Exposure to Artificial Light at Night Alters Ultradian Rhythms in Endothelial Function

Abstract

Introduction

Endogenous biological timing mechanisms are fundamental aspects of living cells, tissues, and organisms. Virtually every aspect of physiology and behavior is mediated by self-sustaining circadian clocks, which depend on light to synchronize with the external daily environment. However, exposure to artificial light at night (ALAN) can impair temporal adaptations and affect health and disease.

Methods

During a study of the effects of long-term ALAN exposure on cardiovascular function, we serendipitously detected ultradian rhythms in muscarinic receptor dependent relaxation of isolated aortic tissue.

Results and Conclusion

Mice exposed to dark nights displayed an ultradian pattern of maximum endothelial-dependent relaxation that was antiphase between the sexes. Rhythmic patterns of relaxation were abolished by ALAN exposure in both sexes suggesting that ALAN exposure can affect ultradian rhythms in physiology and behavior.

Introduction

Endogenous biological timing mechanisms are fundamental features of living cells. Four types of biological rhythms are typically coupled with environmental factors and mimic the periods of the solar day (circadian), the tides (circatidal), the phases of the moon (circalunar), and the annual seasons (circannual) [1]. Most research has focused on circadian rhythms, an internal representation of solar days. Virtually every aspect of physiology and behavior is mediated by these self-sustaining circadian clocks, which depend on light early in the day to precisely synchronize with the external daily environment. However, exposure to artificial light at night (ALAN) can impair temporal adaptations and affect health and disease [2].

Two other common types of biological rhythms are ultradian (shorter than circadian) and infradian (longer than circadian) rhythms; these rhythms display periods that do not correspond to known geophysical cycles. Ultradian rhythms have been reported for processes as diverse as REM (rapid eye movement) sleep, body temperature, locomotor activity, and metabolism [3]. The extent to which ultradian rhythms are coupled to circadian rhythms remains unspecified [3, 4].

In rats, mice, and hamsters, disruption of the primary circadian clock (suprachiasmatic nuclei) does not eliminate ultradian rhythms [4]. Mice lacking normal Bmal1, Clock, or Per2 circadian genes become arrhythmic in constant dark conditions [5]; nonetheless, robust ultradian rhythms are observed in these mice. Disruption of circadian rhythms generally enhances the expression of ultradian rhythms [4].

During an unrelated unpublished study to investigate the effects of long-term exposure to ALAN on cardiovascular function in mice, we serendipitously noticed a pattern of ultradian rhythms in the data set we collected for muscarinic receptor dependent relaxation of isolated aortic tissue. In common with previous reports on aortic function, we detected diel differences in relaxation patterns [6, 7]. Males and females differed in their daily pattern of endothelial-dependent relaxation (EDR). We also observed that ALAN dampens these rhythms. These data may be relevant to generating hypotheses to promote further investigation of the non-circadian rhythms in myocardial infarctions (MIs) and stroke.

Materials and Methods

Animals

Seven week old male and female Swiss Webster mice arrived in our laboratory from Charles River Labs (Wilmington, MA, USA) and, after a minimum 1 week acclimation period, were randomly assigned to be housed for 8 weeks in immediately adjacent identical vivarium housing rooms with one of two lighting conditions: LD 14:10 (33.2 μW/cm²: 0 μW/cm²) or ALAN (14 h of 33.2 μW/cm²; 10 h of dim light [1.1 μW/cm²]), onset of the light phase (ZT0) was at the same clock time in each room. The mice were individually housed in static polypropylene microisolator cages (30 cm × 18 cm × 14 cm) at an ambient temperature of 22 ± 2°C and provided with Teklad 2018 chow (Madison, WI, USA) and filtered tap water ad libitum. Consistent with our previous studies [8–10], low-level ALAN was supplied by standard LUMA5 LED light strips (HitLights Inc.; 1.5 W/ft, 5000 K “cool white,” 1,200 lumens). All experimental procedures were approved by the West Virginia University Institutional Animal Care and Use Committee, and animals were maintained in accordance with the guidelines established by the National Institutes of Health in the Guide for the Care and Use of Laboratory Animals [11].

