Stimulation of the α-adrenoceptor triggers the venom production cycle in the venom gland of Bothrops jararaca

Bothrops jararaca, a Brazilian solenoglyphous venomous snake,belongs to the subfamily Crotalinae, family Viperidae. The venom glands of viperid snakes are related to salivary glands(Kochva and Gans, 1970) with the peculiarity of having a central lumen where the venom produced accumulates(Kochva, 1960, 1987; Warshawsky et al., 1973; Mackessy, 1991). The viperid venom gland contains three distinct cell types: secretory cells, horizontal cells and mitochondria-rich cells. The secretory cells are the main component of the gland epithelium and comprise more than 80% of the total cell population (Kochva, 1978; Warshawsky et al., 1973).

After a bite or manual extraction of venom, the amount of venom inside the lumen decreases and the secretory epithelium undergoes morphological and biochemical changes. The epithelial cells change their shape from cuboid to columnar, the cisternae of the rough endoplasmic reticulum expand, and venom is synthesized. The maximal synthetic activity of the secretory cells and the highest mRNA concentration are observed 4–8 days after manual extraction. After that, the synthetic activity decreases and the venom is gradually accumulated in the gland lumen, while the epithelium returns to a quiescent stage (Ben-Shaul et al.,1971; Rotenberg et al.,1971; Oron and Bdollah,1973; De Lucca et al.,1974; Kochva,1978; Carneiro et al.,1991; Yamanouye et al.,1997). A complete venom production cycle lasts around 30–50 days.

The mechanisms that control the regulation of venom synthesis and secretion are not well understood. The noradrenenergic innervation of the venom gland appears essential for the venom production cycle because depletion of catecholamines by administration of reserpine blocks the cycle(Yamanouye et al., 1997). Bothα- and β-adrenoceptors seem involved, as both theα-adrenoceptor agonist phenylephrine and the β-adrenoceptor agonist isoprenaline reverted many of the morphological changes in the venom gland seen after administration of reserpine(Yamanouye et al., 1997).

The aim of the current study was to investigate the role of theα-adrenoceptor during the venom production cycle in the Bothrops jararaca venom gland. Here, by functional studies using microphysiometry,we demonstrate that the α-adrenoceptor present in the snake venom gland has low sensitivity to noradrenaline and phenylephrine and undergoes a long-term desensitization after stimulation. We also show that stimulation of the α-adrenoceptor is essential for the onset of venom production.

Adult Bothrops jararaca Weid 1824 of both sexes (N=44),weighing 100–400 g, were captured from the wild, classified by the Laboratory of Herpetology from Instituto Butantan, and treated and kept as described by Breno et al.(1990). These snakes had no access to food for 40 days to prevent loss of venom and to make sure that the cells were in quiescent state. It is important to point out that in the wild snakes can fast for 1–2 months but the fast can exceed 1 year(Secor and Nagy, 1994). For manual venom extraction (Belluomini,1968), the snakes were anesthetized with sodium pentobarbital (20 mg kg^–1, s.c.). Animal care and procedures used were in accordance with the guidelines of the Animal Ethics Committee of Instituto Butantan.

l-phenylephrine hydrochloride, (–)-arterenol bitartrate,(–)-propranolol hydrochloride, crystalline reserpine, hyaluronidase,antibiotics and antimycotic powder and 0.4% Trypan Blue solution were purchased from Sigma Chemical Co. (St Louis, MO, USA). Collagenase was purchased from Worthington Biochem. Co. (Lakewood, NJ, USA). Sodium pentobarbital was purchased from Cristália (São Paulo, Brasil). DMEM and foetal bovine serum were purchased from Gibco (Rockville, MD, USA). Other chemicals were of analytical or reagent grade, and were purchased from commercial suppliers.

