INTRODUCTION

The accumulation of amines into storage granules in neurons, into chromaffin granules in the adrenal gland, and into granules of peripheral cells such as mast cells, occurs via a monoamine transporter found in the vesicle membrane which is inhibitable by reserpine, tetrabenazine (TBZ),¹ and ketanserin (for review, see Refs. 1, 2-8). The energy for amine transport is derived from a proton gradient, generated by ATP hydrolysis, and a membrane potential (8-10). Two protons are released from the storage vesicle in exchange for one substrate molecule transported to the inside (11, 12). The proton gradient is coupled, by an unknown mechanism, to the transport of biogenic amines into the synaptic vesicle against a steep concentration gradient. The monoamine transporter transports numerous substrates across the vesicle membrane, including dopamine, norepinephrine, epinephrine, serotonin, histamine, tyramine, meta-iodobenzylguanidine, and the neurotoxin N-methyl-4-pyridinium (13-16).

The monoamine transporter was originally cloned from both rat PC12 cells and rat brain (17). The rat chromaffin granule transporter (VMAT1) contains 521 amino acids, and the rat synaptic vesicle transporter (VMAT2) contains 515 amino acids. Analysis by the method of hydrophobic moments implicates 12 transmembrane helices, which is a characteristic of other known transport proteins (18, 19). The primary sequence of both transporters predicts three potential sites of N-linked glycosylation in the large luminal loop between transmembrane helices 1 and 2. The vesicular transporters show some similarity to the bacterial multidrug resistance transporter, which is also inhibitable by reserpine (20).

A transporter with a sequence identical to that of the rat VMAT2 was expression cloned from rat CV-1 cells (21). The transporter activity was assayed by [³H]serotonin uptake into permeabilized cells. In addition, VMAT2 has been cloned from a human brainstem cDNA library and from bovine chromaffin cell cDNA using probes derived from the rat VMAT2 (22-25). The rat VMAT2 is highly similar to these VMATs; it shares 92 and 88% identity with human and bovine VMAT2, respectively. VMAT2 from rat, human, and bovine has similar pharmacological properties (26-29).

A major obstacle in studying the vesicle monoamine transporters at a molecular level has been the inability to obtain large amounts of purified transporter. In an earlier study, we reported the expression and purification of a recombinant synaptic vesicle monoamine transporter (rVMAT2) using the baculovirus expression system.² In this paper we report the identification of the precise amino acid derivatization site of [¹²⁵I]AZIK and of a small peptide fragment from VMAT2 which is specifically labeled by [¹²⁵I]TBZ-AIPP.

EXPERIMENTAL PROCEDURES

Insect cell culture medium was from JRH Biosciences, and antibiotics were from Life Technologies, Inc. Na¹²⁵I was from NEN Life Science Products. 7-Aminoketanserin was from Research Diagnostics Inc. Trypsin and iminodiacetic acid resin were from Sigma, and Ni²⁺-NTA resin was from Qiagen. TBZ was from Fluka. Digitonin was from Gallard Schlesinger. Concanavalin A was from Vector Laboratories, Inc. Molecular weight standards were purchased from Sigma, and prestained molecular weight standards were from Bio-Rad.

