Differential Expression of SNAP-25 Isoforms and SNAP-23 in the Adrenal Gland

Authors


  • Abbreviations used : LDCV, large dense-core vesicle ; NSF, N-ethylmaleimide-sensitive fusion ; PC12, pheochromocytoma cell line ; PNMT, phenylethanolamine-N-methyltransferase ; SNAP, synaptic NSF-associated protein ; SNAP-25 or SNAP-23, synaptosomal-associated protein of 25 or 23 kDa ; SNARE, SNAP receptor ; t-SNARE and v-SNARE, target and vesicle membrane SNARE, respectively ; SSV, small synaptic vesicle.

Address correspondence and reprint requests to Dr. N. J. Grant at INSERM U338, 5 Rue Blaise Pascal, 67084 Strasbourg Cedex, France.

Abstract

Abstract : In the rat adrenal gland, we previously observed that SNAP-25 is not restricted to the plasmalemma in noradrenergic cells as it is in adrenergic cells, and hypothesized that SNAP-25 isoform expression is different in the two phenotypes. Expression of SNAP-25 isoforms and SNAP-23 was examined by immunoblotting, immunofluorescence, and RT-PCR. Amplifications of SNAP-25 mRNAs were combined with Southern hybridization, restriction enzyme analysis, and sequencing of cloned PCR products to compare SNAP-25 isoform expression in rat and bovine adrenal glands. SNAP-25 and SNAP-23 mRNA and protein are expressed in the glands ; SNAP-23 is enriched in the adrenal cortex, whereas SNAP-25 is restricted to the adrenal medulla. Furthermore, high levels of SNAP-25 and low levels of SNAP-23 are observed in the PC12 cells, whereas both SNAP-25 and SNAP-23 are expressed in adrenal medullary cultures. In all extracts, the SNAP-23 mRNA corresponded to SNAP-23a. SNAP-25a is the major form expressed in rat adrenal glands (75%), as it is in PC12 cells (80%), but both SNAP-25a and SNAP-25b (40% vs. 60%) are expressed in bovine adrenal medulla in situ and in culture. In addition, an enriched population of adrenergic cells (93%) expressed a higher level of SNAP-25b (70%), suggesting that this isoform may not be restricted to fast neurotransmission.

SNAP-25 (synaptosomal associated protein of 25 kDa) is implicated in the docking and fusion of vesicles at the plasma membrane during both constitutive and regulated exocytosis in neurons and neuroendocrine cells (Sollner et al., 1993 ; Sudhof et al., 1993 ; Sollner and Rothman, 1994). The functional importance of SNAP-25 in neurotransmitter release from small synaptic vesicles (SSVs) and large dense-core vesicles (LDCVs) is underscored by evidence that it is a specific target of botulinum toxins A and E, which selectively block neurotransmission (Ahnert-Hilger and Weller, 1993 ; Blasi et al., 1993 ; Schiavo et al., 1993 ; Binz et al., 1994 ; Roth and Burgoyne, 1994). In the hypothesis currently proposed for vesicle-docking, soluble N-ethylmaleimide-sensitive fusion (NSF) protein and the NSF-associated proteins (SNAPs) form a fusion complex with SNAP-25 and syntaxin, the target receptors or t-SNAREs on the plasma membrane, and synaptobrevin and synaptotagmin, the vesicle membrane receptors or v-SNAREs (Sollner et al., 1993 ; Sollner and Rothman, 1994 ; Schiavo et al., 1997). SNAP-25 and syntaxin have been recently reported on synaptic vesicles (Walch-Solimena et al., 1995 ; Kretzschmar et al., 1996) and chromaffin granules (Hohne-Zell and Gratzl, 1996 ; Tagaya et al., 1996), suggesting that the targeting role of these proteins may be oversimplified. The existence of multiple receptors and isoforms of many of these fusion complex proteins may contribute to vesicletarget specificity necessary for distinct biological functions requiring vesicle fusion (Bark et al., 1995 ; Linial, 1997).

Recently, the SNAP-25 proteins have been grouped together in a gene family including the yeast protein Sec 9 (Brennwald et al., 1994) and a newly identified, nonneuronal, mammalian homologue called SNAP-23 (Ravichandran et al., 1996 ; Mollinedo and Lazo, 1997 ; Wang et al., 1997 ; Wong et al., 1997). SNAP-23 is 59% identical to SNAP-25 in amino acid composition, and interacts with the SNAP receptor (SNARE) complex proteins (Ravichandran et al., 1996). Neuronal SNAP-25 itself exists as two isoforms, SNAP-25a and SNAP-25b, which are identical in length and differ by the disposition of just nine amino acids over a sequence of 32 amino acids (Bark, 1993 ; Bark and Wilson, 1994a). The isoforms arise from a single, highly conserved gene by alternative splicing of exons 5a and 5b. This region codes for the membrane-associated domain of SNAP-25. A cluster of four cysteine residues provides the sites for palmitoylation, which probably permit SNAP-25 to interact with the plasma membrane (Hess et al., 1992 ; Viet et al., 1996). The disposition of these cysteine residues differs in the two SNAP-25 isoforms and this change in fatty acylation domains has been proposed to affect their ability to associate with the plasma membrane (Bark and Wilson, 1994b ; Bark et al., 1995).

