Melatonin Production: Proteasomal Proteolysis in Serotonin N-Acetyltransferase Regulation

See allHide authors and affiliations

Science  27 Feb 1998:
Vol. 279, Issue 5355, pp. 1358-1360
DOI: 10.1126/science.279.5355.1358


The nocturnal increase in circulating melatonin in vertebrates is regulated by 10- to 100-fold increases in pineal serotoninN-acetyltransferase (AA-NAT) activity. Changes in the amount of AA-NAT protein were shown to parallel changes in AA-NAT activity. When neural stimulation was switched off by either light exposure or l-propranolol–induced β-adrenergic blockade, both AA-NAT activity and protein decreased rapidly. Effects ofl-propranolol were blocked in vitro by dibutyryl adenosine 3′,5′-monophosphate (cAMP) or inhibitors of proteasomal proteolysis. This result indicates that adrenergic-cAMP regulation of AA-NAT is mediated by rapid reversible control of selective proteasomal proteolysis. Similar proteasome-based mechanisms may function widely as selective molecular switches in vertebrate neural systems.

An important component of vertebrate circadian and seasonal physiology is a large nocturnal increase in circulating melatonin (1), which results from an increase in pineal serotonin N-acetyltransferase (arylalkylamine N-acetyltransferase) (AA-NAT) activity. High nocturnal values decrease rapidly (half-life ∼3.5 min) after light exposure in the middle of the night (2). These changes are regulated by an adrenergic-cAMP mechanism (3), but are otherwise poorly understood.

Analysis of pineal immunoreactive AA-NAT protein (irAA-NAT) (4) indicated a parallel change with AA-NAT activity (5) over a 24-hour period (Fig.1). Exposure to constant light blocked the nocturnal increase (Fig. 1), and light exposure or β-adrenergic blockade in the middle of the night rapidly reduced AA-NAT activity and irAA-NAT (Figs. 1 and 2). This decrease was not due to inhibition of translation, which had only a minor effect (6) (Fig. 2). β-Adrenergic stimulation with isoproterenol blocked the rapid light-induced reduction in AA-NAT activity and irAA-NAT (Fig. 2); this treatment also rapidly reversed the effects of acute light exposure on irAA-NAT (Fig. 2), through a mechanism that required protein synthesis (7) (Fig. 2). These changes were not due to changes in mRNA levels (8) nor were they due to a reversible posttranslational process, because large changes in enzyme activity were always associated with parallel changes in enzyme protein (Figs. to 3). Therefore, changes in AA-NAT activity appeared to result from changes in AA-NAT protein levels.

Figure 1

Rat pineal AA-NAT protein and activity change in parallel. (Bottom) Rats were entrained to a light:dark (LD) 14:10 lighting cycle (light or dark bars) and pineal glands were obtained as indicated. One group of animals was not exposed to darkness on the day tissue was collected (dashed line), and another group was exposed to light for a 15-min period in the middle of the night (dotted line). Samples of pineal glands from two or three animals were analyzed for AA-NAT activity (5) (open circles, solid and dashed lines) or irAA-NAT (4) with As2559 (closed circles, solid and dashed lines); similar results were obtained with As2500 (not shown). (Top) Protein immunoblot analysis of irAA-NAT. Hour of harvest (Zeitgeber time, ZT): (A) 6, (B) 13, (C) 14.5, (D) 16, (E) 17.5, (F) 19, (G) 20.5, (H) 22, (I) 23.5, and (J) 1. Tissue obtained at ZT 21 with (+) or without (−) a 15-min light exposure: (K) −, (L) +; (M) −, (N) +; (O) −, (P) +; the antiserum used for each analysis is indicated. In three independent experiments, AA-NAT activity and irAA-NAT were either undetectable or less than twice background in samples obtained during the day or after several hours of light [P < 0.005, n = 10, Mann-Whitney U test (21)]. Exposure to light (15 min) at night significantly suppressed AA-NAT activity and protein values (P < 0.005, n = 5, Mann-Whitney U test). The SEM of the data presented is less than 30%.