rVMAT2 was expressed and purified as described earlier.² For large scale purification (cell culture volumes greater than 1 liter), the previous purification was modified to a batch procedure. Briefly, the Spodoptera frugiperda (Sf9) cell pellet was washed in 0.3 M sucrose, HEPES pH 7.4 buffer (S/H buffer) and solubilized in 20 mM HEPES, pH 7.4, containing 1% digitonin (40 ml of solubilization buffer/liter of original culture harvested). After centrifugation at 200,000 × g for 40 min, Tris-HCl, pH 7.9, NaCl, and imidazole were added to a final concentration of 20 mM, 500 mM, and 5 mM, respectively (added as an 8 × stock solution). Ni²⁺-NTA resin was added to the solubilized sample (500 µl of resin/50 ml of extract). The resin was collected by centrifugation at 1,600 × g for 5 min, resuspended in 5 mM imidazole, 500 mM NaCl, 20 mM Tris-HCl, pH 7.9, containing 0.08% digitonin, and loaded into a syringe column. The resin was washed with 20 ml of the above ice-cold buffer and eluted with 5 ml of 200 mM imidazole, 500 mM NaCl, 20 mM Tris-HCl, pH 8.1, containing 0.08% digitonin. Concanavalin A-agarose (approximately 300 µl of resuspended resin) was washed with the protease mixture, I₃ (50 mM Tris-HCl, pH 7.4, 5 mM EGTA, 100 µM phenylmethanesulfonyl fluoride, 100 µM benzamidine, 5 µg/ml soybean trypsin inhibitor, 20 µg/ml leupeptin, and 5 mM MgCl₂) containing 0.08% digitonin (I₃-0.08% digitonin) and added to the Ni²⁺ column eluate. This was mixed at room temperature for 30 min. The resin was collected by centrifugation at 1,600 × g for 5 min, resuspended in a small amount of I₃-0.08% digitonin, and loaded into a small column. The resin was washed with 3 ml of ice-cold I₃-0.08% digitonin. Elution was performed at room temperature by stopping the column flow and adding 500 µl of I₃-0.08% digitonin containing 250 mM -methyl mannopyranoside (-mm). The resin was resuspended in the elution buffer several times over the period of 30 min, following which the eluate was collected and stored on ice. A second 500-µl elution was performed in the same manner. Purified rVMAT2 was stable on ice in the cold room for more than 1 month.

The preparation of carrier-free [¹²⁵I]AZIK was performed in the following manner (Fig. 1). To 2 mCi of Na¹²⁵I in 10 µl of 0.1 N NaOH was added 10 µl of 0.1 N HCl, 50 µl of NaOAc, pH 5.6, 10 µl of 7-aminoketanserin I (1 mg/ml in dimethyl sulfoxide), and 10 µl of chloramine T (1 mg/ml in H₂O). The reaction was allowed to proceed for 15 min and extracted three times with 300-µl aliquots of ethyl acetate. The combined extracts were concentrated to approximately 100 µl, streaked onto a 10 × 20-cm silica gel thin layer plate, and developed with chloroform:methanol (95:5). The product was identified by autoradiography, scraped from the plate, and the silica was extracted twice with 500-µl aliquots of methanol. The yield of 7-NH₂-8-[¹²⁵I]ketanserin II was 1.1 mCi (55%). All subsequent steps were performed in the dark. The methanol was removed under a stream of nitrogen gas, and immediately 100 µl of ice-cold 3% H₂SO₄ was added. After 10 min on ice, 10 µl of ice-cold 1 M NaNO₂ was added. After 20 min on ice, 50 µl of ice-cold 1 M NaN₃ was added. The reaction was allowed to proceed on ice for 30 min, at which time the reaction was extracted three times with 300-µl aliquots of ethyl acetate. The extract was checked by thin layer chromatography developed with chloroform:methanol (95:5). The product, [¹²⁵I]AZIK III, migrated with an R_f of 0.5, while II migrated with an R_f of 0.2. The conversion of II to III was quantitative, and no further purification was performed. The yield of III was 888 µCi (an overall yield of 44%).³

Using thick walled Pyrex tubes, purified rVMAT2 (1-5 µg) in I₃-0.08% digitonin and -mm was incubated in 0.3 M sucrose, 10 mM K-HEPES, pH 8.0, containing 0.08% digitonin (S/H/digitonin) in the presence or absence of inhibitors for 30 min at 30 °C. [¹²⁵I]AZIK (1 nM) was added, and the incubation was continued in the dark for an additional 60 min on ice. The samples were photolyzed for 5 s in ice water at a distance of 10 cm from a water-jacketed 1-kilowatt high pressure mercury vapor lamp (AH-6 bulb purchased from Advanced Radiation Corporation). After photolysis, -mercaptoethanol was added to a final concentration of 1%. For analysis by SDS-PAGE, loading buffer was added to a 1 × concentration, and the samples were electrophoresed. For trypsin digestion, samples were either digested immediately or frozen at 20 °C for storage.

For the photoaffinity labeling of large quantities of purified rVMAT2 (100-500 µg), the incubation was performed in I₃-0.08% digitonin containing 250 mM -mm as the incubation buffer.