Preparation of dispersed cells was based on the protocol described by Yamanouye et al. (2000) with some modifications. Briefly, snakes were anesthetized with sodium pentobarbital (30 mg kg^–1, s.c.), decapitated and the venom glands removed. The venom glands were freed from connective tissue and venom,and were cut into slices of 250 μm (McIlwain Tissue Chopper –Brinkmann). Cells were dispersed in an orbital shaker for 90 min at room temperature in Krebs–Hepes solution (composition in mmol l^–1: NaCl 120; KCl 4; MgSO₄ 1.2;KH₂PO₄ 1.2; Hepes 15 and glucose 10; pH 7.4) containing collagenase (3 U mg^–1 of wet tissue), hyaluronidase (3.5 U mg^–1 wet tissue) and antibiotics and antimycotic (100 i.u. ml^–1 penicillin, 100 μg ml^–1 streptomycin and 0.25 μg ml^–1 amphotericin B, respectively). After dispersion, the cells were washed with Dulbecco's Modified Eagle Medium (DMEM)containing sodium bicarbonate (40 mmol l^–1), antibiotics and antimycotic, filtered through a nylon mesh, then washed again and resuspended in DMEM containing 10% foetal bovine serum. All procedures were done under sterile conditions. Viable cells were counted in the presence of 0.4% Trypan Blue and 2×10⁶ cells well^–1 were plated in a 24-well plate. The dispersed cells were cultured at 30°C in a humidified incubator (5% CO₂) for up to 2 days.

Energy metabolism in living cells is tightly coupled to cellular ATP usage,so that any events that perturb cellular ATP levels, such receptor activation and initiation of signal transduction, will result in a change in acid excretion. Changes in the rate of cellular acid secretion have been used as a real-time indicator for changes in functional activity upon receptor stimulation in a wide variety of cells(Hafner, 2000), and have been used to characterize G protein-coupled receptor(Pihlavisto and Scheinin,1999; MacLennan et al.,2000; Meloy et al.,2001).

Extracellular acidification rates were measured using a four-channel Cytosensor Microphysiometer system (Molecular Devices, Menlo Park, CA, USA). Dispersed cells were suspended in a 30% agarose solution and 0.5–1.0×10⁶ cells were placed in a capsule cup. The capsule cups were loaded into the sensor chamber and the chambers were perfused at a flow of 100 μl min^–1 with bicarbonate-free DMEM (containing 40 mmol l^–1 NaCl and 0.57 mmol l^–1 ascorbic acid) at 30°C. The cells were allowed to equilibrate for at least 45 min before exposure to agonists. The pump cycle time was 3 min, which included a 40 s pump-off period during which the acidification rate was measured. Drugs were diluted and perfused through a second fluid path to the sensor chamber. Cells were exposed to agonists for 1 min and a 30 min wash period was employed between successive agonist exposures. This stimulation protocol was found to induce reproducible responses in preliminary experiments. The rate of extracellular acidification was calculated by the Cytosoft Program (Molecular Devices Corporation,Sunnyvale, CA, USA). Changes in the rate of acidification were calculated as the difference between the maximum effect after agonist addition and the measurement taken immediately before agonist addition (basal acidification rate). Baseline acidification rates were normalized to 100% and changes due to agonist exposure were calculated as percent increases over normalized baseline.

Venom glands were removed from snakes from which no venom had been extracted for at least 40 days in order to have cells in a quiescent stage. Dispersed cells were placed in a Cytosensor system for the measurement of extracellular acidification rate. Dose–response curves were constructed for phenylephrine and noradrenaline. These experiments were done in the presence of 10^–4 mol l^–1 propranolol to block β-adrenoceptors present in the venom gland (Yamanouye et al., 1997, 2000). This dose of propranolol completely blocks the response of isoprenaline, aβ-adrenoceptor agonist, in this preparation.

We used snakes that had fasted for 40 days, and from which no venom was extracted during this period. After that, we extracted the venom and waited a variable period (<5 min, and 4, 7, 15 or 30 days) until the snake was killed and cells harvested from the venom gland. A control group was treated similarly except that at the time of sacrifice no venom had been extracted for at least 40 days. Dispersed cells were placed in the Cytosensor system and dose–response curves to phenylephrine were constructed.

Quiescent cells were harvested from snakes from which no venom had been extracted for at least 40 days. The dispersed cells were incubated for 5 min with a high concentration of noradrenaline (10^–4 mol l^–1), kept at 30°C for 24 h, and placed in the Cytosensor system for the construction of dose–response curves to phenylephrine.