Aortic Endothelial Reactivity

Following 8 weeks of experimental light exposure, mice were assigned to one of four timepoints for tissue collection: ZT0 (beginning of light phase), ZT7 (middle of light phase), ZT14 (beginning of dark phase), or ZT19 (middle of dark phase). A total of 176 mice were used in this experiment. All reactivity experiments were performed by the same experimenter in the same order. Two mice of the same sex, one from each lighting condition, were collected per time point for each recording session to achieve a final n = 8–14/group. On each day, a total of 8 mice, 2 at each time point, were collected. Females and males were collected on alternating days. Mice were anesthetized with Euthasol (Patterson Companies, Saint Paul, MN, USA) and then perfused with saline solution and aortas were promptly dissected. To avoid ocular exposure to extraneous ALAN, mice collected at ZT14, and 19 were transported in light-proof chambers and dissections were performed under dim red light; the eyes of the mouse were covered with a foil cap after mice were anesthetized, which remained in place throughout the terminal procedure. Tissue preparation and mounting into force transducer-detector pins was accomplished as described [12, 13], with some modifications. Briefly, following dissection, thoracic aortas were placed in a Petri dish with ice-cold Krebs-Henseleit buffer solution and cleaned of surrounding perivascular adipose tissue. Thoracic aorta were cross sectioned into 2 mm rings, which were rinsed in Krebs-Henseleit buffer and mounted in a myobath chamber (in duplicates), between a fixed point and a force transducer (World Precision Instruments, Sarasota, FL, USA). Rings were incrementally pre-tensioned to ∼1 g and allowed to equilibrate for 30 min in the Krebs-Henseleit buffer bath, aerated with 95% O₂ and 5% CO₂, and maintained at 37°C [13].

Following equilibration at resting 1 g aortic baseline tension, vessel viability was assessed with 50 mM KCl, and rings not generating a rapid response were excluded from the study. To test EDR, aortic rings were pre-constricted with the thromboxane A2 agonist U46619, a stable synthetic analog of the endoperoxide prostaglandin PGH₂ ([1 × 10⁻⁷m] [14]; Sigma-Aldrich; P6126), and a stable tension was reached and recorded after 15 min. Aortic rings were then challenged with increasing doses of the vasodilator methacholine (MCh; from [1 × 10⁻⁹ to 1 × 10⁻⁵m]) every 3 min, a synthetic, non-selective muscarinic receptor agonist. Relaxation was calculated as the percent relaxation for each dose of MCh from the following equation:

$$\text{Percentage}\text{relaxation}=\left(\frac{Z-x}{Z-y}\right)\times 100$$ (1)

where z is the tension after U46619 {1 × 10⁻⁷m} administration, x is the tension after a given dose of MCh, and y is the baseline tension. After the relaxation curve, the system was washed again for 15 min and allowed to return to baseline, after restabilizing baseline tension was readjusted to 1.0 g if necessary. To test endothelial-independent relaxation (EIR), aortic rings were pre-constricted for 10 min with (1 × 10⁻⁷m U46619) and treated with increasing doses of sodium nitroprusside 3 min apart, a vasodilator that relaxes the vascular smooth muscle by producing endothelial nitric oxide (SNP; from 1 × 10⁻⁹ to 1 × 10⁻⁵m; MP Biomedical; 152061) [12, 15–18]. In common with previous findings [12], maximum EDR and EIR were observed at a concentration of 1 × 10⁻⁵m MCh and SNP, respectively. Relaxation values at a concentration of 1 × 10⁻⁵m were then used for downstream analyses for both EDR (MCh) and EIR (SNP).

Statistical Analyses

Maximum EDR values were first analyzed by ANOVA with sex, light condition, and time of day as factors, with follow up pairwise comparisons using SPSS 29.0. One data point was excluded for having a Z score >2.5 from the overall mean. Following significantly different ANOVAs in LD conditions (see results below), we noted that the data appeared to reflect an ultradian rhythm in LD. Based on procedures suggested for identifying circadian rhythmicity in cross-sectional data for a single cycle with an educed rhythm (see Fig. 25 in [19]), we then conducted rhythmicity analysis of all data with an educed period of 12 h via cosinor analysis using Cosinor.Online [20]. ED50 values were determined by nonlinear curve fit to normalized responses using GraphPad Prism 10.5.0. Differences between group means or cosinor curve fit were considered statistically significant at p ≤ 0.05.