Purified rVMAT2 in I₃-0.08% digitonin and -mm was incubated in 0.3 M sucrose, 10 mM HEPES, pH 7.4, containing 0.08% digitonin and 2 nM [¹²⁵I]TBZ-AIPP in the presence or absence of inhibitors at 30 °C for 1 h. Samples were photolyzed for 5 s after which -mercaptoethanol was added to a final concentration of 1%. Samples were either prepared for immediate electrophoresis by the addition of SDS-PAGE loading buffer or were frozen at 20 °C until later use. For the photolabeling of large amounts of rVMAT2 (>50 µg), incubation was performed in I₃-0.08% digitonin containing 250 mM -mm.

For the generation of tryptic peptides for NH₂-terminal microsequencing, 28 µl of 100 mM NaHCO₃, pH 7.8, was added to 50 µg of [¹²⁵I]AZIK- or [¹²⁵I]TBZ-AIPP-labeled rVMAT2 in 250 µl of I₃-0.08% digitonin and 250 mM -mm. Trypsin (10 µg) was added to the sample and incubated at 37 °C for 3 h. SDS-PAGE loading buffer was added to the sample, and the peptides were resolved on a 10-18% SDS-polyacrylamide gradient gel. The following precautions were taken to prevent NH₂-terminal blocking of the peptides by oxidation. (i) The gels were cast using the purest reagents in the preparation of the gel and buffers, including recrystallized SDS (31). (ii) The gels were allowed to polymerize overnight at ambient temperature. (iii) During electrophoresis, 1 mM 3-mercaptopropionic acid was added to the electrophoresis buffer (32). (iv) During staining and destaining of the polyvinylidene difluoride (PVDF) membrane, only aldehyde free acetic acid was used.

After electrophoresis of the peptides, the SDS-polyacrylamide gel was electrotransferred to a PVDF membrane according to the procedure described by Matsudaira (33). Briefly, the gel was equilibrated in transfer buffer (10 mM CAPS, pH 11, 20% methanol) for 5 min. The PVDF membrane (Millipore, 0.45 µm pore size), wetted in methanol and soaked for 15 min in transfer buffer, was placed on the gel. The gel and PVDF membrane were sandwiched between four pieces of Whatman 3MM paper and placed into a Transphor electrophoresis unit (Hoefer Scientific Instruments). Electrotransfer was performed at 70 volts for 45-60 min at ambient temperature. The PVDF membrane was briefly rinsed in water and then stained in 0.025% (w/v) Coomassie Blue R-250 in 40% methanol for 5 min. Excess stain was removed with 45% methanol, 10% acetic acid (aldehyde-free) for 10 min. The blot was allowed to dry for 30 min before autoradiography. Radiolabeled peptides identified by autoradiography were excised from the PVDF membrane using a clean razor blade and sent to the Michigan State University Macromolecular Structure Facility for NH₂-terminal microsequencing. Automated sequence determinations were confirmed by manually reading the HPLC chromatograms.

Purified rVMAT2 (250 µg in 1 ml of I₃-0.08% digitonin and 250 mM -mm) was photoaffinity labeled with 2 nM [¹²⁵I]TBZ-AIPP. To the photoaffinity-labeled rVMAT2 was added 125 µl of 100 mM NaHCO₃, pH 7.8, and 50 µg of trypsin. The sample was incubated at 37 °C for 3 h after which time SDS-PAGE loading buffer was added. Peptides were resolved on a 12% SDS-polyacrylamide gel. An overnight autoradiogram of the "wet" gel at 80 °C was obtained (the gel was not chemically fixed, stained, or destained). The autoradiogram was used as a template to excise the portion of the gel containing the radiolabeled 38-kDa peptide with a razor blade. The gel slices were placed into a 15-ml conical centrifuge tube and minced with a metal spatula. Water was added to form a slurry. The tube was mixed by tumbling overnight at 4 °C. The gel slurry was centrifuged at 1,600 × g for 5 min, and the supernatant was saved. Water was again added to the gel pieces, and this was tumbled for an additional 2 h at 4 °C. The gel slurry was transferred to a Bio-Spin spin-column (Bio-Rad), and the water was collected by centrifugation at 1,600 × g for 5 min. Radioactivity in the water elution and remaining in the gel was quantitated using a Packard -counter. Typical yield from such a procedure was >75% eluted. The eluted material was concentrated by lyophilization. The sample was resuspended in 100 µl of 0.1 N HCl, and CNBr was added to 0.2 M (from a 5 M stock solution in acetonitrile). The cleavage reaction was incubated for 15 h at ambient temperature (34).

