27 November 2012. Perhaps no part of the mammalian nervous system has been pored over more thoroughly than the retina. Van Leeuwenhoek first observed its rods and cones using his microscopes in the early eighteenth century, and by the turn of the twentieth century, the Golgi stain allowed Ramón y Cajal to map out the retina’s layered structure and circuitry in some detail. So it came as something of a scientific surprise in 2002 when researchers demonstrated that some cells in the retina’s ganglion cell layer—generally thought of as a way station where signals from rod and cone photoreceptors were integrated and relayed to the brain—are themselves photosensitive and, rather than mediate vision, help establish circadian rhythms (Berson et al., 2002; Hattar et al., 2002).
These curiosities, called intrinsically photosensitive retinal ganglion cells (ipRGCs), also directly influence mood and learning, according to a new study published online November 14 in Nature. Led by Samer Hattar at Johns Hopkins University in Baltimore, Maryland, the study reports that altering light schedules without disrupting circadian rhythms produced depression-like behavior and learning deficits in mice, though not in mice lacking ipRGCs.
Irregular light schedules and disrupted sleep are linked to impaired mood and learning, and sleep disturbances are commonly reported in psychiatric disorders, including schizophrenia (Wulff et al., 2010; see also SRF related news story). Because of the importance of ipRGCs in setting the circadian clock via projections to the suprachiasmatic nucleus (SCN) and other preoptic areas, there has been growing interest in their indirect role in regulating mood (McCarthy and Welsh, 2012) and learning and memory (Ruby et al., 2008) by affecting SCN function.
But ipRGCs also project to limbic regions such as the lateral habenula and medial amygdala, which suggested that these cells might exert effects on mood and cognition by acting directly in these regions. First authors Tara LeGates and Cara Altimus and colleagues present several lines of evidence that this is indeed the case.
7 equals 24
To alter light input to ipRGCs without affecting SCN function, the researchers made use of a certain schedule of light and dark that does not sway circadian rhythms. Typically, altering the circadian cycle—12 hours of darkness and 12 hours of light, which is known as a T24 cycle—causes disruption in sleep and in normal rhythmic patterns of body temperature, hormone levels, and gene expression. However, in previous work (Altimus et al., 2008), the Hopkins researchers had found that maintaining mice in a so-called T7 cycle of 3.5 hours of light and 3.5 hours of darkness caused no such changes—aside from a slight lengthening of the 24-hour circadian period, T7 mice were indistinguishable from those kept on a T24 regimen.
The group first confirmed that the T7 cycle has an impact in these regions by finding light-dependent expression of the c-fos transcription factor in the amygdala, lateral habenula, and subparaventricular nucleus. They next showed that, compared to T24 mice, T7 mice exhibited depression-like behaviors in the sucrose anhedonia and forced swim tests. T7 mice were also significantly different in cortisol measures: even though the circadian rhythmicity of cortisol secretion was indistinguishable from that in T24 mice, overall cortisol levels were higher—a common finding in animal models of depression and in depressed humans.
Interestingly, no significant differences were seen in the T7 mice in three standard tests of anxiety, which the authors take as evidence “that the T7 cycle does not globally influence behavior, but specifically elicits depression- and not anxiety-like behaviors.”
Light on learning
Because both high cortisol levels and depression-like behaviors have been shown to disrupt performance on hippocampal-dependent spatial learning tasks, the researchers tested the ability of T7 mice to find the underwater platform in several versions of the Morris water maze. The T7 mice performed significantly worse than T24 mice, findings that were complemented by in-vitro assessments of long-term potentiation (LTP), which showed marked impairment of LTP in T7 mice. Finally, on a novel-object recognition task, another measure of hippocampal-dependent learning, the T7 mice performed poorly compared to T24 controls.
To investigate whether the differences between T7 and T24 mice were mediated by ipRGCs, the researchers used the same battery of tests and LTP experiments with mice lacking functional ipRGCs. None of the depression-like, learning, or LTP effects seen in wild-type T7 mice were observed.
The research group found that the impairments of T7 mice on the forced swim and novel-object recognition tests could be rescued, as well as their deficits in LTP, with chronic administration of the antidepressant fluoxetine. Though fluoxetine can alter circadian rhythms, it did not do so in T7 mice, including the rhythmicity of cortisol secretion; however, overall levels of cortisol were significantly lower in fluoxetine-treated T7 mice than in T7 controls, which “could lead to lower depression-like behavior and better learning,” the authors write.
The authors say the T7-induced changes are unlikely to involve the SCN because when the SCN has been disabled genetically or through ablation in previous studies, animals exhibit reduced depression-like behavior (Tataroglu et al., 2004; see also SRF related news story). The team proposes that the T7 mouse may be a useful new animal model of depression, in that no genetic manipulation or aversive environments are required for these mice to exhibit depression-like behavior.—Pete Farley.
LeGates TA, Altimus CM, Wang H, Lee HK, Yang S, Zhao H, Kirkwood A, Weber ET, Hattar S. Aberrant light directly impairs mood and learning through melanopsin-expressing neurons. Nature. 2012 Nov 14. Abstract