Results

In common with previous reports on aortic function, we detected diel differences in relaxation patterns driven by muscarinic receptor activation [6, 7]. Males and females differed in their daily pattern of EDR. We also observed that ALAN dampens these rhythms. Mice exposed to standard dark nights displayed an ultradian pattern of maximum EDR (Fig. 1). The overall ANOVA of EDR was significant (F₍₁₅₎ = 3.466, p < 0.001), with a main effect of sex on relaxation (F₍₁₎ = 7.972, p = 0.005) and an interaction of sex, lighting condition, and time of day to affect maximum relaxation (F₍₃₎ = 4.947, p < 0.003). Pairwise comparisons revealed that in dark night conditions, females had peaks in relaxation at ZT0 and ZT14, whereas males had peaks in relaxation at ZT7 and ZT19 (Fig. 1, left). We performed cosinor analyses with a 24 h period and none of the groups reached significance for a fit to a diel (circadian) rhythm (p > 0.05; Table 1). Given that relaxation appeared to be rhythmic only in dark nights, with peaks and nadirs twice a day approximately 6 h apart, we then performed cosinor analyses with a period set at 12 h to approximate the observed periodicity in the data. Cosine fit with a 12-h period was significant (p < 0.05) only for males and females housed in dark night (LD) conditions. Acrophase for LD females was ZT 0.123, whereas in males, it occurred 7 h later at ZT 7.369, indicating the ultradian rhythm in aortic EDR was antiphase dependent upon sex. Neither sex had rhythmic patterns of relaxation in ALAN conditions based on cosinor analysis (p > 0.05; Fig. 1, right). EIR did not differ by time of day or lighting condition for males (F₍₇₎ = 1.376, p = 0.225) or females (F₍₇₎ = 1.259, p < 0.283), so no further analyses were performed (Table 2). Baseline tension to 10⁻⁷m U46619 did not differ among groups (F₍₁₅₎ = 0.922, p = 0.545), so no further analyses were performed (Table 2).

Fig. 1.

Ultradian rhythms in maximum aortic EDR. Females (top) and males (bottom) exposed to dark nights (LD, left) display antiphase ultradian rhythms in aortic relaxation, whereas significant rhythms are absent in mice of both sexes exposed to dim light at night (ALAN, right). Each point (black circle) represents the average maximum relaxation of two aortic sections from one animal. Gray line is the best fit cosine curve with a 12-h period; p value in lower left is for curve fit by cosinor analysis. N = 8–14/group. *p < 0.05.

Table 1.

Rhythmic parameters determined by cosinor analysis for periods of 12 h (ultradian, left) and 24 h (circadian, right) of maximum aortic EDR in female and male mice exposed to either dark nights (LD) or dim artificial light at night (ALAN)

Cosinor analysis	Ultradian				Circadian
	♀ LD	♀ ALAN	♂ LD	♂ ALAN	♀ LD	♀ ALAN	♂ LD	♂ ALAN
Cosinor model period, hours	12	12	12	12	24	24	24	24
Mesor	52.573	56.261	44.600	46.273	53.220	57.790	44.279	44.975
Amplitude	11.845	13.922	14.111	7.414	4.739	10.069	0.874	10.188
Acrophase, hours	0.123	9.567	7.369	9.319	2.779	3.904	0.711	20.329
Bathyphase, hours	−5.88 or 6.12	3.57 or 15.57	1.37 or 13.37	3.32 or 15.32	−9.22 or 14.78	−8.10 or 15.90	−11.29 or 12.71	8.33 or 32.33
Zero-amplitude test
Df1 (ndf, numerator)	2	2	2	2	2	2	2	2
Df2 (ddf, denominator)	38	43	39	44	38	43	39	44
F value	4.4075	1.9226	9.3937	0.6099	0.4098	3.0328	0.0148	2.8880
p value	0.0190*	0.1586	0.0005*	0.5480	0.6667	0.0586	0.9853	0.0663

Table 2.

LogEC50 values and maximal percent relaxation (at 10⁻⁵m) developed by exposure to MCh (top) and SNP (middle); average tension developed in response to 10^-7 m U46619 (bottom)

​	♀ LD	♀ ALAN	♂ LD	♂ ALAN
% relaxation±SEM 10−5m MCh
ZT0	64.02±7.67	59.07±10.82	33.86±3.52	47.56±2.87
ZT7	49.23±5.49	67.08±7.26	56.70±8.51	38.41±4.07
ZT14	58.58±7.92	45.87±5.76	31.18±3.79	40.13±5.38
ZT19	37.36±5.07	54.18±4.45	59.85±8.10	59.61±9.74
LogEC50 MCh
ZT0	−6.806	−6.992	−6.474	−6.802
ZT7	−6.547	−6.508	−6.736	−6.430
ZT14	−6.898	−6.957	−6.779	−6.738
ZT19	−7.115	−7.122	−6.47	−6.625
% relaxation±SEM 10−5m SNP
ZT0	117.20±5.75	95.78±11.98	102.34±1.84	93.00±4.99
ZT7	96.72±7.07	98.15±2.34	95.13±4.66	109.66±8.87
ZT14	110.80±3.73	110.72±5.85	96.92±3.78	100.85±4.45
ZT19	96.35±7.81	112.53±6.03	98.46±3.11	109.65±4.10
LogEC50 SNP
ZT0	−7.536	−7.603	−7.485	−7.576
ZT7	−7.446	−7.362	−7.636	−7.336
ZT14	−7.745	−7.361	−7.362	−7.462
ZT19	−7.516	−7.286	−7.329	−7.336
Average tension±SEM 10−7m U46619
ZT0	0.82±0.06	0.83±0.08	1.06±0.07	1.13±0.07
ZT7	0.98±0.11	0.87±0.09	0.83±0.10	0.90±0.09
ZT14	0.88±0.07	0.82±0.08	0.98±0.07	0.98±0.05
ZT19	0.93±0.10	0.86±0.09	0.92±0.08	0.87±0.08