SNAP-25 is expressed in cultured chromaffin cells and seems likely to play a role in regulated exocytosis of catecholamines and neuropeptides from LDCVs in these cells (Ahnert-Hilger and Weller, 1993 ; Bartels et al., 1994 ; Hohne-Zell et al., 1994 ; Roth and Burgoyne, 1994). We recently examined SNAP-25 expression in situ in the rat adrenal gland and found that SNAP-25 is differentially expressed in chromaffin cells (Kannan et al., 1996). Although both adrenergic and noradrenergic cells express SNAP-25 in the subplasmalemmal region, only noradrenergic cells display a more extensive distribution. As the SNAP-25 isoforms differ in their membrane-interacting domains, we hypothesized that this difference might be related to modifications in isoform expression in the different catecholaminergic phenotypes ; we developed a specific isoform RT-PCR technique to characterize the SNAP-25 isoforms. Its homologue, SNAP-23, has a similar membrane interacting domain, but has yet to be described in the adrenal gland. Here, we analyze the expression of SNAP-25a and SNAP-25b isoforms and SNAP-23 in the rat and bovine adrenal gland. We used the bovine chromaffin cell in culture as a model because enriched populations of the two major catecholaminergic phenotypes can be obtained (Krause et al., 1996). The relationship between catecholaminergic phenotype and SNAP-25 isoform was then investigated in these enriched populations of adrenergic and noradrenergic chromaffin cells by RT-PCR.

MATERIALS AND METHODS

Tissue and cell preparation

Rat adrenal glands and cerebella were dissected from 12-week-old male Wistar rats. Whole adrenal glands and quartered cerebella were frozen in liquid nitrogen and stored at -80°C before RNA extraction. Rat pheochromocytoma (PC12) cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated horse serum and 5% heat-inactivated fetal calf serum as previously described (Grant et al., 1996) and RNA was extracted after 1 week of culture.

Bovine adrenal glands and calf brains were collected in sterile physiological solution from the abattoir. In the laboratory, the tissues were immediately dissected under sterile conditions, and 50—100-mg aliquots of adrenal medulla and cerebellum were quick-frozen and stored as described above for rat tissues. Chromaffin cells were also isolated from bovine glands, purified by centrifugation on Percoll gradients, and cultured in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum and mitotic inhibitors, as previously described (Grant et al., 1987). Preparations of enriched populations of adrenergic and noradrenergic chromaffin cells were prepared by collecting cells from different fractions of the Percoll gradient following the modified protocol of Krause et al. (1996). Catecholamine content of the subpopulations of chromaffin cells was evaluated in cell lysates using HPLC and electrochemical detection. RNA was extracted from frozen aliquots of freshly isolated chromaffin cells (107), and from cultured cells after 5 days of culture.

PCR primers

A PCR product (559 bp) representing the amplified product from cDNA coding for both SNAP-25 isoforms, hereafter referred to as “total SNAP-25,” was obtained with a forward primer on exon 2 and a reverse primer on exon 7/8 (Table 1). In the isoform-specific PCRs, PCR products of 202 bp were amplified by replacing the reverse primer with one on either exon 5a or exon 5b (Table 1). These primers were modified from the oligonucleotides used by Bark (1993) according to the sequences for rat SNAP-25 (AB003991 and AB003992). Amplification of SNAP-23, using the primers listed in Table 1, distinguishes between SNAP-23a and SNAP-23b, producing products of 392 and 236 bp, respectively (human sequences : Y09567 and Y09568). Amplification of a PCR product corresponding to β-actin was routinely performed as an internal control of RNA quality and the RT step of the RNA extracts. The actin mRNA levels were constant between the different preparations (data not shown).

Table 1. Sequences of oligonucleotides
Primers Sequence (5′-3′) BasesReference
SNAP-25  AB003991
Exon 2aggacgcagacatgcgtaatgaactggagg171-200 
Exon 7/8gttggagtcagccttctccatgatcctgtc564-535 
Exon 5attggttgatatggttcatgccttcttcgacacga207-174 
Exon 5bcttattgatttggtccatcccttcctcaatgcgt207-174AB003992
Exon 4gtcctgatgccagcatctttactctcttcg137-108 
Exon 6gctggctggccactactccatcctgattat351-322 
SNAP-23  AB000822
5′agccattgagtctcaggatgcagga141-165 
3′ctgcccacttgagtcaggttctctt530-506 

TABLE 1.

