Skip to main content
NCBI home page
EMBO Reports logoLink to EMBO Reports
. 2022 Apr 27;23(6):e52316. doi: 10.15252/embr.202052316

External light‐dark cycle shapes gut microbiota through intrinsically photosensitive retinal ganglion cells

Chi‐Chan Lee 1, Feng Liang 1, I‐Chi Lee 1, Tsung‐Hao Lu 1, Yu‐Yau Shan 1, Chih‐Fan Jeng 1, Yan‐Fang Zou 1, Hon‐Tsen Yu 1,2, Shih‐Kuo Chen (Alen) 1,2,✉
PMCID: PMC9171413  PMID: 35476894

Abstract

Gut microbiota are involved in many physiological functions such as metabolism, brain development, and neurodegenerative diseases. Many microbes in the digestive tract do not maintain a constant level of their relative abundance but show daily oscillations under normal conditions. Recent evidence indicates that chronic jetlag, constant darkness, or deletion of the circadian core gene can alter the composition of gut microbiota and dampen the daily oscillation of gut microbes. However, the neuronal circuit responsible for modulating gut microbiota remained unclear. Using genetic mouse models and 16s rRNA metagenomic analysis, we find that light‐dark cycle information transmitted by the intrinsically photosensitive retinal ganglion cells (ipRGCs) is essential for daily oscillations of gut microbes under temporal restricted high‐fat diet conditions. Furthermore, aberrant light exposure such as dim light at night (dLAN) can alter the composition, relative abundance, and daily oscillations of gut microbiota. Together, our results indicate that external light‐dark cycle information can modulate gut microbiota in the direction from the brain to the gut via the sensory system.

Keywords: dim light at night, gut microbiota, ipRGC, melanopsin

Subject Categories: Metabolism; Microbiology, Virology & Host Pathogen Interaction; Neuroscience

In addition to the endogenous circadian clock of the host, the external light‐dark cycle can modulate gut microbiota composition, diversity, and daily oscillation through intrinsically photosensitive retinal ganglion cells.

graphic file with name EMBR-23-e52316-g001.jpg

Introduction

Gut microbiota could influence the central nervous system through a proposed gut‐brain axis to modulate neuronal development, function, and degeneration. It has been shown that germ‐free mice have different hypothalamus metabolites compared with normal mice. In addition, germ‐free or antibiotic‐treated mice have different gene expression profiles in the CNS and lower anxiety levels compared with normal SPF mice (Sudo et al, 2004; Bercik et al, 2011; Diaz Heijtz et al, 2011; Neufeld et al, 2011). The development of the endocrine neuron such as oxytocin neurons is affected by the gut microbe during the postnatal period, which will influence the adult stage social interaction (Buffington et al, 2016). Finally, the gut microbiota will also influence the aggregation of alpha‐synuclein in the gut, which could lead to Parkinson’s disease (Sampson et al, 2016). The composition of gut microbiota and the relative abundance of specific microbes could be influenced by the host’s genetic factors or feeding scheme. Interestingly the relative abundance of many gut microbes displays daily oscillation even under normal condition (Thaiss et al, 2014; Voigt et al, 2014; Zarrinpar et al, 2014; Leone et al, 2015). Although the direct physiological implication of the gut microbe daily oscillation is poorly understood, a recent report showed that arrhythmic gut microbe is associated with patients with type 2 diabetes (Reitmeier et al, 2020). Recent evidence showed that chronic jetlag, constant darkness, and reversal of the light‐dark cycle could disrupt the daily oscillation of gut microbe (Thaiss et al, 2014; Voigt et al, 2014; Liang et al, 2015; Wu et al, 2018). In addition, mice with clock gene knockout also display arrhythmic gut microbiota (Thaiss et al, 2014; Wu et al, 2018). These studies suggest that both external and internal factors could influence the daily oscillation of gut microbe, and their interaction may play an important role in gut homeostasis and microbiota (Mukherji et al, 2013; Thaiss et al, 2014; Voigt et al, 2014; Liang et al, 2015; Wu et al, 2018; Godinho‐Silva et al, 2019; Kuang et al, 2019). However, how external factors influence the gut microbiota is unknown.

One candidate for modulation of daily gut microbe oscillation is the circadian clock system. This system is entrained into the daily light‐dark cycle (LD) via input from the melanopsin‐expressing, intrinsically photosensitive retinal ganglion cells (ipRGCs) to the suprachiasmatic nucleus (SCN) (Berson et al, 2002; Hattar et al, 2002, 2003; Provencio et al, 2002; Guler et al, 2008; Fernandez et al, 2016). In addition, ipRGCs also influence many other non‐image‐forming functions by projecting to brain regions in the thalamus and hypothalamus such as the olivary pretectal nucleus for pupillary light reflex (Lucas et al, 2003; Hattar et al, 2006; Baver et al, 2008; Chen et al, 2011). Genetic elimination of ipRGCs using the Opn4‐DTA or Opn4‐Cre; DTR mouse lines impairs the ability of mice to transmit external light‐dark cycle information for circadian photoentrainment which causes these mice to “free run” under any kind of environmental light‐dark cycle (Hatori et al, 2008; Prigge et al, 2016; Chew et al, 2017). On the other hand, knockout of the photopigment melanopsin (MKO) only produced a light detection phenotype for non‐image‐forming functions under high light intensity (Hattar et al, 2003; Lucas et al, 2003; Schmidt & Kofuji, 2010), which suggests that ipRGCs transmit signals to the brain by combining high luminance signals detected via melanopsin and low luminance signals originating from the canonical rod and cone photoreceptors. Recent evidence showed that ipRGCs can modulate many additional physiological functions such as emotion, hair regeneration, and body temperature (Fan et al, 2018; Fernandez et al, 2018; Rupp et al, 2019). Since ipRGCs provide environmental luminance signals for many non‐image‐forming visual functions, it is likely that they could also influence gut microbiota. Using 16s rRNA analysis and various kinds of ipRGC‐related mutant mice, here we show that light‐dark cycle information could drive the daily oscillation of gut microbes through the ipRGCs‐sympathetic nerve circuit. Furthermore, aberrant light‐dark cycles such as light exposure during the nighttime could cause dysbiosis and dampen the daily oscillation of gut microbes.

