(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Light affects behavioral despair involving the clock gene Period 1

['Iwona Olejniczak', 'Department Of Biology', 'University Of Fribourg', 'Fribourg', 'Jürgen A. Ripperger', 'Federica Sandrelli', 'University Of Padova', 'Padova', 'Anna Schnell', 'Laureen Mansencal-Strittmatter']

Date: 2021-09

Light at night has strong effects on physiology and behavior of mammals. It affects mood in humans, which is exploited as light therapy, and has been shown to reset the circadian clock in the suprachiasmatic nuclei (SCN). This resetting is paramount to align physiological and biochemical timing to the environmental light-dark cycle. Here we provide evidence that light at zeitgeber time (ZT) 22 affects mood-related behaviors also in mice by activating the clock gene Period1 (Per1) in the lateral habenula (LHb), a brain region known to modulate mood-related behaviors. We show that complete deletion of Per1 in mice led to depressive-like behavior and loss of the beneficial effects of light on this behavior. In contrast, specific deletion of Per1 in the region of the LHb did not affect mood-related behavior, but suppressed the beneficial effects of light. RNA sequence analysis in the mesolimbic dopaminergic system revealed profound changes of gene expression after a light pulse at ZT22. In the nucleus accumbens (NAc), sensory perception of smell and G-protein coupled receptor signaling were affected the most. Interestingly, most of these genes were not affected in Per1 knock-out animals, indicating that induction of Per1 by light serves as a filter for light-mediated gene expression in the brain. Taken together we show that light affects mood-related behavior in mice at least in part via induction of Per1 in the LHb with consequences on mood-related behavior and signaling mechanisms in the mesolimbic dopaminergic system.

Day-length has a profound effect on mood status in humans. Short winter days are often associated with seasonal affective disorder, which is a form of depression. Light therapy can alleviate the symptoms of this disorder, but the mechanisms how light can do this are unknown. Using mice as a model system mimicking the effects of light on depressive behavior in humans, we found that the clock gene Period 1 is an important component necessary to mediate beneficial light effects on depressive behavior.

Here we find that light induces Per1 gene expression in the LHb and suppression of it increases immobility in the FST. Furthermore, lack of Per1 in mice increases the number of light-induced genes 10 hours later in the SCN and LHb, while decreasing this number in the NAc. The profound changes in gene expression in the mesolimbic dopaminergic system are linked to sensory perception of smell and G-protein coupled receptor signaling. Our observations suggest an involvement of Per1 in the light-mediated pathway that regulates mood-related behaviors.

In order to adjust the clock to environmental signals such as light, at least one of the clock components needs to be responsive to the stimulus. This leads to a shift of clock phase in order to synchronize the organism to environmental time. Interestingly, expression of the Per1 gene is inducible in the SCN by a nocturnal light pulse at zeitgeber time (ZT) 22 [ 9 , 10 ]. At this time point, light not only causes eventual adaptation of the mammalian circadian clock to a new time zone but is also most effective in the treatment of some forms of depression, such as seasonal affective disorder [ 11 ]. That light exerts powerful effects on mood and cognition has been documented not only in humans [ 12 – 14 ], but also in laboratory animals [ 15 – 17 ]. The neural basis for the influence of light on mood and learning appear to be distinct retina-brain pathways involving ipRGCs that either project to the SCN to influence learning or to the perihabenular nucleus (PHb) in the thalamus to modulate mood [ 18 ]. Furthermore, anti-depressive effects of light therapy require activation of a pathway leading from the retina via the ventral lateral geniculate nucleus (vLGN)/intergeniculate leaflet (IGL) to the lateral habenula (LHb) [ 19 ].