RT-PCR

Total RNA was prepared from cultures by directly lysing cells in extraction buffer (guanidinium thiocyanate/phenol/chloroform) and RNA was transcribed into cDNA using oligo(dT) and Superscript RNase H- Moloney murine leukemia virus reverse transcriptase (GibcoBRL, France), as previously described (Grant et al., 1996). The following PCR cycle profile was used : denaturation at 94°C for 45 s, annealing at 60°C for 60 s, and polymerization at 72°C for 60 s for 28 cycles, unless indicated, followed by an additional polymerization at 72°C for 120 s. The quantity of cDNA amplified corresponded to 50 ng of RNA in RT-PCRs for actin, SNAP-23, and total SNAP-25 unless indicated.

As the amount of total SNAP-25 mRNA varied considerably for the different tissues and cells (the adrenal gland had 100-fold less than cerebellum and 20-fold less than PC12 cells) in preliminary RT-PCRs, the quantity of cDNA used for specific amplifications of SNAP-25 and SNAP-25b was adjusted for each sample, so that approximately equivalent amounts of final PCR product would be obtained for total SNAP-25 amplifications run in parallel. The relative levels of SNAP-25a and SNAP-25b could be compared in the different samples. After amplification, the relative expression of each isoform was quantitated by Southern blot, as described below, using an oligonucleotide on exon 4 (Table 1 ; Eurogentech, Belgium), which is common to both the total and isoform-specific amplification products of SNAP-25.

Characterization of SNAP-25 PCR products

The total SNAP-25 PCR products were also characterized by three separate approaches as alternative means of determining which isoform was expressed. The total SNAP-25 PCR products were analyzed by Southern blots, by restriction enzyme digestion, and finally, by cloning and sequencing the total PCR products.

Southern blot analysis.

Equal aliquots (500 ng) of purified total SNAP-25 PCR products for rat cerebellum, PC12 cells, and rat adrenal and cultured bovine chromaffin cells were hybridized with internal oligonucleotides end-labeled with [γ-32P]ATP to a specific activity of 105 cpm/pmol. The oligonucleotide recognizing both isoforms was on exon 6 (Table 1) and the specific oligonucleotides on exon 5a and exon 5b were the same as those used as primers in the isoform-specific PCRs. Hybridization was performed at 55°C for 3 h, and the blots were washed according to standard protocols (Sambrook et al., 1989). Images were analyzed by using a Bio-Imaging Analyzer FUJIX BAS1000 (Fuji, Tokyo, Japan).

Digestion of total SNAP-25 PCR product.

Total SNAP-25 PCR products (n = 3) for rat cerebellum, rat adrenal gland, PC12 cells, and cultured bovine chromaffin cells were purified using a Wizard PCR Purification kit (Promega, France), digested with StyI, PvuI, and NdeI, and then analyzed on ethidium bromide-stained agarose gels.

Sequencing of total SNAP-25 PCR products.

The total SNAP-25 PCR products obtained for extracts of rat cerebellum, rat adrenal, and cultured bovine chromaffin cells were cloned and sequenced. Immediately after amplification, the bands corresponding to the 559-bp PCR product were excised from low-melting agarose gels and purified using the Wizard PCR Purification kit. The purified SNAP-25 PCR products were then ligated into pMOSBlue T vector and cloned into the pMOS-Blue-competent bacteria following the manufacturer's instructions (Amersham, France). Positive clones were sequenced in both directions using the T7 primer, the universal primer, or the SNAP-25 exon 6 primer as the initiation site, with the Sequenase version 2.0 kit (Amersham).

Immunofluorescence

Cryostat sections of paraformaldehyde-fixed rat adrenal, bovine adrenal medulla, and cultured bovine chromaffin cells were prepared and processed for immunofluorescence, as previously described (Kannan et al., 1996), using a mouse monoclonal antibody against SNAP-25 (1:1,000 dilution ; Sternberger Monoclonals, U.S.A.). To determine the catecholaminergic phenotype in enriched noradrenergic and adrenergic populations, cultures were double-stained with a general marker of chromaffin cells, a monocolonal anti-tyrosine hydroxylase (1:250 dilution ; Euromedex, Strasbourg, France) and a specific marker of adrenergic cells, a rabbit anti-bovine phenylethanolamine-N-methyltransferase (PNMT ; 1:500 dilution). All secondary antibodies (Euromedex) were affinity purified and preadsorbed (1:150 dilution).

Confocal images of SNAP-25 staining of cryostat sections were obtained with a Zeiss laser scanning microscope (LSM 410 invert) equipped with a helium/neon laser (543 nm) for rhodamine excitation and a long-pass 595-nm emission filter. The nonspecific fluorescent signal was removed by subtracting the averaged intensity value obtained by measures made on an adjacent section incubated with the secondary fluorescent antibody.