Low molecular weight peptides were resolved with 16.5:6% Tricine SDS-polyacrylamide gels, according to the procedure of Schägger and von Jagow (31). Precautions were taken to prevent NH₂-terminal blocking of the peptides, as described previously. Following electrophoresis of the peptides, the SDS-polyacrylamide gel was electrotransferred to a PVDF membrane as described previously.

Amino-terminal amino acid sequence analysis was performed by Dr. Joe Leykam of the Michigan State University Macromolecular Structure Facility using automated Edman degradation in an Applied Biosystem 477A gas phase amino acid sequencer. Phenylthiohydantoin amino acid derivatives were identified by HPLC.

Proteins and peptides were analyzed according to the method of Laemmli (35) using either 12% or 10-18% gradient polyacrylamide gels using the following proteins as molecular mass markers: myosin (205 kDa), -galactosidase (116 kDa), phosphorylase b (97.4 kDa), bovine serum albumin (66 kDa), ovalbumin (45 kDa), glyceraldehyde-3-phosphate dehydrogenase (36 kDa), carbonic anhydrase (29 kDa), trypsinogen (24 kDa), soybean trypsin inhibitor (20 kDa), -lactalbumin (14.2 kDa), and aprotinin (6.5 kDa). Prestained molecular mass markers were as follows: myosin (213 kDa), -galactosidase (123 kDa), bovine serum albumin (85 kDa), ovalbumin (50.3 kDa), carbonic anhydrase (33.3 kDa), soybean trypsin inhibitor (28.5 kDa), lysozyme (18.9 kDa), and aprotinin (7.8 kDa). Autoradiography was performed with Kodak X-Omat film at 80 °C using a Quanta III (NEN Life Science Products) intensifying screen.

RESULTS

Purified rVMAT2 was photoaffinity labeled with [¹²⁵I]AZIK and [¹²⁵I]TBZ-AIPP as described under "Experimental Procedures." As seen in the autoradiogram in Fig. 2, both [¹²⁵I]AZIK and [¹²⁵I]TBZ-AIPP labeling of purified rVMAT2 is completely protectable by 10 µM 7-aminoketanserin and TBZ, inhibitors of the transporter (panels A and C). In addition, photoaffinity labeling of the purified transporter was inhibited by a 1 mM concentration of the VMAT substrates dopamine, norepinephrine, and serotonin (panels B and D).

Fig. 3A shows an autoradiogram from a 3-h tryptic digestion of [¹²⁵I]AZIK-labeled rVMAT2 transferred to a PVDF membrane. Four radiolabeled polypeptides, 38, 17, 10, and 5 kDa, were identified on the autoradiogram. The 17-kDa peptide contained the majority of the radiolabel. Photolabeling in the presence of 100 µM 7-aminoketanserin showed that all four peptides were specifically labeled (data not shown). The 17-kDa radiolabeled peptide, identified by the arrow, was excised with a razor blade and NH₂-terminal microsequenced. The sequence obtained and its position in rVMAT2 are shown in Fig. 3B. The sequence started at the NH₂ terminus of the rVMAT2 sequence and matched through 14 cycles with the predicted amino acids. Only one sequence was obtained, supporting the fact that only one peptide migrated at this position. One amino acid at position 14 was not identified with certainty because of the low yield at this cycle and was assigned as a glutamine. The amino acid as predicted by the cDNA is a tryptophan and was confirmed by sequencing of the cDNA insert.

To identify the position of [¹²⁵I]AZIK insertion within the 17-kDa peptide, radiosequencing was performed on this peptide. Release of ¹²⁵I was found to occur at the cycle corresponding to lysine 20 of rVMAT2, a nucleophilic amino acid, which would be expected to react with aryl nitrenes. Because the location of the [¹²⁵I]AZIK insertion site was within the first 25 amino acids from the NH₂ terminus, the label release experiment was repeated using the entire undigested [¹²⁵I]AZIK-labeled rVMAT2. Fig. 4A shows the data obtained from the latter experiment. Lysine 20 was again shown to contain radiolabel from [¹²⁵I]AZIK derivatization. Corresponding amino acids are shown on the abscissa. The position of lysine 20 in the rVMAT2 model is shown in Fig. 4B. As shown, Lys-20 is the last amino acid before the first putative transmembrane domain (TMD; derivatization is indicated by the asterisk).