At the cellular level, the clock mechanism is relying on feedback loops involving a set of clock genes. In mammals the main feedback loop consists of two period (Per1 and Per2) and two cryptochrome genes (Cry1 and Cry2), whose transcription is controlled by the transcriptional activators BMAL1 and CLOCK (or NPAS2). PER and CRY form complexes with additional proteins, enter the nucleus and inhibit the activity of BMAL1/CLOCK complexes stopping their own activation. A second feedback loop involving the nuclear orphan receptors REV-ERB (α, β) and ROR (α, β, γ) regulates the expression of the Bmal1 and Clock genes, whose proteins in turn regulate the Rev-erb and Ror genes [ 7 ]. Posttranslational modifications modulate the activity and stability of clock proteins and thereby contribute to and fine-tune circadian rhythm generation [ 8 ].

The circadian clock has evolved from cyanobacteria to humans in response to the daily light-dark cycle [ 1 ]. The internalization of the regularly recurring alternation of light and darkness allowed organisms to anticipate this change. This enabled them to prepare biochemical and physiological processes in order to optimally respond to the upcoming daily challenges and increase survival in a competitive environment. Malfunctioning or disruption of the circadian clock system results in mammals in various pathologies including obesity, cancer, and neurological dysfunctions [ 2 ]. For the maintenance of synchronicity within the mammalian body and with the environmental light-dark cycle, the suprachiasmatic nuclei (SCN) receive light information directly from intrinsically photosensitive retinal ganglion cells (ipRGCs) [ 3 – 5 ]. This information is converted by the SCN into humoral and neuronal signals to set the phase of circadian oscillators and drive circadian rhythmic coherence in the body [ 6 ].

We were most interested in the 490 genes that were changed by light at ZT8 only in the NAc of wild type mice and appeared to be depending on Per1 gene expression. Some of these genes are likely involved in the light-mediated effects on behavioral despair that we observed ( Fig 1C ). We performed a TopGo analysis of this gene set in order to relate these genes to biological processes and molecular functions ( Fig 4E ). Interestingly, G protein-coupled receptor signaling and sensory perception of smell were the highest-ranking biological processes. This finding was mirrored in the molecular function analysis where G protein-coupled receptor activity and olfactory receptor activity ranked highest ( Fig 4E ). This result is in line with previous observations that showed depression-like behaviors in rats after olfactory bulb removal [ 24 ]. Interestingly, also patients with major depression exhibited olfactory deficits [ 25 ].

We evaluated tissue-specific differences between the genotypes by comparing for each brain region of interest the expression between wild type and Per1 -/- mice (see S1 – S8 Tables). Most of the genes influenced by light in wild type animals appeared to be directly or indirectly regulated by Per1, because only a small fraction of those genes was common to the genes found in Per1 -/- mice after light treatment ( Fig 4D ). Furthermore, we found in wild type the highest number of genes to be modified in the NAc (490) and the lowest number in the VTA (4). These differences in numbers were probably due to the fact that the signal progressing from the LHb to the NAc is a dynamic process and that ZT8 was the optimal time point to detect changes in the NAc, because PER1 protein levels were high at that time point. Interestingly, the number of light-affected genes was higher in all brain regions of Per1 -/- mice, except for the NAc ( Fig 4D ). This observation highlights the important function of PER1 on pathways which modulate gene expression in the NAc at ZT8.

Next, we compared the genes altered in the different brain regions and genotypes ( Fig 4C ). First and most strikingly, all of the genes influenced by light in the VTA, irrespective of the genotype, were specific for this region only, with no Per1 mRNA detectable. Per1 was not detectable, because we were looking for downstream effects of Per1 gene induction 10 hours after the light pulse when PER1 protein is high ( Fig 4A ) and its mRNA was already gone. Second, with the exception of two genes, all the light-affected genes in wild type mice were brain region specific. Conversely, in the absence of Per1 many genes in the NAc and LHb were detected in the SCN as well (21 and 11, respectively). This indicated that many signaling pathways that respond to light and at ZT8 modulate targets common to several brain regions are modified by PER1. Hence, PER1 appeared to contribute to brain region specific shaping of molecular responses to light.

