The influence of neuronal electrical activity on the mammalian central clock metabolome

Abstract

Introduction: Most organisms display circadian rhythms in physiology and behaviour. In mammals, these rhythms are orchestrated by the suprachiasmatic nucleus (SCN). Recently, several metabolites have emerged as important regulators of circadian timekeeping. Metabolomics approaches have aided in identifying some key metabolites in circadian processes in peripheral tissue, but methods to routinely measure metabolites in small brain areas are currently lacking.

Objective: The aim of the study was to establish a reliable method for metabolite quantifications in the central circadian clock and relate them to different states of neuronal excitability.

Methods: We developed a method to collect and process small brain tissue samples (0.2 mm³), suitable for liquid chromatography-mass spectrometry. Metabolites were analysed in the SCN and one of its main hypothalamic targets, the paraventricular nucleus (PVN). Tissue samples were taken at peak (midday) and trough (midnight) of the endogenous rhythm in SCN electrical activity. Additionally, neuronal activity was altered pharmacologically.

Results: We found a minor effect of day/night fluctuations in electrical activity or silencing activity during the day. In contrast, increasing electrical activity during the night significantly upregulated many metabolites in SCN and PVN.

Conclusion: Our method has shown to produce reliable and physiologically relevant metabolite data from small brain samples. Inducing electrical activity at night mimics the effect of a light pulses in the SCN, producing phase shifts of the circadian rhythm. The upregulation of metabolites could have a functional role in this process, since they are not solely products of physiological states, they are significant parts of cellular signalling pathways.

Keywords: Circadian clock metabolites; Neuronal activity; Small brain samples; Suprachiasmatic nucleus; ZIC-cHILIC-MS.

Fig. 1

Overview of tissue sampling and processing. a For extraction and measurements of metabolites in SCN and PVN tissue, the brain was first isolated from the mouse, and the hypothalamic region, containing the SCN was cut in 250 µm thick slices. These slices were incubated in ACSF, or ACSF with either 0.5 µM TTX or 15 mM K⁺. b After incubation, the SCN and PVN were extracted from the slices by a sample corer. Because of the tight control over the thickness of the slices (250 µm) and the diameter of the punch (500 µm), the volume of the sample was constant (0.2 mm³). Punches were placed in 50/50 methanol/water and directly snap frozen in N₂. Samples were kept at − 80 C until metabolite extraction. c Metabolites were extracted from the tissue by using a liquid–liquid extraction method with 100 µL chloroform added to the 100 µL 50/50 methanol/water. Samples were homogenized in an ice-cold sonication bath for 3 × 10 min. Between sonication, samples were snap frozen for a short period in N₂. Proteins were cleared from the solution by centrifuging. The top layer was transferred to a clean 0.5 mL tube and dried in a vacuum concentrator. d The dried samples were then reconstituted in 20 µL of 60/40 methanol/water and analysed by liquid chromatography through a ZIC-cHILIC column, followed by mass spectrometry

Fig. 2

Global analysis of the complete data set. a An PLS-DA analysis separated the high K⁺-night group clearly, and to a lesser extend also separated the control-day group. The TTX-day and control-night groups largely overlap. Since silencing the neurons of the SCN with TTX at midday was intended to imitate the midnight state of the SCN this is in line with expectations. The results are similar for the SCN and PVN. b The contribution of the individual metabolites were calculated from the PLS-DA, giving a value of the VIP. The metabolites are ordered on their VIP score and their relative amount per individual sample shown in a heat plot. Among the metabolites with high VIP scores are all measurable metabolites of the TCA cycle, and several from the glycolysis pathway

Fig. 3

Metabolites of the TCA cycle are mainly affected by incubation in high K⁺. a From the main metabolites of the TCA cycle, five were reliably measurable with the ZIC-cHILIC-MS method (boxed). b In both the SCN and PVN, all these metabolites were significantly upregulated by incubation in high K⁺-medium at midnight. In the SCN, there was higher level of malate in the control day group, compared to control night. *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 4

Schematic overview of electrical activity in the SCN at the time of the experiments and its effect on metabolism. a On the network-level, the multi-unit activity of the SCN reaches its peak during the light period, between ZT0 (start light period) and ZT12 (end light period). b At the time of sampling, at ZT6 (day), most individual neurons are electrically active, and fire at a frequency of around 8 Hz. At this time point, neurons were completely silenced by applying tetrodotoxin, which blocks sodium channels. At the second sampling time point, ZT18 (night), most neurons are electrically silent, or fire at a low frequency (~ 2.5 Hz). By increasing the extracellular K⁺, the membrane potential was depolarized, thereby increasing neuronal firing. c Exposure to high K⁺ at midnight severely upregulated all metabolites of the TCA cycle. The difference between midday and midnight was smaller, with one of the five measurable metabolites upregulated