Purified rVMAT2 was photoaffinity labeled with 2 nM [¹²⁵I]TBZ-AIPP in the absence of protector because protection of [¹²⁵I]TBZ-AIPP labeling of the tryptic peptides by 100 µM TBZ was quantitative (Fig. 2). The photoaffinity-labeled rVMAT2 was then incubated with trypsin for 3 h and the peptides separated on a 10-18% SDS-polyacrylamide gradient gel. The gel was electrotransferred to a PVDF membrane as described under "Experimental Procedures." Fig. 5A shows an autoradiogram of the PVDF membrane. The radiolabeled peptides indicated by the arrows were excised and NH₂-terminal microsequenced. Fig. 5B shows the NH₂-terminal sequences obtained from the peptides and their location on the model of VMAT. The NH₂ terminus of the largest peptide, identified by arrow 1 in Fig. 5A, begins between the third and fourth putative glycosylation sites on the large loop between transmembrane domains 1 and 2. On the basis of its molecular weight, this peptide contains putative TMDs 2-11, with its COOH terminus between TMD11 and TMD12, or in TMD12. Peptides indicated as 2, 3, and 4 were all derived from the NH₂ terminus of rVMAT2. The largest of these peptides, number 2, contains a majority of the luminal loop between putative TMD1 and TMD2, possibly as far as the start of peptide 1. The smaller molecular weight peptides, 2 and 3, are most likely derived from this NH₂-terminal peptide.

The majority of the [¹²⁵I]TBZ-AIPP photoinserted into the 38-kDa tryptic fragment (Fig. 5A). This fragment, based on molecular weight, contains TMDs 2-11/12. To identify the [¹²⁵I]TBZ-AIPP insertion site further, a large amount of purified rVMAT2 (150 µg) was photolabeled with [¹²⁵I]TBZ-AIPP and digested with trypsin. The radiolabeled 38-kDa peptide was eluted from the gel slices as described under "Experimental Procedures." This peptide was then digested with CNBr and peptides resolved on a 16.5:6% Tricine gel (31). The gel was transferred to a PVDF membrane and an autoradiogram obtained. A very small radiolabeled peptide of less than 2 kDa was observed (Fig. 6A). This region of the PVDF membrane was excised and NH₂-terminal sequenced. The NH₂-terminal sequence determined from this peptide is shown in Fig. 6B, as well as its location in a model of VMAT.

DISCUSSION

We had reported previously that rVMAT2 expressed in Sf9 cells could be purified in large quantities.² The Sf9-expressed rVMAT2 was functional as determined by [³H]TBZOH binding and [¹⁴C]dicyclohexylcarbodiimide derivatization. Additionally, rVMAT2 was photoaffinity labeled using [¹²⁵I]AZIK and [¹²⁵I]TBZ-AIPP. Specificity of [¹²⁵I]AZIK and [¹²⁵I]TBZ-AIPP photoaffinity labeling was demonstrated using the VMAT inhibitors 7-aminoketanserin and TBZ. It was shown in this report that the photoaffinity labeling of purified rVMAT2 by [¹²⁵I]AZIK was protected completely by the transporter substrates dopamine, norepinephrine, and serotonin (Fig. 2, A and B). The concentration of the substrates used in this experiment was 1 mM. Serotonin, the highest affinity substrate of the three, inhibits [¹²⁵I]AZIK binding to bovine chromaffin vesicle membranes with an IC₅₀ of 500 µM (36) and inhibits [³H]ketanserin binding to chromaffin vesicle membranes with a K_i of 250 µM. Norepinephrine has a K_i of 800 µM for the inhibition of [³H]ketanserin in the same membranes (5). Dopamine has approximately the same affinity for VMAT as that of norepinephrine. Similar results were obtained for inhibition of [¹²⁵I]TBZ-AIPP photoaffinity labeling by VMAT substrates (Fig. 2, C and D). Serotonin has an IC₅₀ of 345 µM for [³H]TBZOH binding in chromaffin vesicle membranes (37) and 2,200 µM for [³H] TBZOH binding in synaptic vesicles (38). The IC₅₀ values of dopamine and norepinephrine for [³H]TBZOH binding in chromaffin vesicle membranes are approximately 1,000-2,000 µM and 4,600-5,500 µM for [³H]TBZOH binding in synaptic vesicles (37-39).

