Expression of Vesicular Monoamine Transporters, Synaptosomal-associated Protein 25 and Syntaxin1: A Signature of Human Small Cell Lung Carcinoma¹ Free

Amines and peptide hormones are stored in endocrine cells within storage organelles into which they are transported in different ways. Peptide hormones are sequestered cotranslationally within the lumen of the endoplasmic reticulum and then are transported via the Golgi apparatus to the storage organelles. On the other hand, neurotransmitters such as amines synthesized in the cytoplasm are taken up into the storage vesicles by proton-driven transporters (1).

Both peptide hormones and amines are released from endocrine cells by regulated exocytosis. The membrane proteins synaptobrevin, SNAP³-25, and syntaxin1 participate in the release of peptide hormones from the pancreatic islet cells (2, 3) and the hypophysis (4, 5) as well as secretion of amines from adrenal chromaffin cells (6, 7, 8, 9) and gastric ECL cells (10, 11). Moreover, synaptobrevin, SNAP-25, and syntaxin1 participate also in transmitter release by neurons and have been identified in neurons and in endocrine cells by their susceptibility to cleavage and inactivation by clostridial neurotoxins (3, 5, 7, 8, 10, 12, 13, 14, 15, 16, 17). They are collectively termed “SNAREs” for their capacity to bind further proteins such as NSF (a soluble ATPase) via SNAPs (18, 19, 20). The wide occurrence of the SNARE proteins in endocrine cells suggests that they may also serve as general diagnostic markers for endocrine tumors similar to synaptophysin, neuron-specific enolase, and other proteins (21).

Therefore, we explored the expression of the SNARE proteins SNAP-25 and syntaxin1 in lung tumors. About 20% of these tumors belong to a small cell type that is high-grade malignant due to its invasive growth and early metastases. Consequently, small cell type cancers are characterized by short survival of the patients (22). SCLC is often associated with the occurrence of the paraneoplastic syndrome, which is the production and release of peptide hormones by the tumor (23). SCLC cells seem to release serotonin in addition. It is discussed whether serotonin is synthesized by the tumor cells and/or it is taken up from the extracellular fluid (24, 25). Because both the release of peptide hormones and serotonin requires SNAREs in endocrine cells, these proteins are likely to be abundant in SCLC. Furthermore, the storage of serotonin by SCLC would imply that these tumors are equipped with VMATs, which are a prerequisite for proton-driven uptake of amines by intracellular storage organelles (1). Accordingly, we investigated in addition the expression of VMATs by lung tumors.

Our results indicate that the expression of the SNAREs syntaxin1 and SNAP-25 and of the transporters VMAT1 and -2 is a regular and specific signature of SCLC. Other lung tumors (LCLCs, adenocarcinomas, and squamous cell carcinomas) did not express VMATs and, with the exception of 3 of 10 adenocarcinomas, SNAREs. These findings indicate that the proteins analyzed could serve as differentiation markers between SCLC and lung carcinomas with lower malignancy. The presence of VMATs in SCLC is an exciting observation because uptake of amines by the secretory organelles could be used to localize the tumors and/or to target therapeutic agents to the tumors.

The human SCLC cell lines H24 and H69 and the human adenocarcinoma cell line A549 (26, 27) were cultivated in RPMI 1640 (Sigma-Aldrich, Deisenhofen, Germany) containing 5% FCS, as described (28). Paraffin-embedded adrenals, intraoperative samples, and bronchoscopic biopsies of lung tumors and snap-frozen tumors were obtained from a tissue archive of the Institute of Pathology of the Technical University of Munich. The lung tumors investigated had been diagnosed according to histological criteria (21). Tissues were fixed with 4% paraformaldehyde buffered in PBS and embedded in paraffin. Serial sections (5 μm) were cut from paraffin blocks, stained with H&E for initial examination, and then processed for immunohistochemistry.

