Differences in recovery processes of circadian oscillators in various tissues after sevoflurane treatment in vivo.

The inhalation anesthetic sevoflurane reversibly suppresses Period2 (Per2) mRNA expression in the suprachiasmatic nucleus (SCN). However, a discrepancy exists in phase shifting of the Per2 expression rhythm between sevoflurane application in rats (in vivo application) and explants (ex vivo application). This investigation aimed to resolve this issue. First, tissues from the SCN, choroid plexus in the lateral ventricle (CP-LV), and choroid plexus in the fourth ventricle (CP-4V), which are robust circadian oscillators, and pineal gland (PG) tissue, which is a circadian influencer, were prepared from Per2::dLuc transgenic rats. Significant phase responses of bioluminescence rhythms for different preparation times were monitored in the four tissue explant types. Second, tissue explants were prepared from anesthetized rats immediately after sevoflurane treatment, and bioluminescence rhythms were compared with those from non-anesthetized rats at various preparation times. Regarding bioluminescence rhythm phases, in vivo application of sevoflurane induced phase shifts in CP-LV, CP-4V, and PG explants according to the times that rats were administered anesthesia and the explants were prepared. Phase shifts in these peripheral explants were withdrawn due to the recovery period after the anesthetic treatment, which suggests that peripheral tissues require the assistance of related tissues or organs to correct phase shifts. In contrast, no phase shifts were observed in SCN explants. These results indicated that SCN explants can independently correct bioluminescence rhythm phase. The bioluminescence intensity of explants was also decreased after in vivo sevoflurane application. The suppressive effects on SCN explants were withdrawn due to a recovery day after the anesthetic treatment. In contrast, the suppressive effects on the bioluminescence intensities of CP-LV, CP-4V, and PG explants remained at 30 days after anesthesia administration. These results suggest that anesthetic suppression is imprinted within the peripheral tissues.

Introduction

General anesthesia, which has been used for over 175 years, is given to patients for analgesia and sedation during surgical procedures. Enhancement of its safety has brought significant advances in surgical procedures [[1], [2], [3]]. Despite the long history of general anesthesia use, its molecular mechanism is still not well understood. Comprehensive microarray analyses of gene expression related to anesthetics have been performed, and the results revealed that sevoflurane, the most widely used inhalation anesthetic, affected the expression of 1.5% of 10,000 genes in various rat organs, including clock genes [4]. Circadian clock is obviously a key biological function in human. If it is perturbated due to unhealthy life styles that pose circadian rhythm disturbances, the risks of serious illness will increase [5]. Although numerous studies have examined the effects of anesthetics on circadian rhythm, a common understanding has not been obtained [6,7]. Recently, desflurane which is new type of anesthetics used in the clinical practices is also reported to affect the rest/activity rhythm in mice [8].

We have been focusing on Period2 (Per2), a component of the ‘‘core loop’’ of the circadian clock [9,10]. Per2 expression in the suprachiasmatic nucleus (SCN), the master clock of the body, is reversibly suppressed by sevoflurane treatment (in vivo application) [[11], [12], [13]]. Although the Per2 expression level rapidly recovers from suppression after anesthetic treatment, phase shifts of the Per2 expression rhythm were not clearly observed. Furthermore, in vivo sevoflurane application does not affect the rest/activity rhythm clearly [12,13]. In contrast, sevoflurane application to SCN explants from Per2::dLuc transgenic rats (ex vivo application) showed suppressive effects on bioluminescence intensities in the same manner as in vivo application; in addition, obvious phase shifts of the bioluminescence rhythm were observed after anesthetic treatment [13,14]. Therefore, considering the recovery of Per2 mRNA expression/bioluminescence after anesthetic suppression, a discrepancy exists between in vivo and ex vivo application in bioluminescence phase shifts. To resolve this discrepancy, the bioluminescence rhythm phases of explants prepared from both anesthetized and non-anesthetized rats were compared in this study because explants are unaffected by the phase correction phenomenon that occurs in related brain areas. If the bioluminescence phases differ between conditions, the circadian clock in the SCN may be capable of in vivo regulation of the circadian time via related brain areas or peripheral tissues. Conversely, a lack of phase differences would suggest that the explants contain machinery to return the perturbed circadian clock to its original phase.

According to recent reports, the choroid plexus in the lateral ventricle (CP-LV) and the choroid plexus in the fourth ventricle (CP–4V) are robust circadian oscillators [[15], [16], [17]]. The CP-LV and CP-4V are sources of melatonin in the brain in addition to the pineal gland (PG), which affects the circadian phases in the SCN via feed-back information [18,19]. In this study, effects of in vivo sevoflurane application on explants from the CP-LV, CP-4V, and PG, in addition to SCN explants, were examined.