Figure 2

In vivo β-adrenergic control of irAA-NAT. (Top) Animals were treated as indicated between 5 and 7 hours after lights were turned off and then killed under dim red light. AA-NAT activity (5) was measured and irAA-NAT (4) was determined with As2559 and As2500. Data shown are the pooled results of experiments performed on three sets of animals. (Bottom) Treatment groups for immunoblot (lane number): (1) dark, no treatment; (2) dark, then light for 15 min; (3) dark,l-propranolol [5 mg/kg, intraperitoneal (ip)] for 15 min; (4) dark, cycloheximide (CHX, 20 mg/kg ip) for 20 min; (5) dark, isoproterenol (ISO, 5 mg/kg, ip), then exposure to light for 15 min; (6) as in (5) except that cycloheximide was injected 5 min before isoproterenol; (7) dark, then 15 min of light, then isoproterenol (5 mg/kg) in the presence of light for an additional 15 min; (8) same as (7) except that during the initial 15-min period of light exposure, rats were injected with cycloheximide (20 mg/kg). When animals were first exposed to 15-min light at night and then placed in the dark for an additional 15 min, irAA-NAT was not detectable and AA-NAT was barely detectable—consistent with previous reports on AA-NAT activity (7). AA-NAT activity and irAA-NAT values in groups 2 and 3 are statistically lower than values at night [P < 0.025, n = 4, Mann-Whitney U test (21)], group 8 is not statistically different from group 2 (P = 0.09, n = 3, Student'st test), and groups 5, 6, and 7 are different from group 2 (AA-NAT activity: P < 0.001; and irAA-NAT:P < 0.025; n = 3, Student'st test). IrAA-NAT is either undetectable or barely detectable in three experiments in groups 2, 3, and 8, as determined with either As2500 or As2559. Bars indicate SEM.

In vitro studies confirmed that occupancy of the pinealocyte β-adrenergic receptor maintained elevated levels of irAA-NAT because β-adrenergic agonists increased irAA-NAT, and subsequent adrenergic blockade with l-propranolol reversed these effects (Fig. 3). Cyclic AMP appears to be involved because dibutyryl cAMP increased irAA-NAT (9) and preventedl-propranolol–induced reduction in irAA-NAT (Fig. 3) (10).

Figure 3

l-Propranolol–induced decrease in irAA-NAT and AA-NAT activity is prevented by dibutyryl cAMP and lactacystin. Pinealocytes (20) were incubated for 24 hours and then treated with isoproterenol (ISO, 100 nM, 5 hours; not shown) or water (control). All treatment groups shown receivedl-propranolol (PROP, 10 μM, 4 to 5 hours); in addition, one group was also treated with dibutyryl cAMP (DBcAMP, 1 mM, 3.5 to 5.0 hours) and another with lactacystin (LAC, 50 μM, 2.0 to 5.0 hours). IrAA-NAT protein was detected as labeled immunoprecipitated protein (11). AA-NAT activity and irAA-NAT values after 4 and 5 hours of ISO treatment were similar (data not shown). Each determination of activity and protein was based on a pool of cells obtained from two wells; results were normalized to the 4-hour ISO treatment group value (100%) (n = 4). *P ≤ 0.001 compared with ISO + PROP alone for AA-NAT activity;P < 0.025 for irAA-NAT (21).

The l-propranolol–induced reductions in AA-NAT activity and irAA-NAT were prevented by each of six proteasomal protease inhibitors, including lactacystin, calpain inhibitor I, and analogs of these compounds (Fig. 3) (11, 12). This demonstrates that AA-NAT protein is destroyed by the proteasome and suggests that an adrenergic-cAMP mechanism stabilizes AA-NAT activity by preventing proteasomal proteolysis of AA-NAT protein. A reasonable hypothetical mechanism underlying the action of cAMP is inhibition of proteasomal targeting by ubiquitination (13).

Cyclic AMP appears to regulate mammalian AA-NAT activity through complementary stimulation of transcription and inhibition of proteasomal proteolysis of AA-NAT protein. Although transcriptional control is not important in all vertebrates (14), inhibition of AA-NAT proteasomal proteolysis may be conserved (13, 15). β-Adrenergic agents may act in a similar manner to control degradation of proteins in other tissues (13, 16).

These findings indicate that proteasomal proteolysis has a role in neural regulation in vertebrates, as in invertebrates (17). Our results indicate that receptor-regulated proteasomal proteolysis can function as a precise, selective, and very rapid neural switch. In the pineal gland, this mechanism regulates the conversion of minute-to-minute changes in environmental input into profound global changes in physiology (18). Such neurally regulated and selective proteasomal proteolysis may play a similarly important role in other aspects of vertebrate physiology and behavior.

  • * To whom correspondence should be addressed. E-mail: klein{at}


View Abstract

Stay Connected to Science

Navigate This Article