Immunoblots

Immunoblots of Triton X-100 soluble protein from tissues and cell cultures were run on 12% sodium dodecyl sulfate-polyacrylamide gels and transferred to nitrocellulose sheets, as previously described (Kannan et al., 1996). A mouse monoclonal antibody against SNAP-25 (1:6,000 dilution ; Sternberger Monoclonals), and a polyclonal rabbit antibody directed against recombinant human SNAP-23 (1:1,000 dilution ; Galli et al., 1998) were used to compare the relative expression of these proteins in the different extracts. The immunoreaction was revealed with the secondary anti-IgG antibodies coupled to biotin (Euromedex) and 125I-streptavidin (Amersham).

RESULTS

Relative expression of SNAP-25 and SNAP-23 in the adrenal gland

The first five exons of SNAP-25 and SNAP-23 show close sequence homology, and SNAP-23 has been reported to be ubiquitously expressed (Ravichandran et al., 1996) but has not yet been described in the adrenal gland. Immunoblots of rat adrenal gland and pheochromocytoma PC12 cells showed that both SNAP-23 and SNAP-25 are expressed (Fig. 1a). The PC12 cell extracts contained mostly SNAP-25 and very little SNAP-23, whereas the adrenal gland extracts had less SNAP-25 than SNAP-23. As our previous immunofluorescence study had shown that SNAP-25 was restricted to the adrenal medulla (Kannan et al., 1996), we next examined the bovine adrenal gland, because its larger size permits a clean separation of the adrenal cortex and the adrenal medulla. As shown in Fig. 1a, SNAP-23 protein is expressed in the adrenal cortex, but SNAP-25 is hardly detectable. In the adrenal medulla, the expression is reversed ; there is relatively more SNAP-25 than SNAP-23. However, in primary bovine adrenal medullary cultures, SNAP-23 levels were higher, and both SNAP-23 and SNAP-25 were found. Immunofluorescence confirmed the presence of both SNAP-23 and SNAP-25 in chromaffin cells, and only SNAP-23 in the fibroblasts and endothelial cells present in these cultures (data not shown).

Figure 1.

Relative expressions of SNAP-25 and SNAP-23. a : Autoradiograms of immunoblots of adrenal tissue and cell extracts for SNAP-23 and SNAP-25. Blots of 10 μg of Triton X-100-soluble protein fractions (rat adrenal and PC12 cells, and bovine adrenal cortex, adrenal medulla, and adrenal medullary cultures) were reacted with anti-SNAP-23 and anti-SNAP-25 antibodies. b : PCR products for SNAP-23 and SNAP-25 mRNA. cDNA corresponding to 50 ng RNA was amplified in parallel with SNAP-23 or SNAP-25 primers for 30 cycles ; 10-μl aliquots of PCR reactions were run on agarose gels and visualized with ethidium bromide. The PCR product of 392 bp corresponds to the mRNA for SNAP-23a, and that of 559 bp to the mRNA for total SNAP-25, but does not distinguish between a and b isoforms.

FIG. 1.

The distribution of SNAP-23 and SNAP-25 in the adrenal gland was confirmed at the mRNA level, using RT-PCR (Fig. 1b). A PCR product of 392 bp corresponding to the longer transcript SNAP-23a, but not the shorter transcript SNAP-23b, was obtained from all of the adrenal extracts and cells. Lower levels of SNAP-23 mRNA were found in PC12 cells and the bovine adrenal medulla. In contrast to SNAP-25, the level of SNAP-23 mRNA was higher in the cortex than in the medulla.

RT-PCR analysis of SNAP-25 isoforms

Before analyzing SNAP-25 isoform expression in the enriched populations of cultured noradrenergic and adrenergic chromaffin cells, SNAP-25 isoforms were first characterized in rat adrenal gland and bovine adrenal medullary cultures. Extracts of rat cerebellum, which contain essentially SNAP-25b, and extracts of PC12 cells, which contain mostly SNAP-25a (Bark et al., 1995), were run as positive controls for the isoformspecific RT-PCRs. As the SNAP-25 exon 5b primer used shows 78% homology with SNAP-23 sequence, we conducted control PCRs using bovine adrenal cortex extracts containing SNAP-23 mRNA and adrenal medullary culture extracts, which have both SNAP-25 and SNAP-23 mRNA. SNAP-23a mRNA in the gland and adrenal medullary extracts does not interfere with the SNAP-25b isoform-specific PCR amplification, as the 202-bp product expected for SNAP-25b was observed with the adrenal medullary culture and not with adrenal cortex (data not shown).

The isoform-specific RT-PCRs (Fig. 2a) indicated that SNAP-25b was the form expressed in rat cerebellum (>90%), and that SNAP-25a was the major isoform in rat PC12 cells (80%), as we had expected. Like PC12 cells, rat adrenal glands also expressed ~75% of SNAP-25 as the a isoform. However, in bovine adrenal medullary cultures, the expression of SNAP-25a was reduced to 40%, and that of SNAP-25b was higher, ~60%. Comparison of extracts of bovine adrenal medulla and freshly isolated bovine chromaffin cells (Fig. 2b) suggested that both in situ and in culture, bovine adrenal medullary cells express both SNAP-25 isoforms, with slightly more SNAP-25b than SNAP-25a. This isoform expression was particular to the bovine adrenal gland, as SNAP-25b was found to be the major form present in bovine cerebellum (Fig. 2b), as it is in the rat.