The sections were deparaffinized and hydrated. Then they were placed in 10 mm citrate buffer (pH 6.0) and heated twice in a microwave oven for 15 min. Cell cultures cultivated on l-ornithin-coated 24-well plates (NUNC GmbH, Wiesbaden, Germany) were fixed with 3% paraformaldehyde and 0.1% glutaraldehyde in PBS (30 min). Cells and sections were permeabilized with saponin (0.05% in PBS). SNAP-25 was localized with a monoclonal antibody (1:1000; obtained from Sternberger Monoclonals Inc., Baltimore, MD), syntaxin1 with a monoclonal antibody (1:500, HPC-1 antibody; Sigma-Aldrich, Deisenhofen, Germany), VMAT2 and VMAT1 with polyclonal antisera (1:1000; Phoenix Pharmaceuticals Inc., Mountain View, CA; described in Ref. 29), followed by biotin-labeled secondary antibodies (goat antimouse and goat antirabbit; Camon, Wiesbaden, Germany; 1:500), and the avidin-biotin complex method kit (Vectastain; Vector Laboratories, Inc., Burlingame, CA). The chromogen used was 3,3′-diaminobenzidine. In control samples the first antibody was omitted or replaced by mouse or rabbit normal serum (1:10000). In addition, we preabsorbed the polyclonal VMAT antisera with synthetic peptides corresponding to COOH-terminal sequences of human VMAT1 and VMAT2, which were also used for immunization (29). Sections of human adrenals (n = 2) served as positive controls.

Cells and intraoperative tissue samples were homogenized by sonication in 62.5 mm Tris-HCl buffer (pH 6.8) containing 10% sucrose and 2% SDS. Mercaptoethanol was added (10%), and the samples were heated (95°C for 5 min). Samples (15 μg of protein per lane) were loaded on SDS-polyacrylamide gels (12.5%), electrophoretically separated, and blotted onto nitrocellulose. Proteins were identified with the monoclonal antibodies directed against SNAP-25 (1:2000), syntaxin1 (1:2000) as described in the previous section. A polyclonal antiserum was used to identify VMAT2 (1:800; obtained from Alpha Diagnostic, San Antonio, TX; peptide antibody against 13 AAs of the COOH terminus). Peroxidase-labeled secondary antibodies (Camon; 1:3000) and an enhanced chemiluminescence kit (Amersham Buchler, Braunschweig, Germany) were used to detect antigen-antibody complexes.

RNA from cultured cells and snap-frozen tumor samples (n = 2) was prepared using a RNeasy kit (Qiagen, Inc., Hilden, Germany). RNA of paraffin-embedded tissue (n = 2) was isolated using the Purescript kit (Gentra, Minneapolis, MN). RNA (500 ng) from cultured cells was used for reverse transcription using a 18-mer polydeoxythymidine primer and Moloney’s murine leukemia virus reverse transcriptase (Promega, Mannheim, Germany). RNA (500 ng) from snap-frozen tumor samples or RNA extracted from paraffin sections was reverse transcribed using Superscript-RT II (Life Technologies, Inc., Karlsruhe, Germany) with gene-specific primers (Table 1) for SNAP-25 (SNAP25-3′), syntaxin1 (syn-3′), VMAT1 (vmat1-3′), and VMAT2 (vmat2-3′). The reverse transcription product (0.5 μl) was used in subsequent PCRs. For nested PCR, 0.5 μl of the product of the first PCR (diluted 1:1000) served as templates. The following PCR conditions were used: 2-min 94°C initial denaturation, 30-s annealing (T_An), 1-min 72°C extension, 30-s 94°C denaturation, and 33 additional cycles. PCR reactions were carried out in a PTC-200 Peltier Thermal Cycler (Biozym, Hessisch Oldendorf, Germany) using Taq-Polymerase (Promega). PCR products were subcloned into the pGEMT vector (Promega) and sequenced using a fluorescence-based didesoxy sequencing reaction (ABI model 377 DNA sequencer; Perkin-Elmer Corp., Ueberlingen, Germany).

