Propofol Anesthesia Significantly Alters Plasma Blood Levels of Melatonin in Rats

• ❖ Brief propofol anesthesia disrupts the circadian sleep structure in animals and humans

• ❖ Surgery and anesthesia disrupt circadian sleep structure and timing of melatonin release, but the effect of anesthesia alone is not known

• ❖ Propofol anesthesia resulted in a phase advance of secretion of melatonin, suggesting that the disruption of sleep after surgery reflects in part the effects of anesthetics themselves

MELATONIN is a nocturnal hormone mainly released by the pineal gland and acts as a photoneuroendocrine transducer of information on day length.1Pineal melatonin synthesis is under the sympathetic control from the central circadian pacemaker located in the suprachiasmatic nuclei (SCN).2The SCN control the circadian rhythm of rest activity, temperature, and melatonin secretion. These circadian rhythms are synchronized with the light-dark cycle; and the peripheral melatonin secretion provides circadian information to the rest of the body.3Therefore, peripheral melatonin secretion is considered as a biomarker of circadian rhythmicity and is altered in some particular situations of circadian dysregulation, such as the shift work condition4–6and the time zones crossing, the “jet-lag phenomenon.”5,7 

General anesthesia is an altered state of consciousness,8,9which may have some impact on the circadian structure.10Propofol is one of the most widely used intravenous agents because of its pharmacokinetic advantages. In a series of experiments, we previously showed that propofol anesthesia alone (without surgery) disturbed the circadian structure in rodents as well as in humans.10–12We observed a phase advance of the circadian rhythms of body temperature and rest activity after propofol anesthesia in rats under various laboratory conditions,10,12whereas the circadian sleep-wake activity was desynchronized to local time, the days after anesthesia in patients in real life conditions.11Our hypothesis was that propofol anesthesia could equally disturb the circadian rhythm of melatonin.

In this field, general anesthesia combined with surgery decreases the nocturnal peak of melatonin or decreases the total amount of melatonin metabolites in patients.3,13–16Impaired melatonin secretion has been proposed as one of the mechanisms involved in the postoperative sleep disorders observed in patients undergoing anesthesia and surgery.15The difficulty in all these clinical studies is that general anesthesia (using different anesthesia regimens) is always performed in conjunction with surgery and premedications. Thus, it is impossible to distinguish the effects of general anesthesia from those of surgery and other perioperative treatments.17 

The aim of our study was to determine whether propofol anesthesia (without surgery and other medications) disturbs the melatonin secretion during the 24-h postanesthesia period. For this purpose, we examine the effects of an anesthetic dose of propofol on the secretion of peripheral melatonin during 24 h in rats under normal light-dark alternation conditions.

This study, including care of the animals involved, was conducted according to the official edict presented by the French Ministry of Agriculture (Paris, France) and the recommendations of the Declaration of Helsinki. Thus, these experiments were conducted in an authorized laboratory and under the supervision of an authorized researcher (Y.T.). All research procedures were also performed in agreement with standards and ethics for animal biologic rhythm research.18 

Eighty-four male Wistar rats (Janvier, Le Genest-Saint-Isle, France) weighing 200 to 260 g (5-week old) were housed in individual cages at the beginning of the experiments, with food and water available ad libitum , and maintained in a chronobiologic animal facility (Enceinte Autonome d'Animalerie, Ref. A 110-SP-6, ESI Flufrance, Arcueil, France). The chronobiology facility was equipped with equidistant, sound-proof, temperature-controlled (21°± 1.0°C) compartments, each provided with independent light-dark cycles. Rats were maintained for 3 weeks in a 12-h light/12-h dark cycle (L/D 12:12) before the start of the experiment.

The anesthetic regimen of propofol was conforming to previous published data obtained for rats.19,20Propofol (10 mg/ml; 120 mg/kg; Fresenius Kabi, Sèvres, France) or control solution (intralipids 10%, 10 ml/kg; Fresenius Kabi) was administered intraperitoneally. In Wistar male rats, the duration of general propofol anesthesia (120 mg/kg) ranged between 25 and 30 min. All rats have recovered their righting reflex 30 min after propofol administration.

Rats were divided into 14 groups (n = 6 animals per group) according to the treatment received (propofol or intralipids) and the time of blood samples' collections at +1, +4, +8, +12, +16, +20, and +24 h after propofol or intralipids administration. Thus, for each time point, we injected 12 rats (half with propofol and half with intralipids).

Propofol or intralipids were administered at one Zeitgeber time (ZT) ZT16. In circadian terminology, ZT0 and ZT12 define light on (offset of activity period) and light off (onset of activity period), respectively (fig. 1). Thus, rats were injected 4 h after the onset of the activity period in nocturnal rodents.

Rats were killed by decapitation. Trunk blood samples were collected in plastic tubes with heparin. Samples were kept on ice, centrifuged at 4°C (4,000 g ) for 15 min, and the plasma was stored at −40°C until assay.

