lung cancer: biology and molecular biology come of age. Chest, 117: 163S–168S, 2000. 12. Woll, P. J., and Rozengurt, E. [D-Arg1,D-Phe5,D-Trp7,9,Leu11]substance P, a potent cell lines and tumors will have different pharmacological responses to bombesin antagonist in murine Swiss 3T3 cells, inhibits the growth of human small nicotine because of differences in nAChR type as well as in density. cell lung cancer cells in vitro. Proc. Natl. Acad. Sci. USA, 85: 1859–1863, 1988. 220

13. Rozengurt, E. Autocrine loops, signal transduction, and cell cycle abnormalities in the 37. Morris, D. The choline acetyltransferase of human placenta. Biochem. J., 98: 754– molecular biology of lung cancer. Curr. Opin. Oncol., 11: 116–122, 1999. 762, 1966. 14. Williams, C. L., and Lennon, V. A. Activation of M3 muscarinic acetylcholine 38. Sastry, B. V. Human placental cholinergic system. Biochem. Pharmacol., 53: 1577– receptors inhibits voltage-dependent calcium influx in small cell lung carcinoma. 1586, 1997. J. Biol. Chem., 265: 1443–1447, 1990. 39. Lan, C. T., Shieh, J. Y., Wen, C. Y., Tan, C. K., and Ling, E. A. Ultrastructural 15. Tarroni, P., Rubboli, F., Chini, B., Zwart, R., Oortgiesen, M., Sher, E., and Clementi, localization of acetylcholinesterase and choline acetyltransferase in oligodendrocytes, F. Neuronal-type nicotinic receptors in human neuroblastoma and small-cell lung glioblasts and vascular endothelial cells in the external cuneate nucleus of the gerbil. carcinoma cell lines. FEBS Lett., 312: 66–70, 1992. Anat. Embryol. (Berl.), 194: 177–185, 1996. 16. Cunningham, J. M., Lennon, V. A., Lambert, E. H., and Scheithauer, B. Acetylcholine 40. Fujii, T., Tsuchiya, T., Yamada, S., Fujimoto, K., Suzuki, T., Kasahara, T., and receptors in small cell carcinomas. J. Neurochem., 45: 159–167, 1985. Kawashima, K. Localization and synthesis of acetylcholine in human leukemic T cell 17. Maneckjee, R., and Minna, J. D. Opioid and nicotine receptors affect growth regulation of lines. J. Neurosci. Res., 44: 66–72, 1996. human lung cancer cell lines. Proc. Natl. Acad. Sci. USA, 87: 3294–3298, 1990. 41. Klapproth, H., Reinheimer, T., Metzen, J., Munch, M., Bittinger, F., Kirkpatrick, 18. Quik, M., Chan, J., and Patrick, J. ␣-Bungarotoxin blocks the nicotinic receptor C. J., Hohle, K. D., Schemann, M., Racke, K., and Wessler, I. Non-neuronal mediated increase in cell number in a neuroendocrine cell line. Brain Res., 655: 161–167, 1994. acetylcholine, a signalling molecule synthesized by surface cells of rat and man. 19. Schuller, H. M., Nylen, E. S., Park, P., and Becker, K. L. Nicotine, acetylcholine and Naunyn-Schmiedebergs Arch. Pharmakol., 355: 515–523, 1997. bombesin are trophic growth factors in neuroendocrine cell lines derived from 42. Friedrich, A., George, R. L., Bridges, C. C., Prasad, P. D., and Ganapathy, V. experimental hamster lung tumors. Life Sci., 47: 571–578, 1990. Transport of choline and its relationship to the expression of the organic cation 20. Fucile, S., Napolitano, M., and Mattei, E. Cholinergic stimulation of human micro- transporters in a rat brain microvessel endothelial cell line (RBE4). Biochim. Bio- cytoma cell line H69. Biochem. Biophys. Res. Commun., 230: 501–504, 1997. phys. Acta, 1512: 299–307, 2001. 