Figure 2.

Identification of SNAP-25 isoforms by specific RT-PCR. a : RT-PCRs (28 cycles) for total SNAP-25 (559 bp) and for isoform a (202 bp) and isoform b (202 bp) from extracts of rat tissues and bovine adrenal medullary cultures were run in parallel in triplicate. Initial cDNA was adjusted to give similar quantities of total SNAP-25 product for each extract (1 ng for rat cerebellum, 5 ng for PC12 cells, and 100 ng for rat adrenal and bovine chromaffin cells, as determined from serial dilutions of cDNA). Aliquots (10 μl) of each PCR reaction were run on agarose gels and visualized with ethidium bromide. b : RT-PCR amplification of SNAP-25 isoforms a and b in bovine tissues and cells, i.e., adrenal medulla, freshly isolated chromaffin cells, cultured chromaffin cells, and cerebellar extracts. Initial cDNA corresponded to 1 ng of RNA for the cerebellar extract and 100 ng of RNA for the others.

FIG. 2.

Characterization of total SNAP-25 PCR products

Southern blotting.

These apparent differences in SNAP-25 isoform expression between the rat and bovine adrenal glands were confirmed by further analyses of the total SNAP-25 PCR products. First, Southern blots of total SNAP-25 products were probed either with an oligonucleotide specific for exon 6 and common to both isoforms, or with the oligonucleotide specific for exon 5a or exon 5b (Fig. 3a). Loading of equal amounts of purified PCR products was verified by ethidium bromide staining (Fig. 3a, row I). As shown in Fig. 3a, row II, in agreement with the isoform-specific PCR results reported above, SNAP-25b is the isoform expressed in rat cerebellum and SNAP-25a is the major form in PC12 cells and rat adrenal gland. These results also provide additional evidence that bovine chromaffin cells express slightly more SNAP-25b than SNAP-25a, and that the isoform pattern differs from the rat adrenal gland.

Figure 3.

Isoform characterization of total SNAP-25 PCR products. a : Southern blots of total SNAP-25 PCR products, i.e., rat cerebellum, PC12 cells, rat adrenal, and bovine adrenal medullary cultures. Purified PCR products (500 ng) were separated on agarose gels, stained with ethidium bromide, as shown in row I, blotted, and hybridized with either an oligonucleotide (exon 6) common to both isoforms or with an oligonucleotide specific for either exon 5a or exon 5b (row II). The common oligonucleotide reacted with the total SNAP-25 PCR products obtained from rat cerebellum, PC12 cells, rat adrenal gland, and bovine adrenal medullary cultures. The exon 5a oligonucleotide hybridized very strongly with the PCR product from PC12 cells and rat adrenal gland, weakly with the bovine chromaffin cell PCR product, and negligibly with the rat cerebellum PCR product. In contrast, the exon 5b oligonucleotide reacted strongly with the rat cerebellum PCR product, moderately with the bovine chromaffin cell PCR product, and only weakly with the PC12 cell and rat adrenal PCR products. b : Restriction enzyme digestion of total SNAP-25 PCR products (559 bp) obtained from rat cerebellum, PC12 cells, rat adrenal, and bovine adrenal medullary cultures ; 300 ng of purified PCR product was digested, separated on agarose gels, and stained with ethidium bromide. A 100-bp ladder is shown on the left and an HinfI digest of pBR322 is shown on the right. Digestion of SNAP-25b with StyI should yield two fragments of 233 and 326 bp, and digestion of SNAP-25a with PvuI should yield two fragments of 164 and 395 bp. Although PvuI digestion was never complete, the larger fragment was evident in digests of PC12 cell and rat adrenal PCR products (the 164 bp was barely detectable). According to the mouse sequence, NdeI digest fragments of 477 and 82 bp should be obtained for SNAP-25b, and fragments of 257 and 220 bp (which appear together as a single broad band) and 82 bp for SNAP-25a (the 82 bp was not detectable).

FIG. 3.

Restriction enzyme and sequence analysis.

The restriction maps of mouse SNAP-25b and human SNAP-25a and b contain unique sites for StyI and PvuI in exon 5b and exon 5a, respectively. In Fig. 3b, StyI digestion of the purified total PCR product from rat cerebellum produced the fragments expected for SNAP-25b, whereas PvuI digestion, although incomplete, cut the PCR product from PC12 cells and rat adrenal extracts, confirming the presence of SNAP-25a. It is surprising that the bovine chromaffin cell PCR product was apparently not digested with StyI or PvuI. To clarify this ambiguity, the PCR products were also digested with NdeI. This digestion produced a single band for rat cerebellum, as expected for isoform b. For the other samples, this same band was observed, but also a broad lower band. In the digests of the PCR products from PC12 cells and rat adrenal, the lower band was most intense, as would be expected for isoform a. For the bovine chromaffin cell PCR product, the higher band was major, but was fainter than that observed for the digestion of the rat cerebellum PCR product, and the lower band was barely detectable. The NdeI data provide further evidence that both isoforms are expressed in bovine chromaffin cells.