Plasma melatonin was measured after dichloromethane extraction by radioimmunoassay with a specific rabbit antiserum (R19540; INRA, Nouzilly, France) and ¹²⁵I-labeled melatonin tracer (Perkin-Elmer, France) according to the method of Brown et al.  21Two hundred microliters of samples were used for the assay. The sensitivity of the melatonin assay was 5 to 10 pg/ml. The intraassay coefficients of variation were 4, 7, and 10%, and the interassay coefficients of variation were 10, 9, and 15% for melatonin concentrations 50, 200, and 1000 pg/ml, respectively.

Statistical analysis of data was performed using two-way analysis of variance (between factor: ZT of blood sample [ZT17, ZT20, ZT24, ZT4, ZT8, ZT12, and ZT16] and treatment [propofol vs.  intralipids]; dependent factor: melatonin concentration). At each time point, multiple comparisons between propofol and intralipids were performed using t  test with Bonferroni correction. Significance was defined as P < 0.05.

Melatonin is secreted according to a circadian (nearly 24 h) rhythm. In rodents, melatonin decreases to its lowest concentrations at ZT0, which corresponds to the offset of activity period, and reaches its highest concentrations during the early hours of the activity period (between ZT12 and ZT16). This variation may be fit to a sinusoidal function by the cosinor method, a linear method of least squares. Cosinor analysis detects a circadian rhythm when the rhythm amplitude is different from zero with a P < 0.05.22Propofol not only modifies the peripheral blood melatonin at different time points but also disrupts its circadian rhythm. Any shift in the circadian rhythm (phase advance or phase delay) has a profound effect on the regulation of sleep-wake cycle and could lead to sleep disorders.4,5,7To examine a putative effect of propofol on the circadian rhythm of peripheral melatonin, cosinor analysis and comparison of linear regressions for the two treatments (intralipids and propofol) were performed using SigmaPlot software (Jandel Scientific, Chicago, IL).

A significant modification of melatonin secretion was observed after propofol administration at ZT16 when compared with control administration (fig. 2). Two-way analysis of variance detected a significant effect of time (F [6,63]= 64.13; P < 10⁻⁴) and no effect of treatment (F [1,63]= 0.018), but a significant interaction between treatment and time (F [6,63]= 3.906; P < 0.002). Post hoc  analysis showed significant differences at ZT17 (P = 0.015), ZT20 (P = 0.004), and ZT16 (P = 0.019). For the first two ZTs after injections (ZT17 and ZT20), propofol significantly decreased melatonin secretion. On the contrary, propofol increased melatonin secretion 24 h after the propofol administration at ZT16. Cosinor analysis detected circadian rhythms of peripheral melatonin for both treatments: propofol (F [2,39]= 48.62, P < 10⁻⁴) and intralipids (F [2,36]= 30.46, P < 10⁻⁴). Comparison between treatments revealed only a trend with a 40-min (CI 95%: 1-81 min; P = 0.068) phase advance melatonin peak cosinor parameter (ZT in h:min ± SD in min) after propofol anesthesia when compared with intralipids (ZT 17:50 ± 27 vs.  ZT 18:30 ± 31 for propofol and intralipids, respectively; fig. 3).

This study shows that general propofol anesthesia decreases the plasma melatonin concentration during the first 4 h after anesthesia and increases the melatonin secretion 20 h after anesthesia. These results demonstrate the disturbing effects of general propofol anesthesia on the circadian rhythm of melatonin in rats under usual alternation of light and darkness.

Many clinical studies suggest that general anesthesia impacts melatonin secretion, although this has never been demonstrated. Importantly, most of the studies have been done when general anesthesia was combined with surgery. General anesthesia for hysterectomy surgery was associated with a decrease in melatonin secretion on the first night after anesthesia.15The amount of 6-sulfatoxymelatonin, the urinary metabolite of melatonin, decreased significantly on the first postoperative night after thiopental or isoflurane anesthesia for orthopaedic surgery when compared with the preoperative night.3The plasma melatonin concentration has been shown to decrease under general anesthesia and surgery16but seemed to be unchanged during the first hours after general anesthesia and surgery.23Conversely, two studies showed an opposite effect. General anesthesia (thiopental and fentanyl or propofol and fentanyl) increased melatonin secretion during anesthesia24or during the 8 h after general anesthesia for surgery.14Finally, the melatonin concentration was unchanged immediately after thiopental or midazolam injection in children.25Such differences from one study to another could be easily explained by differences in anesthesia regimen, co-medications, surgeries performed, and time of sampling for melatonin concentration assessment. From all these previous results, the effect of general anesthesia on melatonin secretion itself remains undetermined.