Because PER1 protein presence and hence its biological function is increased at ZT8 after a light pulse at ZT22 compared to controls, we collected brain tissues at ZT8. We chose the SCN as reference tissue and the LHb, VTA and NAc as tissues known to be involved in the regulation of behaviors related to mood and despair as part of the mesolimbic dopaminergic system [ 19 , 22 ]. The tissues from wild type and Per1 -/- mice receiving no light or light at ZT22 were isolated at ZT8. We validated the specificity of the tissues isolated using marker genes such as lrs4 (SCN), Gpr151 (LHb), Tacr3 (VTA) and Tac1 (NAc) ( S4 Fig ) that were determined from the Allan brain atlas and normalized according to [ 23 ]. RNA sequencing was performed and the volcano plots comparing gene expression under no light versus light at ZT22 for the various tissues and genotypes are shown in Fig 4B . In wild type mice the SCN and LHb showed several genes are up- or down-regulated 10 hours after the light pulse (black dots) with the VTA displaying a much lower number and the NAc showing a massive number of affected genes. In contrast, the change in gene expression is opposite in the Per1 -/- mice. A large number of genes affected by light were observed in the SCN and LHb with low numbers in the VTA and the NAc. These results are consistent with the role of PER1 as a regulator of transcription [ 7 ]. Since the flow of information driven by the light signal goes from the LHb via the VTA to the NAc [ 19 ] the gene expression in the NAc has an opposite dynamic in presence or absence of PER1 ( Fig 4B ).

(A) Expression of PER1 protein at ZT8 in the SCN and the LHb after a light pulse at ZT22 compared to no light pulse. The left panel depicts an example of a Western blot and the right panel shows the quantification of three independent experiments. Unpaired t-test, n = 3, *p<0.05, values are means ± SEM. (B) Volcano plots depicting the changes of gene expression in response to a light pulse at ZT22 in the indicated tissues of wild type and Per1 -/- mice. The RNA sequencing was perfomed 10 hours after the light pulse (ZT8). Black dots indicate significant changes, n = 6. (C) Venn-diagrams illustrating the overlap of gene expression changes in the four nuclei. Top comparison in wild type animals, bottom comparison in Per1 -/- mice. Numbers indicate the amount of genes affected by the light pulse. For the gene names see S1 Table . (D) Venn-diagram comparing wild type and Per1 -/- mice in a single nucleus. Numbers indicate number of genes affected by the light pulse. Genes are listed in S1 – S9 Tables; SCN: wt ( S2 Table ), Per1 -/- ( S3 Table ); LHb: wt ( S4 Table ), Per1 -/- ( S5 Table ); VTA: wt ( S6 Table ), Per1 -/- ( S7 Table ); NAc: wt ( S8 Table ), Per1 -/- ( S9 Table ). (E) TopGO analysis of biological processes (left) and molecular functions (right) in the NAc of wild type animals.

Presence or absence of the Per1 gene changes behavior of mice in the FST. In order to relate this behavioral effect to alterations in the transcriptome, we performed RNA sequencing experiments. Because we were interested in the downstream impact of Per1, we first studied the amount of PER1 protein after the light pulse at ZT22 in the SCN and the LHb. Previous studies showed that after a light pulse at ZT22 murine PER1 protein in the SCN was increased 10 hours later (ZT8) when compared to mice that received no light [ 21 ]. We confirmed this observation in the SCN by Western blot and found a similar increase of PER1 protein in the LHb ( Fig 4A ). This suggested that Per1 is light-inducible in the LHb.