To try to resolve why StyI failed to digest the bovine chromaffin cell PCR product, but it completely digested the rat cerebellum PCR product, total SNAP-25 PCR products were cloned and sequenced. The exon 5 sequences obtained for SNAP-25a from rat adrenal, SNAP-25b from rat cerebellum, and SNAP-25a and SNAP-25b from cultured bovine chromaffin cells are presented in Fig. 4. For both isoforms, the rat and bovine sequences were almost identical and the predicted amino acid compositions were the same. Four conservative base changes were noted in bovine exon 5a, one of them being in the exon 5a primer. In the sequence of bovine exon 5b, there were only two changes. The change at base 80 eliminates the StyI restriction site and explains why bovine SNAP-25b was not digested with StyI. Furthermore, the full sequence of the total SNAP-25 product obtained for bovine SNAP-25a and b indicated only a few conservative changes in the other exons compared with the sequences for rat (unpublished observations).

Figure 4.

Exon 5 sequence of cloned total SNAP-25 PCR products. Clones for bovine SNAP-25a and SNAP-25b were obtained from bovine adrenal medullary cultures and those for rat SNAP-25a and SNAP-25b from adrenal gland and cerebellum, respectively. Of the eight bovine clones sequenced, five clones corresponded to SNAP-25b and three clones to SNAP-25a. Differences in the predicted amino acid sequences between a and b isoforms are indicated by the boxes, and the underlined sequences correspond to exon 5a and exon 5b primers.

FIG. 4.

Chromaffin cell phenotype and SNAP-25 isoform expression

To address more directly the question of whether both of the principle chromaffin cell phenotypes expressed the same profile of SNAP-25 isoforms, we used cultured bovine chromaffin cells, as they can be isolated in sufficient quantities to permit preparations of enriched adrenergic and noradrenergic cell populations. Catecholamine analyses of these preparations after 4 days in culture indicated that the adrenergic cell fraction contained 93% adrenaline, whereas the noradrenergic cell fraction contained 65% noradrenaline (Table 2). These values can be compared with control primary cultures, which contain 20% noradrenaline/80% adrenaline. Cell counts of these fractions double-labeled by immunofluorescence for dopamine β-hydroxylase and PNMT confirmed that the relative catecholamine contents are approximately the same as the percentages of noradrenergic and adrenergic cells in each fraction (Table 2). It is curious that the intensity of the immunofluorescent PNMT labeling of cells in the noradrenergic enriched fraction was weaker than that observed in the enriched adrenergic fraction, suggesting that the apparent adrenergic cells in this fraction may represent a second adrenergic phenotype.

Table 2. Characterization of fractionated bovine chromaffin cell cultures
Cell fractionCatecholamine content aPercentage of cells b
  1. aTotal catecholamine content corresponded to 48-52 μM for each fraction as determined from lysates of 500,000 cells per well (n = 4). Data are given as percentages of adrenaline.

  2. bCells were immunostained with a general marker of chromaffin cells (tyrosine hydroxylase) and a marker of adrenergic cells (PNMT), and the percentages of cells positive for PNMT were calculated by counting a minimum of 450 cells in at least eight different fields for each sample.

Adrenergic93.7 ± 0.489.8 ± 7.2
Noradrenergic33.8 ± 1.234.5 ± 4.7

TABLE 2.

Analysis of SNAP-25 mRNA expression in the enriched adrenergic and noradrenergic fractions of bovine chromaffin cells demonstrated that although the total SNAP-25 mRNA levels were similar, the isoform profile was not identical for each phenotype (Fig. 5a). In the noradrenergic enriched fraction, which nevertheless contained 35% adrenergic cells, the profile indicated that both isoforms are expressed in about equal quantities. However, for the adrenergic fraction, which contained nearly all adrenergic cells, the Southern blot analysis indicated that the amount of a isoform had decreased significantly, by ~50%, and that SNAP-25b represented 70% of the SNAP-25 mRNA in this population (Fig. 5b).

Figure 5.

SNAP-25 isoform expression in the major chromaffin cell phenotypes. a : RT-PCR products for total SNAP-25 (559 bp), the isoform a (202 bp), and the isoform b (202 bp) on agarose gel stained with ethidium bromide for enriched noradrenergic population (NA) and enriched adrenergic population (A). Initial cDNA corresponded to 100 ng of mRNA. b : Quantitation of Southern blots ; 10 μl of PCR reactions for isoforms a and b was blotted and probed with the SNAP-25 exon 4 oligonucleotide for NA-enriched population (n = 6) and A-enriched population (n = 4). Data were obtained from two cultures (mean ± SD). Bands were quantitated, and the density expressed in arbitrary units. The decrease in SNAP-25a in the A-enriched population was significantly different from the NA-enriched population (p < 0.05 ; ANOVA).