Snakes were injected with reserpine (20 mg kg^–1 body mass,s.c.) to deplete endogenous catecholamine stores. Venom was extracted 24 h later, and venom extraction was immediately followed by an injection of phenylephrine (100 mg kg^–1 body mass, s.c.) or vehicle. On the day of venom extraction and the subsequent 13 days, snakes continued to receive injections with reserpine (5 mg kg^–1day^–1, s.c.). 1 day after the last injection, the snakes were anesthetized and killed and their venom glands removed. The venom glands were prepared for electron microscopy as described by Carneiro et al.(1991). Ultrathin sections (70 nm) were analyzed using an LEO 906E transmission electron microscope.

Dose–response curves were fitted through a non-linear regression and pD₂ values (–logEC₅₀) were calculated using the curve-fitting program Graph-Pad PRISM 3.0 (GraphPad Software, San Diego, CA,USA). The results are expressed as means ± s.e.m. of at least 3 independent experiments with 5–12 replicates each. Statistical significance (P<0.05) was assessed using Student's t-test to compare two values, or analysis of variance (ANOVA) followed by the Newman–Keuls test for multiple comparisons.

To stimulate α-adrenoceptors, we used noradrenaline or phenylephrine. To avoid stimulation of β-adrenoceptors, all experiments were done in the presence of propranolol, which is known to block increases in cAMP production in the venom gland induced by the β-agonist isoprenaline(Yamanouye et al., 2000). Phenylephrine is an α-adrenoceptor agonist in mammals, and reverses the effect of reserpine in the venom gland(Yamanouye et al., 1997),suggesting that it is an α-adrenoceptor agonist in snakes.

Noradrenaline and phenylephrine increased the rate of extracellular acidification in a dose-dependent manner in dispersed cells in the quiescent stage of the venom production cycle (Fig. 1). The pD₂ values were 4.77±0.09 (N=7)and 3.75±0.07 (N=11), respectively. The pD₂ values for phenylephrine in absence or presence of propranolol were not significantly different (data not shown).

In this experiment we measured the effect of the α-adrenoceptor agonist phenylephrine on venom gland cells that were harvested from the snake at various times after venom extraction. The pD₂ fell from 3.75±0.07 (N=11) in quiescent cells to 3.27±0.02(N=6) immediately after venom extraction (within 5 min, P<0.05, Fig. 2),and remained reduced for at least 15 days: day 4, 3.32±0.04(N=6); day 7, 3.31±0.03 (N=5) and day 15,3.29±0.04 (N=6), P<0.05). 30 days after venom extraction, pD₂ had returned to quiescent cells values(3.64±0.04, N=6).

We hypothesized that the long-term desensitization might be due to stimulation of the α-adrenoceptor by noradrenaline released during venom extraction. To test if administration of noradrenaline desensitizes theα-adrenoceptors in the venom gland, quiescent cells were dispersed,incubated for 5 min with a high concentration of noradrenaline(10^–4 mol l^–1), and kept at 30°C for the next 24 h. This treatment displaced the dose–response curve for phenylephrine to the right (P<0.05, Fig. 3). The pD₂values in cells stimulated with noradrenaline (3.20±0.02, N=6)was close to the value seen in cells harvested soon after extraction of venom(3.27±0.02, N=6).

After venom extraction, the secretory cells increase in size and assume a columnar shape, the rough endoplasmic reticulum intracisternal spaces expand,many secretory vesicles appear near apical membrane and the Golgi apparatus become well developed (Yamanouye et al.,1997). In snakes that were treated with reserpine for 15 days,beginning the day before the extraction of venom, the cisternae of the rough endoplasmic reticulum were narrow and parallel and the Golgi apparatus appeared quiescent (Fig. 4A; Yamanouye et al., 1997). In a previous study we show that chronic treatment with phenylephrine (100 mg kg^–1 body mass, s.c., 10 days after venom extraction)reverses the effect of reserpine on morphology of venom gland cells(Yamanouye et al., 1997). Considering that α-adrenoceptors undergo a long-term desensitization we hypothesized that their stimulation can trigger the venom production cycle. In this study, a single injection of phenylephrine (100 mg kg^–1body mass, s.c.) given immediately after the extraction of venom, reversed the effect of reserpine: the morphology of venom gland cells after this treatment(Fig. 4B) seemed similar to that of untreated cells. This corroborates our hypothesis about the importance of the α-adrenoceptors at the onset of the venom production cycle.