(A) Scheme illustrating the generation of a conditional Per1 allele. For details see methods section. Numbers above black triangles indicate exons; E, EcoRI; neo, neomycin resistance; green triangles, FRT recombination sites; red triangles, loxP recombination sites; box S, short arm of homology; box L, long arm of homology; grey box, Southern probe; blue double arrows with numbers, PCR products representing the wild-type (wt), Per1 loxP (fl), and the Per1 - (Dfl) alleles. (B) 1.5% agarose gel visualizing the PCR products representing the wild-type (wt), Per1 loxP (fl), and the Per1 - (Dfl) alleles. (C) Immunohistochemistry visualizing PER1 (pink) in the LHb region of Per1 fl/fl mice after injection of a control AAV (top panels) or an AAV expressing CRE (red) (bottom panesl). Cell nuclei are visualized by DAPI staining (blue) and GFP (green) shows viral vector infected cells. Scale bar: 50 μm. (D) Quantification of knock-down efficiency in the LHb compared to SCN. Per1 expression is significantly reduced by AVV Syn-iCre compared to control (ctrl). One-way ANOVA with Tukey’s post-test, n = 4, **p<0.01. (E) Western blot analysis. Protein extracts from the LHb of control AAV or AAV iCre-injected animals were separated on a 7% SDS-PAGE and PER1 and CLOCK were detected with the respective antibodies. Sizes of marker proteins are indicated on the left. (F) Immobility time without and with a light pulse at ZT22 in the FST of male mice at ZT6. Animals were injected with AAV expressing Cre recombinase under the synapsin promoter (vSPer1) or control vector. One-way ANOVA with Dunn’s multiple comparisons test, n = 6, *p<0.05 (G) Time in open arm of an O-maze at ZT6. Unpaired t-test, n = 11, p>0.05, values in all experiments are means ± SEM.

In order to challenge the hypothesis that expression of Per1 in the LHb plays an important role in the regulation of despair based and anxiety-related behavior, we deleted Per1 in neurons of the LHb. To this end we generated mice with a floxed Per1 allele ( Fig 3A ). Deletion of exons 4–16 resulted in a reading frame that could only be resumed in exon 22, thereby deleting almost the entire gene. The various alleles could be analyzed by PCR resulting in a 280 bp amplicon for the wild-type allele (wt), a 468 bp amplicon for the floxed allele (fl), and a 585 bp amplicon for the Per1 deleted allele (Per1 - , Dfl) ( Fig 3A and 3B ). To ensure efficient delivery of the Cre recombinase into the LHb we tested the transduction efficiency and tissue penetration of various serotypes of adeno-associated virus (AAV) expressing EGFP ( S3A Fig ). We injected AAV6 containing expression vectors for either a control construct that expresses EGFP under the synapsin promoter or the same construct additionally expressing iCre into the LHb ( Fig 3C ). We observed that presence of CRE (red color) led to strong reduction of PER1 immunohistochemistry signal (pink color), indicating successful Per1 gene deletion. After recovery from the procedure the animals were tested for efficiency of Per1 deletion. A significant decrease of Per1 expression was observed in the LHb but not in the SCN ( Fig 3D , tissue control S3B Fig ). The remaining Per1 mRNA detected in the LHb probably stemmed from Per1 expression in other cells than neurons, such as glial cells. In line with these results was the significant reduction of PER1 protein in the LHb after deletion of Per1 ( Fig 3E ). Animals injected with the virus expressing iCre, termed vSPer1, or with the control virus, were subjected to the FST at ZT6. Mice that received the control virus still responded to a light pulse at ZT22 with a reduction of immobility, whereas the vSPer1 animals did not respond to the light pulse ( Fig 3F ). Interestingly, deletion of Per1 in the LHb was not sufficient to increase immobility time in the FST ( Fig 3F ), which is in contrast to the total body Per1 knock-out ( Fig 1C ). We also evaluated the vSPer1 animals in the elevated O-maze test. Compared to the control mice these animals spent a similar amount of time in the open arm ( Fig 3G ). Overall these results indicate that Per1 expression in neurons of the LHb is highly important for mediating the light effects observed in the FST. However, deletion of Per1 in the LHb is not sufficient to affect immobility time in the FST without a light pulse, nor the time spent in the open arm of the O-maze. This suggests that Per1 expression in other brain regions is involved in the phenotypes observed in Fig 1 .

To corroborate our observations, we performed quantitative real-time PCR analysis. In order to demonstrate the accuracy of the isolation of SCN, LHb, VTA, and NAc tissue from wild type mouse brains, we used markers specific to the corresponding brain region ( S2B Fig ). Mice that received a light pulse at ZT22 were compared to controls not receiving any light. We detected induction of Per1 but not Per2 in the SCN, the LHb and the VTA one hour after the light pulse ( Fig 2D ), which was consistent with the data we obtained by using in situ hybridization. Taken together our data suggest a role of Per1 in the light-mediated effects on despair related behavior and that this may involve the LHb.