FIG. 5.

Localization of SNAP-25 in bovine chromaffin cells

As our previous immunolocalization of SNAP-25 in the rat adrenal gland (Kannan et al., 1996) had suggested that SNAP-25 is more extensively expressed in noradrenergic chromaffin cells than in adrenergic chromaffin cells, we next compared SNAP-25 distribution in the bovine adrenal medulla with that of the rat by immunofluorescence. As shown in confocal sections of the adrenal medulla, the more intensely fluorescent patches of chromaffin cells, showing immunoreactivity for SNAP-25 both in the cytoplasm and at the plasma membranes, in the rat gland (Fig. 6a), are not apparent in bovine adrenal medulla (Fig. 6b). Instead, SNAP-25 was localized at the plasmalemma of all cells, although some variations in fluorescence intensity was evident between individual cells. These individual staining differences were also observed in cultured bovine chromaffin cells (Fig. 6c), but no clear-cut correlation between the intensity of the plasmalemma staining and the catecholaminergic phenotype was established.

Figure 6.

Cellular localization of SNAP-25. Immunofluorescence of SNAP-25 revealed with a rhodamine-conjugated secondary antibody in cryostat sections of rat adrenal medulla (a) and bovine adrenal medulla (b) and in cultured bovine chromaffin cells (c). The cultured cells were also stained for PNMT (c'). In the confocal sections (a and b), note that patches of intensely fluorescent cells with cytoplasmic staining that are observed in the rat gland [arrow (a)] are not evident in the bovine gland (b). Bars = 10 μm (a and b) and 5 μm (c).

FIG. 6.

DISCUSSION

In our original hypothesis, we suggested that the differential distribution of SNAP-25 in noradrenergic and adrenergic chromaffin cells in the rat adrenal gland might be due to a difference in SNAP-25 isoform expression. This difference does not appear to be linked to expression of the SNAP-25 homologue SNAP-23. Although SNAP-23 is found as the a isoform in chromaffin cells of both rat and bovine adrenal glands, it is expressed at higher levels in the cortex than in the medulla. SNAP-25a is identified as the major form in the rat adrenal and PC12 cells, but, unexpectedly, both isoforms were found in the bovine adrenal medulla in situ and in cultured chromaffin cells. More important, the expression of SNAP-25b was correlated with the adrenergic phenotype in bovine adrenal medulla.

The characterization of SNAP-25 isoforms a and b in rat and bovine adrenal glands by a direct RT-PCR technique, using specific primers for exon 5a or exon 5b, was supported by additional analyses of the total SNAP-25 PCR products by Southern blotting, restriction enzyme digestion, and finally, by cloning and sequencing. The reduced expression of the SNAP-25a isoform and higher levels of SNAP-25b found in the bovine adrenal gland are unlikely to result from the PCR technique itself, as there is only a single base change in the bovine sequence corresponding to the exon 5a primer and none in that of the exon 5b primer. In addition, the overall base sequences of the bovine SNAP-25a and b clones were very similar to those of the rat clones (94% similar ; data not shown), and the amino acid compositions were identical, confirming that the SNAP-25 gene is highly conserved, as has been previously reported for other species as varied as chicken, mouse, human, Torpedo, and Drosophila (Risinger et al., 1993 ; Bark and Wilson, 1994a). These sequence data support the isoform-specific PCR data showing a distinct species-dependent difference in SNAP-25 isoform expression in the adrenal gland. In addition, these data confirm that the PCR products correspond indeed to the SNAP-25 gene, and not to the closely related SNAP-23 gene (Mollinedo and Lazo, 1997 ; Wang et al., 1997).

Exon 5 codes for the membrane-interacting domain, and the two SNAP-25 isoforms are thought to differ in their interactions with the plasmalemma, perhaps due to alterations in their capacity to be palmitoylated due to a rearrangement of the cysteine quartet (Bark et al., 1995). Alternatively, the regroupment of amino acids in this domain results in an overall net increase in the positive charge of SNAP-25a, and this may influence its interaction with other membrane-bound proteins, such as syntaxin (Bark et al., 1995 ; Lane and Liu, 1997). Recently, transfection of mutated forms of this domain, lacking the cysteine residues, has been shown to remain cytoplasmic, confirming that this region is essential for membrane association (Viet et al., 1996). Furthermore, the degree of palmitoylation in transfected mutants with one or two modified cysteine residues also seems to be correlated with membrane association (Lane and Liu, 1997). Transfection of tagged SNAP-25 isoforms in nerve growth factor-treated PC12 cells indicated that isoform b concentrated in the varicosities and neurite endings, whereas isoform a has a more diffuse distribution (Bark et al., 1995). In the rat adrenal gland where the major isoform is SNAP-25a, the present confocal images suggest a more extensive distribution of SNAP-25 in the cytoplasmic compartment of noradrenergic chromaffin cells. Differential regulation of the palmitoylation of SNAP-25a between adrenergic and noradrenergic chromaffin cells in the rat adrenal gland may contribute, in part, to the more extensive distribution of SNAP-25 in noradrenergic chromaffin cells. However, in the bovine adrenal gland where the expression of SNAP-25a expression is lower and SNAP-25b is higher, SNAP-25 appears more closely associated with the plasmalemma of all chromaffin cells. All of these data would be consistent with the idea that SNAP-25b is preferentially bound to the plasmalemma.