Our experimental design was as follows. Four types of explants were prepared at various times of the day, and the bioluminescence phase responses of each tissue according to the explant preparation time were examined. On the basis of the results, the bioluminescence from explants of anesthetized and non-anesthetized rats was compared to examine the effects of anesthetics on the phase and intensity of bioluminescence. Finally, bioluminescence from the explants was examined to investigate whether the anesthetic effects were withdrawn due to the recovery period after the anesthetic treatment.

Materials and methods

Animals

Per2::dLuc transgenic male rats (8–14 weeks old) were used to monitor the bioluminescence of explants [13,14,20]. Rats were kept on a 12 h light/12 h dark cycle at 21 °C with unrestricted access to food (Lab MR stock, NOSAN, Kanagawa, JAPAN) and water. The Committee of Animal Research of the International University of Health and Welfare approved this study (No. 17006).

Anesthetic treatment of individual rats (in vivo application of sevoflurane)

Rats were placed in a chamber (44 x 32 x 16 cm) and exposed to a gas mixture containing 4% sevoflurane (NIKKO Pharmaceutical Co., Gifu, Japan) and 60% O2 at a flow rate of 1 L/min via a vaporizer (NARCOBIT-E(II), Natsume Seisakusho Co., Ltd., Tokyo, JAPAN). Rats were anesthetized for 6 h while the temperature in the chamber was maintained between 26 °C and 29 °C by a heating pad [13].

Experiment 1: detection of the bioluminescence phase response to time for preparing explants

Transgenic rats (n = 4–8) were sacrificed by decapitation at Zeitgeber Time (ZT) 2, 8, 11, 14, and 20. Immediately after decapitation, explants were prepared from the SCN, CP-4V, CP-LV, and PG. Explant preparation and bioluminescence monitoring were conducted in the same manner as in our previous report [17]. The ZT of each peak time point of the bioluminescence rhythm in the explants was determined for the four types of tissue, which were prepared at various times during the day.

The methods used to examine the degree of phase shift are described in the supplemental experimental procedures.

Experiment 2: effects of in vivo sevoflurane application to the four explant types on the phase and intensity of the bioluminescence rhythm

Four types of tissue explants were prepared from anesthetized rats at ZT2, 8, 14, and 20, and the bioluminescence rhythm was monitored. The peaks in anesthetized rats, which appeared 12 h before and after the peaks in non-anesthetized rats, were identified. To predict the effects of anesthetic administration on bioluminescence rhythm phases in explants, the ZTs of bioluminescence peaks in anesthetized rats were compared with the corresponding peaks of explants from non-anesthetized rats that were prepared at the same ZT.

To examine suppressive effects of sevoflurane on the explants, the intensities of each peak (peak intensity of bioluminescence: PIB) in the explants from anesthetized rats were compared with the corresponding PIBs of non-anesthetized rats. The dataset obtained in Experiment 1 was used in Experiment 2 because the experimental conditions of non-anesthetized rats were identical.

Experiment 3: influence of recovery period after the anesthesia

After anesthetic treatment (ZT14–20), the rats were housed in cages in the same manner as before anesthesia. The explants were prepared at ZT20 at 24 h, 7 days, and 30 days after anesthetic administration. The PIB and peak ZT of the bioluminescence rhythm were compared with those of non-anesthetized rats that were prepared at ZT20. The datasets obtained from non-anesthetized rats in Experiment 1 and the datasets from Experiment 2 in which explant-preparation was performed immediately after wearing off anesthesia, were used in Experiment 3.

Statistical analysis

Statistical analysis of the phase response curve was performed using one-way analysis of variance with a Bonferroni post-hoc test. The bioluminescence phase and intensity in the two groups were compared using Welch's t-test or Student's t-test.

Results

Phase response of the bioluminescence rhythm to the explant preparation time

The peak ZT of each type of explant fluctuated depending on the explant preparation ZT (Supplemental Fig. 1). For the first peak, the time lags between the ZT of the earliest and latest peaks were 3.4 h in the SCN, 5.8 h in the CP-LV, 5.9 h in the CP-4V, and 5.5 h in the PG (Supplemental Fig. 2). Cosine curve fitting of the bioluminescence intensity at 1 h after explant preparation (Supplemental Fig. 3) showed that the in vivo peak times were ZT12 in the SCN, ZT12 in the CP-LV, ZT 15 in the CP-4V, and ZT 21 in the PG (Supplemental Fig. 4). As shown in Fig. 1, which is based on Supplemental Fig. 2 and Supplemental Fig. 4, the phase response curve indicated that the most significant phase delay in the SCN occurred at ZT11. However, in the CP-LV, CP-4V, and PG, the most significant delay occurred at ZT14.