In situ hybridization revealing expression of Per1 (A) or Per2 (B) in the SCN, the medial habenula (MHb) and the lateral habenula (LHb). The top panel shows the quantification of Per1 expression in the SCN, the MHb and the LHb, values are means ± SEM. CircWave analysis revealed circadian expression of both genes in the tissues shown (n = 3, p<0.05). The bottom panel shows representative images of brain sections at ZT6 and ZT18 with expression signal (black) of Per1 or Per2 in the respective brain regions. Scale bar: 5 mm. (C) Induction of Per1 mRNA expression after a LP at ZT22 in the SCN (left panels) and in the lateral habenula (right panels). The blue color depicts cell nuclei (Hoechst staining) and the yellow color shows radioactively labeled antisense Per1 riboprobe hybridized to Per1 mRNA. Below the photo-micrographs quantification is shown. Two-way ANOVA with Bonferroni multiple comparisons test, n = 3, *p<0.05, **p<0.01, ****p<0.0001, values are means ± SEM. Scale bar = 200 μm. (D) Quantitative PCR revealed mRNA expression levels of Per1 (left panel) and Per2 (right panel) in the suprachiasmatic nuclei (SCN), lateral habenula (LHb), ventral tegmental area (VTA) and nucleus accumbens (NAc) after 60 min. (light grey bars) and 120 min. (dark grey bars) of a light pulse at ZT22 (0 min., black bars). Per1 mRNA was induced in the SCN, LHb, and VTA, whereas Per2 was not induced in any of the tissues investigated. One-way ANOVA, Tukey’s multiple comparisons test, n = 12–16, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, values are means ± SEM.

Because the LHb is involved in the regulation of the behavioral response of mice in the FST [ 19 ], we performed in situ hybridization experiments on mouse brain sections in order to evaluate, whether the Per1 and Per2 genes were expressed in the LHb and the medial habenula (MHb). We observed, that Per1 was expressed in both the LHb and MHb although with a much lower amplitude than in the SCN ( Fig 2A ). Per2 mRNA was expressed in both the LHb and the MHb but to a much lower extent ( Fig 2B ). Since light at ZT22 affects immobility in the FST ( Fig 1C ), and induces Per1 but not Per2 expression in the SCN [ 9 , 10 ] we tested the effect of a light pulse at ZT22 on expression of the Per1 gene. We detected strong induction of Per1 in the SCN one hour after the light pulse, which was weaker after two hours ( Fig 2C , left panel) consistent with previous findings [ 9 , 10 ]. A similar pattern of Per1 gene induction was observed in the LHb although this was less pronounced when compared to the SCN ( Fig 2C , right panel). Interestingly, a light pulse at ZT14 did not induce Per1 in the LHb contrary to the SCN ( S2A Fig ) paralleling the absence of a light effect in the FST ( Fig 1A ).

Based on these data we performed the subsequent experiments with female mice and applied a light pulse only at ZT22. The FST was performed at ZT6 and immobility time was assessed for the four following days. The data of days 2–4 were pooled to compare immobility times between animals ( Fig 1B ). Consistent with the results in Fig 1A wild type mice showed reduced immobility time at ZT6 when they received a light pulse at ZT22 ( Fig 1C , left panel), but not when they were assessed at ZT18 ( Fig 1C , right panel). Since Per1 is a light-inducible gene in the SCN [ 9 , 10 ], we tested mice lacking this gene in the same paradigm. At ZT6, as well as at ZT18, Per1 knock-out (Per1 -/- ) mice showed increased immobility time compared to wild type animals, and a light pulse did not affect their immobility time in the FST ( Fig 1C ). Hence, Per1 -/- animals showed an increase in behavioral despair, which was not decreased by light in contrast to wild type animals. Next, we used the sucrose preference test, which provides a measure for anhedonia, another characteristic of depression (decreased ability to experience pleasure; [ 20 ]). We observed no difference between wild type and Per1 -/- mice in this experiment ( S1B Fig ). To test anxiety-related behavior in the two genotypes we used the elevated O-maze test [ 20 ]. The Per1 -/- mice spent less time in the open area and did enter it significantly less compared to wild type control animals ( Fig 1D and 1E ). These results indicated that Per1 -/- animals were more anxious than controls.

