Absence of a serum melatonin rhythm under acutely extended darkness in the horse

Abstract

Background: In contrast to studies showing gradual adaptation of melatonin (MT) rhythms to an advanced photoperiod in humans and rodents, we previously demonstrated that equine MT rhythms complete a 6-h light/dark (LD) phase advance on the first post-shift day. This suggested the possibility that melatonin secretion in the horse may be more strongly light-driven as opposed to endogenously rhythmic and light entrained. The present study investigates whether equine melatonin is endogenously rhythmic in extended darkness (DD).

Methods: Six healthy, young mares were maintained in a lightproof barn under an LD cycle that mimicked the ambient natural photoperiod outside. Blood samples were collected at 2-h intervals for 48 consecutive h: 24-h in LD, followed by 24-h in extended dark (DD). Serum was harvested and stored at -20°C until melatonin and cortisol were measured by commercial RIA kits.

Results: Two-way repeated measures ANOVA (n = 6/time point) revealed a significant circadian time (CT) x lighting condition interaction (p < .0001) for melatonin with levels non-rhythmic and consistently high during DD (CT 0-24). In contrast, cortisol displayed significant clock-time variation throughout LD and DD (p = .0009) with no CT x light treatment interaction (p = .4018). Cosinor analysis confirmed a significant 24-h temporal variation for melatonin in LD (p = .0002) that was absent in DD (p = .51), while there was an apparent circadian component in cortisol, which approached significance in LD (p = .076), and was highly significant in DD (p = .0059).

Conclusions: The present finding of no 24 h oscillation in melatonin in DD is the first evidence indicating that melatonin is not gated by a self-sustained circadian process in the horse. Melatonin is therefore not a suitable marker of circadian phase in this species. In conjunction with recent similar findings in reindeer, it appears that biosynthesis of melatonin in the pineal glands of some ungulates is strongly driven by the environmental light cycle with little input from the circadian oscillator known to reside in the SCN of the mammalian hypothalamus.

Figure 1

(A-B): Averaged equine MT (A) and cortisol (B) rhythms under conditions of light dark (LD 13.5:10.5) and constant darkness (DD). The barn LD cycle is depicted above each graph: white bars represent light in LD and subjective day in DD; black bars and internal shading represent darkness in LD and subjective night in DD (CT14-24). Sampling began at ZT/CT0 in LD and ended at CT22 in DD after 32 h in continuous darkness. Hormone data are presented as mean ± SE for six mares (n = 6). CT0 represents 0700 h; CT2 0900 h, etc. (A) MT remained low during hours of light (L) in LD but not during the corresponding times (subjective day, CT2-CT10) in DD. A 24-h MT rhythm is evident under LD conditions, but not under DD (p < 0.0001). *, ** denote significant difference (p < .05, p < .01) at specific time points (Bonferoni post hoc tests). (B) In contrast, cortisol showed similar 24-h patterns in LD and DD.

Figure 2

(A-D): Individual equine MT (A-C) and cortisol (D) time series throughout the experimental LD and constant dark (DD) conditions described for Figure 1. Due to substantial individual differences in peak MT levels expressed in the first hours of darkness individual MT data were normalized and expressed as a percentage of the ZT16-ZT22 mean (set to 100%). The resulting plots (B, C) illustrate the two different temporal patterns discussed in the text: continuously high levels in B contrasting with eventual MT declines in C. Panel D illustrates the substantial individual and ultradian variation in blood cortisol. Other conventions are the same as in Figure 1.