For Northern blotting, 15 μg of RNA of cultured cells were denatured at 65°C in a formamide/formaldehyde/4-morpholinepropanesulfonic acid buffer, loaded onto a 1.1% agarose gel containing 2.2 m formaldehyde, and blotted with a vacuum blotter (Appligene Oncor, Heidelberg, Germany) onto a nylon membrane (Hybond-N; Amersham) using 20× SSC. RNA probes were made from sequenced human VMAT1 or human VMAT2 clones obtained by PCR (see above), followed by linearization (NotI or NcoI) and in vitro transcription using 32P-UTP and T7- or SP6-RNA-polymerase (Promega). Transcripts were purified with nick-columns (Pharmacia Amersham, Freiburg, Germany) and hybridized at 60°C overnight. Subsequently, blots were washed five times at 65°C in 0.1× SSC and 0.1% SDS and dried. Autoradiograms were developed after 1–5 days. Sense probes served as control.

To examine the occurrence of VMATs and SNAREs we performed immunocytochemistry with two well established SCLC cell lines (H24 and H69). Both exhibited specific immunoreactivity for VMAT1, VMAT2, SNAP-25, and syntaxin1. In Fig. 1, we show the immunostaining of the cell line H69 as an example. To characterize the immunostained proteins in detail, we investigated the occurrence of VMATs and SNAREs in SCLC cell lines by Western blot analysis. As shown in Fig. 2, we detected single bands for SNAP-25 (M_r 25,000) and syntaxin1 (M_r 35,000) in protein extracts of both cell lines investigated. Both well characterized antisera (29) used for immunocytochemistry for the detection of VMAT1 and VMAT2 revealed no specific bands in immunoblots. However, a different available antiserum directed against VMAT2 (see “Materials and Methods”) allowed to show the presence of VMAT2 in the SCLC cell lines (see Fig. 2).

To extend the results obtained at the protein level, we examined the presence of mRNA coding for VMATs and SNAREs by RT-PCR in the SCLC cell lines (H24 and H69) and in the adenocarcinoma cell line A549. Using the primers shown in Table 1, we obtained bands indicating the presence of SNAP-25, syntaxin1, VMAT1, and VMAT2 in the SCLC cell lines but not in the adenocarcinoma cell line (Fig. 3). The PCR products were characterized by sequencing. Comparison with the published data (see accession numbers in Table 1) revealed the expected sequences.

Finally, we performed Northern blotting to substantiate the occurrence of VMAT1 and VMAT2 in the SCLC cell lines. We used radioactive probes transcribed from the fragments obtained by cloned RT-PCR products. With both probes we obtained a single band for each transcript. VMAT1 mRNA was detected at 3 kb and VMAT2 mRNA at 4 kb (Fig. 4), which is consistent with the published values (29). We conclude from our investigations of the established SCLC cell lines H24 and H69 that they express the SNAREs SNAP-25 and syntaxin1 as well as the vesicular monoamine transporters VMAT1 and VMAT2.

Our results obtained during the analysis of SCLC cell lines lead to the question whether the expression of VMATs and SNAREs is also a feature of SCLC tumors. For this purpose, we used paraffin-embedded tissue of SCLC tumors and snap-frozen samples.

We performed immunohistochemistry on serial sections. Fig. 5 shows an immunostained paraffin section of a SCLC biopsy (case 24 in Table 2) using monoclonal antibodies directed against the SNAREs SNAP-25 and syntaxin1. The image at lower magnification indicates that virtually all cells exhibit the same staining intensity, suggesting a homogeneous expression of the SNAREs by the tumor cells. At higher magnifications (Fig. 5) the presence of the peroxidase reaction product is evident mostly at the surface of the tumor cells, similar to observations in adrenal medullary cells (data not shown). This staining pattern is consistent with the reported presence of SNAREs at the plasmalemma (30, 31). SNAREs have also been found in storage organelles of adrenal chromaffin cells (9) and in neurons (32). These observations suggest that part of the immunostaining observed in the tumor cells may also be due to secretory organelles. All intraoperative samples of SCLC tumors (n = 13) and biopsies obtained by bronchoscopy (n = 12) exhibited specific immunoreactivity for SNAP-25 and syntaxin1 (Table 2). LCLCs (n = 10) and squamous cell carcinomas (n = 10) showed no immunostaining. Three of 10 adenocarcinomas exhibited intracellular granular staining (data not shown), which contrasts to the predominant staining of the cell surface observed in SCLC (see above).

