Daily melatonin administration to rats and some hamster strains, by injection or by infusion, will synchronise free-running activity and temperature rhythms in constant darkness. It is also reported to partially or completely synchronize disrupted activity rhythms in constant light (19). Circadian phase can be set in fetal hamsters by maternal injections of melatonin at 24-hour intervals at specific circadian times (45). Timed administration to rats hastens adaptation of activity and melatonin production to forced phase shift, and can change the direction of re-entrainment (49, 53). A phase response curve (PRC) to single injections of melatonin can be demonstrated with small phase advances of at most one hour during the late subjective day (49).
As the pineal is involved in circadian timing, the presumption must be that it is concerned with the timing of the LH surge and indeed with general estrous timing. There is evidence that in rats, timed melatonin administration can mimic the effects of extending the light-dark cycle on the timing of the LH surge. Observations of the peripheral melatonin rhythm itself show a decreased amplitude during proestrus in rodents but with conflicting reports in other species (see (12, 54) for reviews). In the rat, gestation length depends on the ambient light-dark cycle. Small advances or delays of parturition can be induced by daylengths shorter or longer than 24 hours, and the effect can partially be mimicked by timed melatonin administration (55).
A fairly consistent observation in pineal research is the decline in amplitude of the melatonin rhythm in old age (see (12) for references). Pinealectomy accelerates the aging process, and there has been some considerable publicity concerning claims for an anti-aging effect of melatonin (56, 57). The most widely published hypothesis is that melatonin acts as a free radical scavenger and anti-oxidant (58, 59). Being an easily oxidised molecule melatonin does indeed have some anti-oxidant activity. However the quantities of exogenous melatonin required to generate clinically relevant anti-oxidant activity in vivo remain to be specified.
Melatonin inhibits 2-deoxyglucose uptake into the SCN in late subjective day, with no effect at other times, and inhibits electrical activity also during late subjective day (19). In this way melatonin may counter a 'wake' signal from the SCN. The most convincing evidence for a direct influence on the circadian system is the phase-advancing effect of melatonin on the circadian rhythm of electrical activity in cultured SCN (60). The effect was large, acute, and time-dependent, with shifts of up to several hours being observed.
Melatonin appears to function as a paracrine signal within the retina. It enhances retinal function in low intensity light by inducing photomechanical changes and regulating the turnover rates of the photoreceptive apparatuses of rods, cones and the surrounding pigment epithelium (61).
The pineal, the retina, and the SCN together form the basic structures perceiving and transducing non-visual effects of light. Melatonin provides a closed loop to this system. It is reasonable to conclude that in adult mammals melatonin serves to modulate circadian phase and strengthen coupling. Since optimal circadian phase is important to health, this is clearly a very significant function. In fetal and neonatal mammals it may help to program the circadian system and to determine the timing of developmental stages, especially puberty.
The numerous reports of other effects of melatonin in animals and in vitro are beyond the scope of this article. Many of these concern ‘protective’ effects attributed to anti-oxidant activity (62)(review). Evidence for anti-tumour activity of melatonin is now strong and possible mechanisms have been proposed (63). Circadian control of metabolism means that there is much scope for effects of melatonin in this area. Most recently, anti-apoptopic activity of melatonin has been investigated and attributed to mitochondrial mechanisms (64)