Fig. 1

Phase response curves of bioluminescence rhythms according to explant preparation times for different types of tissue. The phase shift compared with the in vivo peak time (Supplemental Fig. 4) is presented as the mean time difference ±standard error. Significant differences between two values are indicated by * P < 0.05 and **P < 0.01.

Effects of and recovery from in vivo sevoflurane application with regard to the bioluminescence rhythm of SCN explants

Regardless of the ZT when the explants were prepared, no difference was observed between anesthetized and non-anesthetized rats regarding the appearance of the first peak of the bioluminescence rhythm in the SCN. These results indicated that in vivo application of sevoflurane did not cause a phase shift in SCN explants (Fig. 2A).

Fig. 2

Effects of in vivo sevoflurane application on bioluminescence rhythm in SCN explants. A: Time lag between first peak ZT of explants from anesthetized rats and those from non-anesthetized rats are shown for each explant preparation ZT. Phase advances (shown in red) and phase delays (shown in blue) are presented as the mean time difference ± standard error. No difference was observed between groups. B: Recovery of bioluminescence intensity of SCN explants after in vivo sevoflurane application in rats. The bioluminescence intensities of first peaks were recorded in rats of 0 h, 24 h after anesthetic treatment (ZT14–20, red) and compared with those of non-anesthetized rats (blue) (*P < 0.05). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

The most common pattern of bioluminescence in explants from the SCN was a lower intensity in anesthetized rats than in non-anesthetized rats (Supplemental Fig. 5). In particular, the first, second, and third PIBs of SCN explants from anesthetized rats that were prepared at ZT2 and ZT20 were significantly lower than those from non-anesthetized rats prepared at the same ZT (Supplemental Fig. 6).

Regarding the recovery of PIB, when SCN explants were prepared 24 h after anesthesia, the PIBs did not show a significant decline compared with those of non-anesthetized rats (Fig. 2B). These results indicated that the SCN required a day to get recovered from the perturbation due to sevoflurane.

Effects of and recovery from in vivo application of sevoflurane with regard to bioluminescence rhythms of peripheral explants

Regarding the phase shift, CP-LV explants that were prepared at ZT20 showed significant delays in the first peak. CL-4V and PG explants also showed phase shifts depending on the preparation time (Fig. 3A). Regarding recovery to the original phase, significant phase delays were observed in CP-LV and CL-4V explants that were prepared at ZT20 at 24 h after anesthesia administration. In contrast, explants from rats that completed 7 days of recovery period after anesthesia showed no significant phase shifts (Fig. 3B). These results indicated that CP-LV and CL-4V explants require the recovery period of several days to withdraw the effects of anesthetics on phase shifts.

Fig. 3

A: Effects of in vivo sevoflurane application on the bioluminescence rhythm in peripheral explants. The time lag between the first peak ZT of explants from anesthetized rats and those from non-anesthetized rats is shown for each explant-preparation ZT. Phase advances (shown in red) and phase delays (shown in blue) are presented as the mean time difference ± standard error (*P < 0.05, **P < 0.01). B: Recovery of bioluminescence phase of explants after in vivo sevoflurane application to rats. ZTs of the first peaks were recorded in rats of 0 h, 24 h, and 7 days recovery after anesthetic treatment (ZT14–20) and compared with those of non-anesthetized rats. C: The bioluminescence intensities of first peaks were recorded in rats of 0 h, 24 h, 7 days, and 30 days recovery after anesthetic treatment (ZT14–20, red) and compared with those of non-anesthetized rats (blue) (*P < 0.05, **P < 0.01). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Regarding the bioluminescence intensity, comparable suppressive patterns were observed in peripheral explants and the SCN (Supplemental Fig. 5). The PIBs of CP-LV, CP-4V, and PG explants from anesthetized rats appeared to exhibit significantly lower intensity than those of corresponding explants from non-anesthetized rats, depending on the time of explant preparation (Supplemental Fig. 6). These results indicated that the suppressive effects of in vivo sevoflurane application persisted in peripheral explants even after wearing off anesthesia.

Regarding the recovery of bioluminescence intensity, the CP-LV, CL-4V, and PG explants prepared at ZT20 showed significantly lower PIBs, even at 30 days after anesthesia administration (Fig. 3C), indicating that the anesthetic-induced suppression of bioluminescence intensity is imprinted within the peripheral tissues.

Discussion

The present study revealed that explant preparation acted as a Zeitgeber on the bioluminescence rhythm of four types of tissue explants due to Per2 transcription machinery. In contrast, vanderLeest et al. reported that the phase of the electrical activity rhythm in SCN explants was not altered by time when the explants were prepared within one day [21]. The electrical activity of the SCN has a significant impact on the Per1 expression machinery [22]. Per2 mRNA expression is driven by the same transcription mechanism as Per1 expression. In the SCN, conflicting results were obtained between our study and the report by vanderLeest et al.. Therefore, the circadian rhythm of transcriptional activity demonstrated in our study was divergent from the electrical activity rhythm. Furthermore, the time lag between the earliest and latest peaks of bioluminescence rhythm in the four types of tissue, as indicated by the phase-response curve, were shorter in SCN explants than those in CL-LV, CL-4V, and PG explants. This may have occurred because of the higher resilience of SCN explants to stress exerted by explant preparation than that of other peripheral explants. The bioluminescence rhythm phases observed in the four types of explants from non-anesthetized rats were set to the “original phase”.