Several proteases were tested for cleavage of the radiolabeled rVMAT2. It was observed that trypsin cleaved both [¹²⁵I]AZIK- and [¹²⁵I]TBZ-AIPP-labeled rVMAT2 rapidly. Because the rVMAT2 was purified in a mixture of protease inhibitors, it was necessary to use large amounts of trypsin to digest the radiolabeled samples; ratios of trypsin to rVMAT2 as high as 1:2 were used, but a ratio of 1:5 was used regularly. The radiolabeled tryptic peptides from both [¹²⁵I]AZIK- and [¹²⁵I]TBZ-AIPP-labeled rVMAT2 were similar in molecular size but differed in their susceptibility to trypsin over time. Trypsin cleavage of [¹²⁵I]AZIK-labeled rVMAT2 resulted in three (four with extended trypsin cleavage) radiolabeled peptides: 38, 17, and 10 kDa (with a fourth peptide around 5 kDa). The 17-kDa peptide contained the majority of the radioactivity (Fig. 3A). Trypsin cleavage of [¹²⁵I]TBZ-AIPP-labeled rVMAT2 resulted in a similar 38-kDa, 17-kDa, and 10-kDa peptide. A fourth peptide of 8 kDa appeared with the concomitant loss of 17-kDa peptide. An interesting observation is that although the 17-kDa peptide from both photoaffinity labels migrated with similar electrophoretic mobility, and both peptides have the same NH₂-terminal sequence, the peptide resulting from the [¹²⁵I]TBZ-AIPP label is cleaved further by trypsin. There are two possible explanations for this: (i) the peptides are different but migrate similarly in this electrophoretic system; or (ii) the peptides are the same, but the presence of [¹²⁵I]AZIK on the transporter protects the labeled transporter from further cleavage by trypsin. The latter is possible because the transporter is cleaved under nondenaturing conditions.

When the similar peptide from [¹²⁵I]TBZ-AIPP-labeled rVMAT2 was sequenced (peptide 2, Fig. 5A), the same sequence as above was obtained. The same NH₂-terminal sequence was also obtained when the smaller peptides from [¹²⁵I]TBZ-AIPP-labeled peptides were sequenced (peptides 3 and 4, Fig. 5A). For each of these peptides, only one peptide was visible on the Coomassie-stained PVDF membrane, and only one sequence was obtained from the sequencer, indicating that these peptides were homogeneous. These data, taken together, indicate that [¹²⁵I]AZIK reacts primarily at the NH₂ terminus of rVMAT2, photoinserting into lysine 20, and [¹²⁵I]TBZ-AIPP inserts partially into this region (see the model presented in Fig. 7).

The other tryptic peptide generated with both labels is the 38-kDa peptide. [¹²⁵I]TBZ-AIPP primarily labels this peptide, whereas [¹²⁵I]AZIK only labels this peptide slightly. The NH₂ terminus of this peptide begins in the large loop between TMD1 and TMD2 (Fig. 5), and the molecular weight is consistent with this fragment extending to TMD12. This tryptic fragment derived from [¹²⁵I]TBZ-AIPP-labeled rVMAT2 was purified and further cleaved by CNBr. This cleavage was performed under denaturing conditions (0.1 N HCl), after elution of the 38-kDa peptide from an SDS-polyacrylamide gel slice. Cleavage by CNBr resulted in a small radiolabeled polypeptide of less than 2 kDa (Fig. 6A). When this peptide was excised from the PVDF membrane and NH₂-terminal microsequenced, a single sequence was identified as shown in Fig. 6B. The amino acid derivatized by [¹²⁵I]TBZ-AIPP in this fragment has not been identified at this point. The size of the peptide corresponds to approximately 20 amino acids; based on the predicted sequence of rVMAT2, CNBr cleavage should result in a 24-amino acid peptide in this region. This sequence contains a histidine, an arginine, and a serine, which are potential sites of reaction with the nitrene (40-43). It is important to note that this peptide and Lys-20, which is derivatized by [¹²⁵I]AZIK, are both on the putative cytoplasmic surface of the transporter. The insertion site for [¹²⁵I]AZIK within this large fragment has not been identified, but it would be interesting if it also labeled the same site as [¹²⁵I]TBZ-AIPP.