All SCLC tumors examined (n = 25) were specifically immunostained for both vesicular monoamine transporters (see Fig. 5 and Table 2). Strong immunoreactivity of the small cytoplasmic space of the tumor cells was apparent (see Fig. 5), which appeared to be moderately granular as one would expect for an intracellular vesicular antigen. None of the other types of lung carcinomas examined (LCLCs, adenocarcinomas, and squamous carcinomas) exhibited specific immunostaining for VMAT1 and VMAT2 (see Fig. 6 and Table 2). In accordance with published observations, (29) we found strong specific immunoreactivity in human adrenal medullary chromaffin cells with the polyclonal antisera directed against human VMAT1 and VMAT2 (Fig. 7).

We also attempted to analyze SNAREs and VMATs in human tumors by RT-PCR experiments using unstained paraffin sections (cases 6 and 10 in Table 2). RNA was reverse transcribed with gene-specific primers and analyzed by PCR with the primers shown in Table 1. Rarely, we obtained signals with the first PCR. However, if nested PCR (primers in Table 1) was performed, clear signals indicating the expression of SNAP-25 and syntaxin1 in the SCLC tumors were obtained (Fig. 8, top). As positive controls (cases 56 and 57 in Table 2), we performed PCR with paraffin-embedded human adrenal sections (Fig. 8, bottom).

We were unable to detect mRNA specific for VMAT1 and VMAT2 with the same strategy (in fixed and unstained paraffin sections). However, in another approach, we isolated mRNA from snap-frozen samples of SCLC tumors obtained during tumor resection. Conventional histology of adjacent tissue provided no evidence for the presence of nerve or endocrine tissues in these samples. In these extracts bands indicating VMAT1 and VMAT2 were obtained after reverse transcription with gene-specific primers, followed by nested PCR (Fig. 9, top). The snap-frozen samples were also used for Western blot analysis, followed by immunodetection (Fig. 9, bottom). Besides SNAP-25 and syntaxin1 (data not shown), we detected VMAT2 with the same antiserum used for the analysis of SCLC cell lines (see Fig. 2).

Lung tumors are the most frequent tumors observed in both males and females (22). SCLC is one of four major types of lung carcinomas (about 20% of the lung tumors) and belongs to the most malignant tumors known. However, information on the biology of SCLC is still fragmentary, and specific tracers based on the biochemical properties of SCLC suitable to detect and characterize these tumors by PET are not available.

SCLC accounts for 60% of all cases of ectopic Cushing’s syndrome, and the syndrome of inappropriate antidiuretic hormone secretion is due to SCLC in 60% of all cases (23). In addition to the production and secretion of peptide hormones (paraneoplastic syndrome) there is evidence that SCLC may release serotonin: serotonin and the enzyme tryptophan hydroxylase have been identified in SCLC cell lines (24, 33). Both secretion of amines and peptide hormones requires the expression of SNAREs, the core proteins of the exocytotic apparatus shared by endocrine cells and neurons. The storage of serotonin demands VMATs within SCLC cells.