The decrease in melatonin secretion immediately after propofol anesthesia was followed by an increase in melatonin secretion 20 h after the wake up from anesthesia. The results suggest that this increase may be due to a shift in the circadian rhythm of melatonin secretion (figs. 2 and 3). In a series of experiments, we recently demonstrated that general propofol anesthesia per se  (without surgery and premedication) disturbed the circadian pattern of the rest-activity rhythm during the first postanesthesia days in both rats and humans.10–12Propofol given at an anesthetic dose in rats induced a phase advance of the rest-activity and temperature rhythms.10,12Melatonin secretion and rest-activity rhythm are controlled by the same circadian pacemaker (SCN).16,26Thus, general anesthesia can theoretically create a phase advance in the circadian rhythm of melatonin, which is similar to the phase advance previously observed for the circadian rhythms of rest activity and temperature. In this study, we examined the effects of propofol on plasma melatonin concentration by using seven sampling times during a 24-h period. Our result showed a 40-min phase advance in the circadian melatonin rhythm in the animals submitted to propofol anesthesia when compared (statistical trend) with the control situation (intralipids). From a methodological point of view, we may have missed the onset of melatonin secretion after anesthesia by the limited seven sampling times examined or the statistical significance because of the higher amplitude of the circadian rhythm of melatonin.

Propofol acts via  a positive modulation of the inhibitory function of the neurotransmitter γ-aminobutyric acid (GABA) through GABA receptors type A (GABA^A). GABA^Areceptors are expressed in many brain areas, including the outputs from SCN.27Indeed, GABA seems to be involved in transmitted signals from the SCN to the paraventricular nucleus stimulation of the melatonin synthesis from pineal gland. Infusion of GABA during the activity period inhibits melatonin secretion, whereas infusion during the rest period decreases melatonin synthesis.28However, recent data showed that GABAergic agonists such as muscimol, triazolam, and phenobarbital induce phase advance of circadian rhythm, whereas they modify the expression of genes clock such as per  1 and per  2 within the SCN.29,30The activation of GABA^Areceptors within the SCN may inhibit the clock genes expression and may constitute a common mechanism for these drugs to induce a phase advance of the circadian rhythms. One hypothesis is that propofol, agonist GABA, has an inhibitory effect on the outputs from SCN to paraventricular nucleus. The alternative mechanism might be a direct effect on clock genes expression within the SCN, which induces a phase advance of the circadian rhythms, including the circadian melatonin peripheral secretion. This is currently under examination.

Propofol anesthesia clearly disturbed the endogenous melatonin rhythm in rodents. These results extend and emphasize the effects of propofol on circadian markers. Indeed, the resetting effect of propofol anesthesia on internal clock, as observed in specific laboratory conditions,10seems to be responsible not only for a desynchronization of rest-activity rhythms to local time in both rodents12and patients11but also for a desynchronization of peripheral melatonin secretion. Melatonin is a key marker to provide day-length information to the body and is especially involved in the regulation of the sleep-wake cycle.31A desynchronization of the sleep-wake rhythm to the local time is well known to be associated with altered vigilance and obvious health consequences (fatigue). However, the circadian rhythm of melatonin is also closely associated with human mood and performance.32In anaesthesiology, and more particularly in ambulatory practice, the quality of recovery has to be considered in terms of health and economic factors. Any disturbance of the peripheral melatonin rhythm is, therefore, clinically important. Unexplained clinical symptoms occur in the days after anesthesia, including fatigue, drowsiness, sleep disorders, and mood alteration even for a short-duration anesthesia for a minor surgical procedure. Such complaints from the patients might be sustained at least by a disturbance in melatonin secretion induced by propofol anesthesia.

There are a number of limitations in this study. Rats are nocturnal animals. Therefore, the peak of melatonin occurs in rats during the activity period, whereas it takes place during the rest period in humans. We acknowledge that our results on rats cannot be directly applied to propofol effects on human melatonin. However, the time chosen for this study mimicked the time period when anesthesia is largely performed in patients, that is, the activity period. This is the reason why we examined the impact of rat propofol anesthesia on the peak of melatonin in the activity period. From our data obtained in rats, we cannot demonstrate that the fatigue, drowsiness, and sleep disorders observed in patients are related to a disturbed circadian pattern of human melatonin. However, our results open new lines of researches to better understand and perhaps correct these symptoms.

However, our study demonstrates for the first time that propofol administered at an anesthetic dose impacts the circadian secretion of melatonin during at least 24 h after anesthesia in rats submitted to light-dark alternation. From a clinical point of view, propofol-induced dysregulation of circadian melatonin rhythm might be one factor that impairs recovery from anesthesia, independent of the surgical procedure.

The authors thank Etienne Challet, Ph.D. (Département de Neurobiologie des Rythmes, Institut de Neurosciences Cellulaires et Intégratives, UMR7168, Centre National de la Recherche Scientifique, Université Louis Pasteur, Strasbourg, France), for his help on cosinor analysis.