In the present study, SNAP-25b mRNA was found to be the major form in both calf and adult rat cerebellum. Earlier morphological studies of the rodent brain had clearly established a differential distribution of SNAP-25 within the CNS (Oyler et al., 1989, 1991). Overall SNAP-25 expression increases during development of the rodent brain, and in some regions, such as the caudate nucleus, it is localized initially in the axons, but later concentrates in presynaptic regions (Oyler et al., 1992). Subsequent isoform characterization has shown that SNAP-25a mRNA predominates during embryonic and early postnatal development when SNAP-25 is present in nerve fibers, whereas SNAP25b mRNA is up-regulated and predominates when SNAP-25 relocalizes to presynaptic regions in the adult brain (Bark et al., 1995 ; Boschert et al., 1996). This dual localization leads to the proposal that SNAP-25 is multifunctional, playing roles in neurite outgrowth and synaptic maturation, in addition to neurosecretion. Increased levels of SNAP-25 expression during neurite outgrowth or after neuronal injury (Sanna et al., 1991 ; Jacobsson et al., 1996), and functional studies using antisense oligonucleotides to inhibit neurite elongation in vitro and in vivo (Osen-Sand et al., 1993) also implicate SNAP-25 in neuroplasticity phenomena. On the basis of the distribution of SNAP-25 isoforms in the rat, Bark and Wilson (1994b) proposed that SNAP-25a has a more broad function, varying from constitutive exocytosis to regulated exocytosis of LD-CVs and SSVs, whereas SNAP-25b is restricted to a specialized role in the fast release of classic neurotransmitters.

The present results on rat neuroendocrine cells, showing that SNAP-25a is the major isoform in the rat adrenal gland and PC12 cells, and confirming a previous report (Bark et al., 1995), fits with the view that this isoform is involved in secretion of LDCVs. SNAP-25 is expressed in other rodent neuroendocrine tissues, such as the pancreas (Jacobsson et al., 1994 ; Sadoul et al., 1995), the anterior and intermediate lobes of the pituitary gland (Aguado et al., 1996), and the pineal gland (Redecker et al., 1997). In the pituitary gland whereas SNAP-25 isoform expression has been examined, SNAP-25a is the predominant isoform (Bark et al., 1995 ; Jacobsson and Meister, 1996). Finding isoforms a and b in about equal quantities in bovine chromaffin cells, and the apparent higher level of isoform b in adrenergic cells, was thus unexpected. If SNAP-25b is linked to fast classic neurotransmission, the higher level of SNAP-25b in the bovine adrenal gland, in particular in adrenergic chromaffin cells, may imply that more regulated release of SSVs occurs than in the rat gland. SSV-like vesicles, although in a minority, do coexist with LDCVs in neuroendocrine cells (Llona, 1995 ; Ahnert-Hilger et al., 1996 ; Redecker, 1996). In bovine chromaffin cells, acetylcholine or catecholamines are also stocked in SSVs (Llona, 1995).

An alternative possibility is that this species-dependent difference in SNAP-25 expression may reflect a fundamental difference in the degree of differentiation in the chromaffin cell population. In the bovine adrenal gland, the increased expression of SNAP-25b, which characterizes mature postmitotic brain neurons, may indicate that the neuroendocrine phenotype of the bovine chromaffin cells is also more mature. Indeed, SNAP-25b seems to be associated with the adrenergic population, which expresses an additional catecholaminergic enzyme, PNMT, and thus may be considered to be the more terminally differentiated neuroendocrine phenotype of the sympathoadrenergic lineage. The suppression of certain proteins associated with the neuronal phenotype, like GAP-43 (growth-associated protein of 43 kDa ; Grant et al., 1994), and the cell-adhesion molecule L1 (Leon et al., 1992) in adrenergic chromaffin cells is in line with this view. Furthermore, cultured bovine chromaffin cells are reported to be less likely to undergo nerve growth factor-induced transdifferentiation toward a neuronal phenotype (Unsicker and Hofmann, 1983) than those from rat (Doupe et al., 1985).

Although the significance of this species-dependent expression of SNAP-25 isoforms in the adrenal gland remains to be established, the present finding that the SNAP-25b in bovine adrenergic chromaffin cells raises the question of the function of this isoform. Determining how the switch in SNAP-25 isoform expression can be provoked should help to clarify our understanding of SNAP-25 expression during development and may shed light on the species-dependent difference observed here in the adrenal gland.

Ancillary