The phase shifts in SCN explants from anesthetized rats were not notable in this study. This result indicated that SCN explants could independently restore the perturbed circadian phases to those observed before anesthesia. According to a previous study, suppression of Per2 mRNA expression in the brain restored the normal expression pattern, while no obvious phase shift was exhibited after anesthetic treatment [[11], [12], [13]]. Extrapolating from this ex vivo experimental result, the SCN in the brain can induce phase correction under in vivo conditions by itself in the same manner as that observed under ex vivo conditions.

Mori et al. reported that in vivo application of sevoflurane inhibited binding of the histone acetyltransferase CLOCK to the Per2 promoter region in the SCN, leading to suppression of Per2 mRNA expression [23]. Immediately after wearing off anesthesia, Per2 expression in the brain or bioluminescence in SCN explants rapidly restored the original phases. Even during the effects of anesthesia, CLOCK/BMAL in the SCN showed enhanced ability to transcribe Per2/luciferase mRNA. Further investigation into the mechanism of this phenomenon involving CLOCK/BMAL is necessary.

Phase shifts in the bioluminescence rhythm appeared after ex vivo sevoflurane application, while the phases showed an increasing trend, which is closely associated with gamma-aminobutyric acid (GABA) receptors [14]. GABA receptors are expressed in the SCN [24], and GABAA receptors have a particular protein region that responds to sevoflurane [25]. GABAA receptor activation suppresses Per1 and Per2 mRNA in the SCN during the middle of the subjective day [26]. These reports suggest that the SCN itself has a mechanism to sense sevoflurane anesthetics. However, the same machinery may not function in the brain when the anesthetic is applied under in vivo conditions. In a previous report, sevoflurane was intermittently administered to a small glass chamber containing explants, and the same concentration used under in vivo conditions was maintained. Additionally, the explants were exposed to air movement [13,14]. However, comparable circumstances such as air movement or high anesthetic concentrations cannot be expected inside the brain in practice. The mechanism responsible for the suppressive effects of Per2 mRNA expression under in vivo conditions may differ from the phase shift mechanism of ex vivo anesthetic application. Regarding suppressing Per2 mRNA expression or bioluminescence, this could be explained if anesthetics act on neurons outside the SCN as follows. The SCN is innervated mainly by the lateral geniculate body and raphe nucleus. Neuropeptide Y and serotonin, which are produced in the lateral geniculate body and raphe nucleus, respectively, have been reported to affect Per2 mRNA expression in the SCN [[27], [28], [29], [30], [31], [32]]. In addition, orexin, which is related to sleep, is produced in the lateral hypothalamic area and has also been reported to affect Per2 mRNA expression [33].

The present study showed that in vivo application of sevoflurane during late night induces phase shifts in CP-LV, CP-4V, and PG explants. These three types of tissues are known to be innervated via the autonomic nervous system [[34], [35], [36]]. The electrical activity of the SCN is decreased at night. Therefore, even if suppressive effects occur in CP-LV, CP-4V, and PG explants, the SCN is not a candidate area for sevoflurane action. For example, autonomic nervous system nuclei could be affected by sevoflurane.

This investigation also revealed that anesthetic treatment suppressed PIBs of four types of tissues. The results suggest that anesthetic treatment suppresses Per2 expression/the bioluminescence rhythm of four types of explants during consecutive hours. Suppression of PIBs in SCN explants was restored after one day of recovery. According to the results of similar reports, Per2 mRNA expression was decreased on the day after anesthesia was applied in mice under in vivo conditions compared with mice without anesthetic treatment [11,12]. After wearing off anesthesia, an in vivo intervention using related tissues is necessary for the SCN to restore the normal bioluminescence intensity level. These related tissues may include the three tissues other than SCN in the brain examined in this study. In contrast, the PIBs of peripheral explants did not recover from suppression after anesthesia even if the rats were bred for 30 days before being sacrificed. The results may suggest that certain epigenetic events such as DNA or histone methylation may contribute to the effects on bioluminescence intensity.

Conclusion

The resilience of the circadian oscillator in the SCN to the effects of sevoflurane was predominant over that of other tissues. In contrast, the bioluminescence intensity, which indicated gene expression in peripheral tissues was prone to remain low after sevoflurane treatment.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.