21. Wessler, I. K., and Kirkpatrick, C. J. The Non-neuronal cholinergic system: an 43. Kleinzeller, A., Dodia, C., Chander, A., and Fisher, A. B. Na(ϩ)-dependent and emerging drug target in the airways. Pulm. Pharmacol. Ther., 14: 423–434, 2001. Na(ϩ)-independent systems of choline transport by plasma membrane vesicles of 22. Reinheimer, T., Bernedo, P., Klapproth, H., Oelert, H., Zeiske, B., Racke, K., and A549 cell line. Am. J. Physiol, 267: C1279–C1287, 1994. Wessler, I. Acetylcholine in isolated airways of rat, guinea pig, and human: species 44. Auld, D. S., Day, J. C., Mennicken, F., and Quirion, R. Pharmacological character- differences in role of airway mucosa. Am. J. Physiol., 270: L722–L728, 1996. ization of endogenous acetylcholine release from primary septal cultures. J. Pharma- 23. Conti-Fine, B. M., Navaneetham, D., Lei, S., and Maus, A. D. Neuronal nicotinic col. Exp. Ther., 292: 692–697, 2000. receptors in non-neuronal cells: new mediators of tobacco toxicity? Eur. J Pharmacol., 45. Varoqui, H., and Erickson, J. D. Active transport of acetylcholine by the human 393: 279–294, 2000. vesicular acetylcholine transporter. J. Biol. Chem., 271: 27229–27232, 1996. 24. Ohno, K., Tsujino, A., Brengman, J. M., Harper, C. M., Bajzer, Z., Udd, B., Beyring, 46. Wessler, I., Roth, E., Schwarze, S., Weikel, W., Bittinger, F., Kirkpatrick, C. J., and R., Robb, S., Kirkham, F. J., and Engel, A. G. Choline acetyltransferase mutations Kilbinger, H. Release of non-neuronal acetylcholine from the human placenta: dif- cause myasthenic syndrome associated with episodic apnea in humans. Proc. Natl. ference to neuronal acetylcholine. Naunyn-Schmiedebergs Arch. Pharmakol., 364: Acad. Sci. USA, 98: 2017–2022, 2001. 205–212, 2001. 25. Eiden, L. E. The cholinergic gene locus. J. Neurochem., 70: 2227–2240, 1998. 47. Lindstrom, J. Neuronal nicotinic acetylcholine receptors. In T. Narahashi (ed.), Ion 26. Okuda, T., and Haga, T. Functional characterization of the human high-affinity Channels, vol. IV, pp. 377–450. New York: Plenum, 1996. choline transporter. FEBS Lett., 484: 92–97, 2000. 48. Lauweryns, J. M., Cokelaere, M., Deleersynder, M., and Liebens, M. Intrapulmonary 27. Apparsundaram, S., Ferguson, S. M., George, A. L., Jr., and Blakely, R. D. Molecular neuroepithelial bodies in newborn rabbits. Influence of hypoxia, hyperoxia, hyper- cloning of a human, hemicholinium-3-sensitive choline transporter. Biochem. Bio- phys. Res. Commun., 276: 862–867, 2000. capnia, nicotine, reserpine, LDOPA and 5HTP. Cell Tissue Res., 182: 425–440, 28. Phelps, R. M., Johnson, B. E., Ihde, D. C., Gazdar, A. F., Carbone, D. P., McClintock, 1977. P. R., Linnoila, R. I., Matthews, M. J., Bunn, P. A., Jr., Carney, D., Minna, J. D., and 49. Quigley, R. L., Shafer, S. H., and Williams, C. L. Regulation of integrin-mediated Mulshine, J. L. NCI-Navy Medical Oncology Branch cell line data base. J. Cell adhesion by muscarinic acetylcholine receptors and protein kinase C in small cell lung Biochem. Suppl., 24: 32–91, 1996. carcinoma. Chest, 114: 839–846, 1998. 29. Carney, D. N., Gazdar, A. F., Bepler, G., Guccion, J. G., Marangos, P. J., Moody, 50. Zwart, R., and Vijverberg, H. P. Potentiation and inhibition of neuronal nicotinic T. W., Zweig, M. H., and Minna, J. D. Establishment and identification of small cell receptors by atropine: competitive and noncompetitive effects. Mol. Pharmacol., 52: lung cancer cell lines having classic and variant features. Cancer Res., 45: 2913– 886–895, 1997. 2923, 1985. 51. Nylen, E. S., Becker, K. L., Snider, R. H., Tabassian, A. R., Cassidy, M. M., and 30. Chini, B., Clementi, F., Hukovic, N., and Sher, E. Neuronal-type ␣-bungarotoxin Linnoila, R. I. Cholinergic-nicotinic control of growth and secretion of cultured receptors and the ␣5-nicotinic receptor subunit gene are expressed in neuronal and pulmonary neuroendocrine cells. Anat. Rec., 236: 129–135, 1993. nonneuronal human cell lines. Proc. Natl. Acad. Sci. USA, 89: 1572–1576, 1992. 