(A) Immobility time in the forced swim test (FST) of wild type female mice assessed over several days at ZT6 after no light pulse (LP) (black bars), after a LP at ZT14 (red bars), and after a LP at ZT22 (blue bars). Two-way repeated measures ANOVA (n = 13–15), ZT14 LP p = 0.255, ZT22 LP p = 0.014, values are means ± SEM (B) Light pulse treatment and assessment protocol using the FST. The first day after the light pulse the FST was performed but only the data from days 2–4 were pooled and used for further analysis. (C) Immobility time of wild type and Per1 -/- female mice with and without LP at ZT22 assessed at ZT6 (left panel, n = 11–16) or ZT18 (right panel, n = 5–6). Unpaired t-test, *p<0.05, **p<0.01, values are means ± SEM. (D) Time spent in the open parts of an O-maze. Unpaired t-test, n = 8 for each genotype, *p<0.05, values are means ± SEM. (E) Entries into the open parts of an O-maze. Unpaired t-test, n = 8 for each genotype, *p<0.05, values are means ± SEM.

In order to test the effect of light on mood-related behavior in mice, we applied a 30-min polychromatic light pulse (white light) to animals that were kept in a 12-hour light/12-hour dark (12:12 LD) cycle. The light pulse was applied in the dark phase at zeitgeber time (ZT) 14 or 22, with ZT0 being lights on and ZT12 lights off. The mice were then subjected to a forced swim test (FST) at ZT6 for the next five days where their immobility time was assessed (Figs 1A and S1A ). Animals that received no light pulse (black bars) or a light pulse at ZT14 (red bars) showed comparable immobility times (Figs 1A and S1A ). In contrast, mice that received a light pulse at ZT22 (blue bars) displayed lower immobility times compared to control animals. This difference was statistically significant in females ( Fig 1A ), but only showed a similar trend in males ( S1A Fig ), which is consistent with previous findings in rats [ 17 ].

Discussion

A growing body of evidence supports the notion that light therapy can be efficient in the treatment of individuals with seasonal and non-seasonal depression [26–32], while light deprivation can increase depressive-like behavior in various species [33–36]. However, many unresolved questions remain regarding how light therapy can produce its beneficial effects. In this study, we offer evidence that light-mediated induction of the clock gene Per1 in the LHb is involved in the anti-depressant effects of light therapy.

We studied the effect of light on behavioral despair in mice and uncovered a significant contribution of the clock gene Per1 in this process. Similar to the observations in humans, where light exerts powerful effects on mood and cognition [12–14], we observed that light at night affected mice in a comparable way (Fig 1A), although mice are nocturnal and not diurnal. We found that light affected behavior in a despair-based paradigm when given in the late part of the dark phase (ZT22). On the other hand, we did not observe any change in this behavior when light was given in the early period of the dark phase (ZT14), as has been reported for rats [17]. This is very similar to light treatment being more efficient in the early morning than evening for patients with seasonal affective disorder (SAD) (reviewed in [11]). Furthermore, we found that light treatment at ZT22 appeared to be more effective in female than male mice (Figs 1A and S1A). However, in humans such a sex-related discrepancy wasn’t observed in patients treated for SAD with light [37,38]. Although the incidence of depression seems to be higher in females compared to males [39]. Of note is that the testing of mice in the FST was in their rest phase, whereas in humans it is in the activity phase. This puts a limitation on our interpretations.