In this report we show that the α-adrenoceptor present in the snake venom gland has a low sensitivity for noradrenaline and phenylephrine and undergoes a long-term desensitization. The sensitivity of theα-adrenoceptor in the venom gland of Bothrops jararaca is lower than the value we reported for the α-adrenoceptor in the aorta of this snake (Yamanouye et al., 1992; Yamanouye and Picarelli,1992), suggesting that this snake expresses different subtypes ofα-adrenoceptors. The low sensitivity found for the α-adrenoceptor of the venom gland is similar to those found in rat vas deferens circular muscle or spleen (Caricati-Neto et al.,1992; Buckner et al.,1996).

Although the sensitivity of the α-adrenoceptor in the venom gland is low, it may have an important role because the amount of noradrenaline in the venom gland is high (approximately 3 ng mg^–1 wet tissue; Yamanouye et al., 1997), much higher than in the salivary glands, trachea and heart of the rat or salivary glands and brain of the mouse (De Avellar and Markus, 1990; Murai et al., 1995). We have previously shown that the β-adrenoceptor in the venom gland also has a low sensitivity to β-adrenoceptor ligands(Yamanouye et al., 2000).

Our study shows that the α-adrenoceptor became desensitized immediately after the extraction of venom. Desensitization also occurred when venom gland cells were incubated for 5 min with 10^–4 mol l^–1 noradrenaline, which suggests that theα-adrenoceptor desensitization after venom extraction is due to release of noradrenaline in the venom gland. Whether this rapid desensitization was due to uncoupling or to internalization of the receptor is not clear. We have previously shown that β-adrenoceptors in the venom gland become desensitized by extraction of venom(Yamanouye et al., 2000),suggesting that this desensitization is due to uncoupling. Theα-adrenoceptor remained desensitized for at least 15 days. The long time necessary for recovery of the response suggests that a downregulation process occurred. However, sensitivity was restored 30 days after the extraction of venom. Since both the venom production cycle and α-adrenoceptor resensitization take around 30 days, it seems that the machinery necessary for initiating a new cycle is only ready when the gland is full of venom and the secretory cells have entered the quiescent stage.

We have previously shown that the effect of reserpine on morphology of the venom gland can be blocked with a chronic treatment of phenylephrine (100 mg kg^–1 body mass day, s.c., for 10 days). The current finding that a single dose of phenylephrine of 100 mg kg^–1 body mass,given immediately after venom extraction, is as effective as chronic treatment supports our idea that release of noradrenaline in the venom gland is especially pronounced immediately after the extraction of venom or biting. In previous work we have shown that the stimulation of β-adrenoceptor increases cAMP production in quiescent but not in activated cells, showing its relevance to starting the venom production cycle(Yamanouye et al., 2000). Therefore, taking all these data together, we conclude that noradrenergic innervation in the Bothrops jararaca venom gland is crucial to trigger the venom production and secretion machinery.

In conclusion, when the venom gland is full of venom, the secretory cells are in the quiescent state. Biting or manual extraction of venom triggers the production of new venom, which involves a change in the shape of the cells of the secretory epithelium from cuboid to columnar, expansion of the cisternae of the rough endoplasmic reticulum, and the production of secretory vesicles by the Golgi apparatus. To start this cycle, the presence of noradrenaline is essential: no venom is produced in the presence of reserpine. However, our data show that it is sufficient to have noradrenaline present only in the beginning of the venom production cycle, and that the noradrenergic innervation is not essential once the venom production is on the way. Our data also suggest that extraction of venom induces the release of noradrenaline,which indicates that noradrenaline is a causative rather than a permissive factor in the start of the venom production cycle. Knowledge of the dynamic of the venom gland will be helpful for understanding cellular mechanisms involved in venom production and secretion, and would aid the future production of venom in culture. Moreover, this peculiar exocrine gland is an attractive preparation for the investigation of cellular mechanisms of protein synthesis and secretion.

We thank Dr Guus H. M. Schoorlemmer for comments on this manuscript and Wanda R. Carrela da Silva and Débora Aparecida Moura for technical assistance. This work was supported by grants from FAPESP(Fundação de Amparo à Pesquisa do Estado de São Paulo) to N.Y. (97/07990-0, 02/00422-8). N.Y. and R.P.M. are supported by researcher fellowships from CNPq (Conselho Nacional de Pesquisa) and C.M.K. is supported by a FAPESP graduate fellowship. C.M.K. is a graduate student at the Department of Histology and Embriology, University of São Paulo.