Syntaxin1 has been reported to exist in SCLC cell lines (34), but it was unknown whether this protein occurs also in SCLC tumor tissue. SNAP-25 immunoreactivity has been observed in paraffin sections of SCLCs (35), but the nature of the antigen has not been further analyzed. We revealed in this study the expression of mRNA and protein of both SNAREs, SNAP-25 and syntaxin1, in established SCLC cell lines and in resected SCLCs and in biopsies. The mRNA of the SNAREs in paraffin sections could be isolated and characterized by RT-PCR analysis, followed by sequencing even after years. Immunohistological examination of the major types of lung carcinomas revealed that SNAREs can serve to distinguish SCLC from LCLC and squamous cell carcinoma. Interestingly, we observed SNAP-25 and syntaxin1 immunoreactivity in 30% of the adenocarcinomas analyzed. However the staining could be clearly distinguished from the one in SCLC because it shows a granular pattern. The immunoreactivity in lung adenocarcinomas may be due to the known heterogeneity in this type of lung cancer (36, 37).

Together, our data show that SCLC consistently express two of the major SNAREs necessary to secrete peptide hormones and amines by endocrine cells. The SNAREs, thus, add to other proteins such as synaptophysin, NCAM, chromogranins, and neuron-specific enolase used as diagnostic markers for SCLC to distinguish them from other types (21, 28, 38). The expression of some of these proteins is often variable or inconsistently observed in paraffin sections of SCLC tumors. The strong and specific immunostaining for the SNAREs SNAP-25 and syntaxin1 for SCLC may be of advantage when compared with other markers. SNAREs have been reported to occur also in peptide hormone-secreting cells of the adenohypophysis and have been also detected in derived tumors such as in growth hormone-adenomas and prolactinomas (39). Thus, it seems that SNAREs may serve as differentiation markers for the highly malignant SCLC and for endocrine tumors of low malignancy.

Endocrine cells secreting biogenic amines synthesize their secretory products by cytoplasmic enzymes and transport the amines subsequently into the intracellular storage organelles using a transmembrane proton gradient provided by a proton pumping ATPase (1). This process requires VMATs, which have been characterized on the molecular level within the last few years (40, 41, 42). Serotonin, synthesized and present in SCLC cells (24), is transported by both of the known transporters VMAT1 and VMAT2, although with different affinities (29). There are few cell lines previously shown to express enogeneous VMATs, and SCLC lines may, therefore, be a valuable research tool for studies of these transporters.

Our findings that VMATs are expressed by established SCLC cell lines and SCLC tumors indicate that serotonin-derived tracers may be suitable to detect and analyze SCLC tumors noninvasively by PET imaging. PET imaging has already been successfully applied to less malignant but biochemically related tumors. Carcinoids store and secrete serotonin and express VMATs (43). [¹¹C]-labeled 5-hydroxytryptophane taken up by carcinoids can be used to localize carcinoids in patients by PET (44). VMAT2 found in histamine-secreting ECL cells and in derived tumors allows to identify and characterize ECL hyperplasia and ECL tumors (43). VMATs are present in adrenal chromaffin cells (40) and in pheochromocytomas (45). Suitable tracers allow to analyze pheochromocytomas and metastases by PET imaging, which seemed to be more sensitive and to have higher resolution than traditional ¹³¹I-meta-iodobenzyl guanidine scintigraphy (46). Besides substrates of amine transporters, inhibitors may serve as a tool to examine tumors endowed with amine transporters including SCLC. Recently, [¹¹C]dihydrotetrabenazine, which has a high affinity for VMAT2, was used to examine the density of striatal presynaptic monoaminergic terminals in patients (47, 48).

In conclusion, our observations of the robust expression of the SNAREs SNAP-25 and syntaxin1 and the transporters VMAT1 and VMAT2 in SCLC might be a further step in the biochemical characterization of this frequent and malignant type of lung tumors. Moreover, our study provides a biochemical basis for the potential application of PET imaging in lung cancer diagnosis.

We thank Johannes Grosse, Andreas Bulling, and Britta Sommersberg for help during this work; Martina Haasemann for critical reading of the manuscript; and Marlies Rauchfuß and Andrea Thalhammer for excellent technical assistance.