52. Cattaneo, M. G., Codignola, A., Vicentini, L. M., Clementi, F., and Sher, E. Nicotine 31. Sekhon, H. S., Jia, Y., Raab, R., Kuryatov, A., Pankow, J. F., Whitsett, J. A., stimulates a serotonergic autocrine loop in human small cell lung carcinoma. Cancer Lindstrom, J., and Spindel, E. R. Prenatal nicotine increases pulmonary ␣7 nicotinic Res., 53: 5566–5568, 1993. receptor expression and alters fetal lung development in monkeys. J. Clin. Investig., 53. Nylen, E. S., Snider, R. H., Moore, C. F., and Becker, K. L. Modulation of pulmonary 103: 637–647, 1999. bombesin by nicotine and vagotomy. Life Sci., 47: 213–218, 1990. 32. Schoepfer, R., Conroy, W. G., Whiting, P., Gore, M., and Lindstrom, J. Brain 54. Novak, J., Escobedo-Morse, A., Kelley, K., Boose, D., Kautzman-Eades, D., Meyer, ␣ -bungarotoxin binding protein cDNAs and MAbs reveal subtypes of this branch of M., and Kane, M. A. Nicotine effects on proliferation and the bombesin-like peptide the ligand-gated ion channel gene superfamily. Neuron, 5: 35–48, 1990. autocrine system in human small cell lung carcinoma SHP77 cells in culture. Lung 33. Sekhon, H. S., Keller, J. A., Proskocil, B. J., Martin, E. L., and Spindel, E. R. Cancer, 29: 1–10, 2000. Maternal nicotine exposure upregulates collagen gene expression in fetal monkey 55. Codignola, A., Tarroni, P., Cattaneo, M. G., Vicentini, L. M., Clementi, F., and Sher, lung. Association with ␣7 nicotinic acetylcholine receptors. Am. J. Respir. Cell. Mol. E. Serotonin release and cell proliferation are under the control of ␣-bungarotoxin- Biol., 26: 31–41, 2002. 34. Pedersen, W. A., Berse, B., Schuler, U., Wainer, B. H., and Blusztajn, J. K. All-trans- sensitive nicotinic receptors in small-cell lung carcinoma cell lines. FEBS Lett., 342: and 9-cis-retinoic acid enhance the cholinergic properties of a murine septal cell line: 286–290, 1994. evidence that the effects are mediated by activation of retinoic acid receptor-␣. 56. Carpentier, A. F., and Delattre, J. Y. The Lambert-Eaton myasthenic syndrome. Clin. J. Neurochem., 65: 50–58, 1995. Rev. Allergy Immunol., 20: 155–158, 2001. 35. Gazdar, A. F., Carney, D. N., Nau, M. N., and Minna, J. D. Characterization of variant 57. Benatar, M., Blaes, F., Johnston, I., Wilson, K., Vincent, A., Beeson, D., and Lang, subclasses of cell lines derived from small cell lung cancer having distinctive biochemical, B. Presynaptic neuronal antigens expressed by a small cell lung carcinoma cell line. morphological, and growth properties. Cancer Res., 45: 2924–2930, 1985. J. Neuroimmunol., 113: 153–162, 2001. 36. Tooyama, I., and Kimura, H. A protein encoded by an alternative splice variant of 58. Maddison, P., Newsom-Davis, J., Mills, K. R., and Souhami, R. L. Favourable choline acetyltransferase mRNA is localized preferentially in peripheral nerve cells prognosis in Lambert-Eaton myasthenic syndrome and small-cell lung carcinoma. and fibers. J. Chem. Neuroanat., 17: 217–226, 2000. Lancet, 353: 117–118, 1999.

221

Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 2003 American Association for Cancer Research. Acetylcholine Is Synthesized by and Acts as an Autocrine Growth Factor for Small Cell Lung Carcinoma

Pingfang Song, Harmanjatinder S. Sekhon, Yibing Jia, et al.

Cancer Res 2003;63:214-221.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/63/1/214

Cited articles This article cites 55 articles, 14 of which you can access for free at: http://cancerres.aacrjournals.org/content/63/1/214.full#ref-list-1

Citing articles This article has been cited by 17 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/63/1/214.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/63/1/214. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.