It is hypothesized that TMD1 is involved in the recognition and binding of inhibitors and substrates because the azide moiety of AZIK is in a region of the ketanserin backbone which has structural similarity to substrates of VMAT. This hypothesis is supported by data obtained from the VMAT1-VMAT2 chimera work reported by Peter et al. (44). The replacement of TMD1 from VMAT1 with TMD1 of VMAT2 increased substrate and inhibitor affinities to an intermediate level between that of VMAT1 and VMAT2. All of the inhibitors for VMAT contain a portion of the molecule which is structurally similar to substrates: reserpine has an indole portion similar to serotonin, and the dimethoxy phenyl portion of TBZ is similar to dopamine. Using TMD3 as an interacting site for the dopamine analogous region of [¹²⁵I]TBZ-AIPP, a determination of the insertion site for the nitrene of [¹²⁵I]TBZ-AIPP should identify domains near TMD1. (The length of the extended [¹²⁵I]TBZ-AIPP molecule as measured with the SYBYL molecular modeling computer program is approximately 20 Å. An -helical domain was measured to be approximately 10 Å in diameter using the same program.) Because [¹²⁵I]TBZ-AIPP labeled both TMD1 and TMD10/11, these two domains must be close spatially. Therefore, our model, as shown in Fig. 7, is that TMD1 and TMD10/11 are juxtaposed.

Because [¹²⁵I]AZIK derivatizes Lys-20 at the NH₂ terminus of rVMAT2, the azide moiety must be positioned within covalent bond-forming distance of this residue. The putative TMDs may form a channel in the vesicle membrane through which the protons and amines pass. Therefore, the putative TMDs in this model are arranged in a fashion that juxtaposes TMD1 with the TMD10/11 region. A recent report by Merickel et al. (30) is consistent with this model. Using site-directed mutagenesis, the authors determined that there is an ion pair formed between Lys-139 and Asp-427 in TMD2 and TMD11, respectively.

The model is also supported by the work performed using chimeric VMAT1/VMAT2 proteins (44). Replacing TMD1 of VMAT1 with TMD1 of VMAT2 increased the affinity of the transporter for inhibitors and substrates to an intermediate level between VMAT1 and VMAT2. The authors propose the possible interaction of this region with either the ligand or other domains of the transporter. Data from this study suggest that TMD1 interacts with both the ligand (inhibitors and substrates) as well as the TMD10/11 region of the transporter. This hypothesis is supported by the observation that chimeric transporters containing TMD9-TMD12 of VMAT2 had VMAT2-like binding affinities only in the presence of NH₂-terminal VMAT2 sequence and a recent report that residues Lys-139 in TMD2 and Asp-427 in TMD 11 form an ion pair. These observations are suggestive of an interaction between the NH₂ and COOH termini of VMAT2. An interesting set of experiments to address the question of NH₂ and COOH termini interaction would be to synthesize a chimeric transporter that contains TMD1 and TMD10-12 of VMAT2 and the remaining sequence of VMAT1. It would be predicted, based on the model, that this construct would possess properties of VMAT2 (i.e. high affinity for TBZ and substrates).

The multiple insertion sites for [¹²⁵I]TBZ-AIPP can be explained using the model in Fig. 7B. The iodophenylazide moiety is connected to the TBZ backbone by a long, flexible alkyl chain. If TMD1 and TMD10/11 are juxtaposed, as shown in the model, the azide could derivatize either one of these regions. Nitrene insertion is favored by TMD10/11, perhaps due to a more favorable positioning of reactive amino acid side chains. Determination of the insertion site for [¹²⁵I]AZIK in the 38 kDa tryptic fragment should resolve the proximity hypothesis of TMD1 and TMD10/11.

The data presented in this study indicate that TMD1 and TMD10/11 are juxtaposed and perhaps interacting in a functionally significant manner. Additional experiments are needed to confirm this model, such as the use of ketanserin and TBZ molecules with the photoactive moiety in various positions on the molecule. Of particular interest would be a TBZ derivative in which the azide is positioned in the dopamine analogous region. Another photoaffinity label which would provide useful information would be a ketanserin derivative with the photoactive moiety at the opposite end of the ketanserin molecule.