Interestingly, we could see the light effect when we assessed animals at ZT6 but not at ZT18. This may indicate that either the light-dark transition at ZT12 may have an impact on the behavioral outcome, or that the circadian clock modulates the amount of immobility in the FST. The latter is more likely, because without any light pulse immobility at ZT18 is lower than at ZT6 in the FST ([40], Fig 1C). Consistently, the lack of Per1 not only resulted in the absence of a light response but also in an increased immobility time (Fig 1C). Furthermore, it has been shown that the LHb was not activated by a dark-light transition [19].

Mood-related behavior is a complex trait. Therefore, in addition to despair, we tested sucrose preference and behavior in the elevated O-maze, which relate to anhedonia and anxiety, respectively. The results indicated that wild type and Per1-/- mice are similar in their ability to experience pleasure (tasting of sucrose) (S1B Fig), but Per1-/- mice appeared to be significantly more anxious (Fig 1D and 1E). Hence, Per1 is likely involved in the despair and anxiety aspects of mood-related behaviors. Altogether, these results support the notion that Per1 participates in the regulation of mood-related behaviors.

In mammals, light at night can induce gene expression in the SCN [41]. The clock gene Per1 is among these light inducible genes [9,10]. Since the ipRGCs not only project to the SCN, but also to other brain regions [42], we investigated the habenula [19] for light-inducible Per1 gene expression. We observed that Per1 was expressed in the lateral (LHb) and medial habenula (MHb) (Fig 2A), while Per2 was expressed at a much lower level in these structures (Fig 2B). Per1 gene expression was induced by light in the LHb when applied at ZT22 (Fig 2C), but not when applied at ZT14 (S2A Fig). These results indicated a time-specific inducibility of Per1 in the LHb, which mirrored the effect of light at ZT22 observed in the FST as immobility (Fig 1A). Our findings suggest that light induction of Per1 in the LHb and probably in other brain regions played an important role (Fig 2D). Of note is that Per2 was most likely not involved, because its gene induction by light at ZT22 was minimal or absent (Fig 2D). Interestingly, light at ZT22 elicits phase-advances, which was abrogated in mice lacking Per1 but not Per2 [43]. Hence lack of phase advance and increased immobility in the FST of Per1 knock-out mice inversely correlated with the observation in humans in which advance of sleep phase had a positive effect on depression [11]. Interestingly, the daily amount of sleep was not different between control and Per1 knock-out animals (S5 Fig), indicating that the light response in the FST was not simply due to a more awake state of the Per1 knock-out mice. Taken together, these results suggested that the amount of Per1 gene expression in the SCN and the LHb correlate with despair-based behavior as observed in the FST (Fig 1C).

In order to challenge the hypothesis that the amount of Per1 gene expression in the LHb was relevant for the level of immobility time in the FST, we deleted Per1 by injecting AAVs expressing iCre into the area of the LHb of Per1fl/fl mice (Fig 3C–3E). We observed no increase of immobility in the FST in contrast to Per1 knock-out animals (Figs 1C and 3F), which suggested that Per1 expression in the LHb is not important for immobility time in the FST. Hence, the oscillator properties of the LHb [44] most likely are not involved in mood related behaviors. However, we observed that the response to a light pulse is abolished when Per1 is deleted in neurons of the LHb (Fig 3F). This indicated that the light signal leading to a change of immobility in the FST is depending on Per1 in the LHb. Hence, the LHb clock may contribute to the gating of the light induction of Per1.

The importance of the LHb area in mood regulation has been described previously [45]. Light modulated LHb activity via M4-type melanopsin-expressing retinal ganglion cells (RGCs) thereby regulating depressive-like behaviors [19]. Our observations support these findings, although we do not know whether Per1 is induced in the LHb via the M4-type melanopsin-expressing RGCs. Interestingly, another study described light effects via intrinsically photosensitive RGCs on the SCN and the perihabenula (PHb), regulating hippocampal learning or mood, respectively [18]. Although our experimental set up was different from that study, our findings are not contradictory, because we cannot exclude an involvement of the PHb in our study. The knock-down of Per1 in the area of the LHb had no effect on anxiety (Fig 3G) compared to the Per1 knock-out (Fig 1D and 1E). This highlights that Per1 in other brain regions or body tissues contributed to the phenotype as well. Overall, our results and data by others support the notion, that the area of the LHb and expression of Per1 are involved in the light mediated effects on mood-related behavior.