Discussion

Previous work has demonstrated a diel rhythm in ex vivo responses to vasoactive drugs in aorta isolated from male rodents [21, 22]. However, to our knowledge, only one study has reported an ultradian rhythm in vasoconstriction of the aorta from male rats [23], and there have been no reports of ultradian rhythms in vasodilation/relaxation or any reports from females. In common with the report on vasoconstriction in male rats [23], in the current study, mice of both sexes exposed to dark nights displayed an ultradian pattern of muscarinic receptor EDR. Furthermore, aorta from males and females displayed antiphase patterns of EDR under dark nights. Exposure to ALAN dampened the ultradian pattern of EDR in tissue from both male and female mice. EIR was not significantly affected by light exposure at night or time of day in aorta tissue either males or females. We have previously reported that cerebrovasculature is altered by exposure to ALAN [10] and that the recovery from insults to the cardiovascular system is altered by ALAN [24, 25]; taken together, the current results from aorta tissue suggest that peripheral microvasculature may also be affected by ALAN.

Although numerous studies have established physiological and behavioral changes in response to disrupted circadian rhythms by exposure to ALAN including changes in metabolism and feeding [26–32], to our knowledge, this is the first demonstration that ALAN can disrupt ultradian rhythms. Although the evidence remains equivocal regarding the interaction between circadian and ultradian rhythms, these serendipitous results suggest that disrupting circadian rhythms by ALAN may also affect an ultradian rhythm. Further studies will be necessary to confirm and extend these serendipitous observations of ultradian rhythmicity and sex differences in aortic function, including those that enhance the temporal resolution of the rhythm description (sampling every 2–4 h) and those that incorporate both sexes from different species and strains.

Cardiovascular function is strongly influenced by circadian rhythms and is under direct control of the molecular circadian clock (reviewed in [33]). Both sympathetic and parasympathetic autonomic vascular responses vary across the day [34–36], and ALAN affects the balance of these responses on cardiovascular systems [37, 38]. Daily variation in the onset of MIs and other negative cardiac events are well documented; risk factors for cardiovascular dysfunction are often coincident with disrupted circadian rhythms [39]. Serious and well-documented cardiovascular events, including MIs, sudden cardiac death, as well as atrial and ventricular fibrillation, generally occur during the early morning [39, 40]. These cardiovascular events often display bimodal peaks (ultradian) in occurrence across the day; again, the predominant peak is during early morning coincident with a daily increase of platelet and prothrombotic factor production [41]. The aforementioned studies implicate both branches of the autonomic system, and the cellular clocks in vascular tissues, in daily rhythmicity in cardiovascular function. Our data suggest that these biological and molecular pathways may warrant future investigation for ultradian rhythms and their implications in vascular health and disease. Further investigation of the regulation of the interaction among circadian rhythms, ultradian rhythms, and EDR seems warranted.

Acknowledgments

We thank James Frazer and Emily Burrage for their technical assistance.

Statement of Ethics

All experimental procedures were reviewed and approved by the West Virginia University Institutional Animal Care and Use Committee approval No. 180102607_R1, and animals were maintained in accordance with the guidelines established by the National Institutes of Health in the Guide for the Care and Use of Laboratory Animals.

Conflict of Interest Statement

Paul Chantler was a member of the journal’s Editorial Board at the time of submission. The authors have no other conflicts of interest to disclose.

Funding Sources

This research was supported by NIH grants R01 NS092388 and R21 AT011238 awarded to R.J.N., as well as by P20 GM109098. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Author Contributions

J.C.W.: formal analyses, data curation, and writing – review and editing. O.H.M.F.: writing, visualization, methodology, investigation, formal analysis, data curation, and conceptualization. A.C.D.: writing – review and editing and resources. P.D.C.: writing – review and editing, investigation, data curation, methodology, and investigation. R.J.N.: writing – review and editing, writing – original draft, visualization, supervision, resources, project administration, investigation, funding acquisition, and conceptualization.

Funding Statement

This research was supported by NIH grants R01 NS092388 and R21 AT011238 awarded to R.J.N., as well as by P20 GM109098. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Data Availability Statement

The raw data summarized in this publication, other directly associated recorded data, and other research materials are not published due to journal limitations but are available upon reasonable request to the corresponding author (james.walton@hsc.wvu.edu).