The molecular mechanisms through which light elicits its beneficial effects on mood-related behaviors are poorly understood. The observation that light induces the expression of the Per1 gene in the area of the LHb provides an opportunity to get a first glimpse at potential molecular pathways that are initiated by the activation of Per1. In order to identify targets of PER1 that are involved in the behavior we observed in the FST, we performed RNA sequencing experiments. Since clock genes regulate behavior in the FST partly via the mesolimbic dopaminergic system [40,46–48] we used tissue from the LHb, VTA, and NAc with SCN tissue as control. The tissues were isolated at ZT8, because light-induced PER1 protein was highest at that time in the SCN [21] and LHb (Fig 4A) and very close to our behavioral assessment in the FST at ZT6. We observed that light-modulated genes were mostly specific to a particular tissue with virtually no overlap with other tissues assessed. In contrast, lack of Per1 increased the overlap of light-affected genes between the tissues (Fig 4C), suggesting a role of PER1 in regulating common light-responsive genes. This evidence is consistent with the known role of PER1 as a suppressor of clock and tumor-related genes [7,49]. Interestingly, this can only be observed in the SCN, LHb and VTA, but not the NAc (Fig 4D). This could be due to indirect regulation of genes in the NAc by PER1 (e.g. suppression of suppressors specific to the NAc) and/or neurotransmitter related gene regulation in the NAc by the VTA or other brain regions. Remarkably, the VTA showed the fewest number of genes affected by light. The reason for this may be the time of assessment, because at ZT8, most of the light-modulated genes in the VTA may have already been silenced again. This highlights the highly dynamic nature of light effects on behavior from initial gene expression changes, activation of signaling pathways to neuronal communication and neurotransmitter release. In order to understand this process better, a dynamic assessment of the transcriptome at several time points (e.g. every two hours after the light pulse) would be necessary. The dynamic temporal change of the transcriptome in the LHb, VTA and NAc after the light pulse may then provide insights into the relationships between the various nuclei to translate the initial light signal into a systemic change that affects mood-related behaviors. Nevertheless, our analysis of the light-responsive transcriptomic change at ZT8 revealed significant contributions of genes involved in the sensory perception of smell/olfactory receptor activity and G protein-coupled receptor signaling/activity. This parallels previous findings that described depression-like behaviors in rats after olfactory bulb removal [24] and patients with major depression displaying olfactory deficits [25]. However, it remains elusive how the olfactory system and depression are mechanistically related.

Integrated genome-wide association and hypothalamic eQTL studies indicated a link between Per1 and coping behavior in humans [50]. Furthermore, a single nucleotide polymorphism in the human Per1 promoter, as well as animal experimentation, revealed a role of Per1 in psychosocial stress-induced alcohol drinking [51]. These studies are consistent with our finding that Per1 was involved in the regulation of behavioral despair and anxiety, two aspects of depression. Interestingly, the profiles of Per1 and Rev-erbα were advanced in patients in the manic phase compared with those in the depression phase [52]. This correlates with the light-induced phase-shifting properties of Per1 when light was applied at ZT22 and eliciting a phase-advance in activity and gene expression rhythms [9,10,43]. Phase shifts of the circadian clock involve the SCN, which are intact in our vSPer1 mice and still express Per1 (Fig 3D). Therefore, these animals would phase advance normally. If a phase-advance of the cock would be the only reason for the anti-depressant effects of light at ZT22 then deletion of Per1 in neurons of the LHb would not be sufficient to abolish the light mediated effects in the FST. Therefore, it is very likely that the effects of light we observe in the FST are due to Per1 induction in the LHb rather than due to SCN mediated phase advances of the circadian clock.

Taken together, our study provides evidence that the benefit of light on mood-related behaviors involves the clock gene Per1 and that induction of this gene in neurons in the area of the LHb plays an important role.

[END]

[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1009625

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.