Results

Light/dark cycle information is important to drive daily oscillations of gut microbes

To determine how the light‐dark cycle may directly influence the composition of gut microbiota, we designed a specific dim light at night (dLAN) conditions to provide light exposure while minimizing the disruption of the central circadian clock. To control feeding time, we modified the dLAN condition by incorporating a temporal restricted feeding cycle with a high‐fat diet to eliminate daytime feeding, a behavioral phenotype that could disrupt the circadian clock (Fonken et al, 2010). Mice housed under our modified dLAN conditions could only eat during the nighttime and have similar circadian behaviors and gene expression patterns compared with normal light‐dark cycle (Fig EV1). The average nighttime food intake is similar between all experimental setups (Fig EV2). Together, our results indicated that the central clock remains entrained and shows daily oscillation similar to the normal LD cycle when mice were exposed to 25 lux of light during the subjective nighttime under the temporal restricted high‐fat diet condition. Therefore, our modified dLAN conditions allowed us to test the influence of additional light exposure during the nighttime while keeping mice entrained to 24 h daily light‐dark cycle.

Figure EV1. Mice housed in dLAN condition display similar activity and clock gene expression patterns to LD condition.

Figure EV1

Open in a new tab
  1. Schematic representation of the experimental design for mice exposed to dim light at night (dLAN) or normal light/dark (LD). LD mice were housed under a 12h:12h LD cycle with 0 lux during the dark phase; dLAN mice were housed under 25 lux during the dark phase. Food availability was restricted to the dark phase for both conditions.
  2. Temporal activity profile of mice, plotted by the moving distance in the home cages per hour, under dim light at night (dLAN, red) or normal light/dark (LD, blue) conditions. n = 5 for each group.
  3. qPCR analysis of expression levels of the circadian clock genes Bmal1 and Per2 in the suprachiasmatic nucleus (SCN) and liver from control mice housed under conditions of dLAN (gray) and LD (black).
  4. Representative images of Per2 immuno‐positive cells in the SCN at ZT4, ZT10, ZT16, and ZT22 from control mice housed under LD (upper) and dLAN (lower).
  5. Quantification of Per2 immuno‐positive cell in the SCN. There is no significant difference in the Per2 positive cell number between mice housed under LD and dLAN conditions.
  6. Daily food intake measured in Kcal is similar between all experimental groups. n = 4–5 mice per group.

Data information: No significant difference is reported using 2‐way ANOVA. n = 4–5 for qPCR, n = 3 for IHC. Scale bar is 100 μm for (D). Data are presented as mean ± SEM.

Figure EV2. Actogram of Opn4DTA/DTA mice.

Figure EV2

Open in a new tab

Wheel‐running activity of Opn4DTA/DTA mice immediately after final fecal sample collection.

  1. Opn4DTA/DTA mice, which activity onsets are at least 3 h in advance or delay to the light off, housed under LD or dLAN condition.
  2. Opn4DTA/DTA mice housed under dLAN condition with activity onsets less than 3 h in advance or delay to the light off.

Data information: Yellow circles indicate predicted onset for fecal sample‐collecting days. Red triangles indicate fecal sample‐collecting times.

To determine the impact of the light‐dark cycle on gut microbes, we first compared the gut microbiota under normal LD conditions, under constant darkness (DD) with reduced light exposure, or under dLAN conditions with extra light exposure. After 2 weeks of normal LD cycle, constant darkness, or dLAN conditions, fecal samples were collected from WT mice at six time points throughout the day. We perform 16S rRNA‐based next‐generation sequencing and use the operational taxonomic unit (OTU) based JTK_CYCLE (Hughes et al, 2010) to analyze the daily gut microbial oscillation in mice housed under different light‐dark cycles. For WT mice, many microbes display daily oscillations under LD conditions (Fig 1A and D). To compare the amount of oscillating gut microbe between different light‐dark cycles, we added all microbes with both P‐value and q‐value (false‐positive rate) smaller than 0.05 in JTK _CYCLE analysis to calculate the total percentage of gut microbe displayed daily oscillation (Fig 1C). Surprisingly, although the locomotor activity of mice and most circadian clock genes were entrained to 24 h cycle under dLAN conditions (Fig EV1), percentages of oscillating gut microbes were highly reduced in mice housed under dLAN conditions (Fig 1A, C and D). We also confirm that the daily oscillation of gut microbes is significantly reduced in mice housed under the DD condition, even with free‐running endogenous clock (Fig 1B, C and D). The temporal restriction feeding with a high‐fat diet is not sufficient to drive the full daily oscillation of gut microbe for mice housed under dLAN and constant darkness conditions. Together, our results suggest that the environmental light‐dark cycle is one of the major factors driving the daily oscillation of gut microbe.

Figure 1. Light‐dark cycle is important to drive the daily oscillation of gut microbes.

Figure 1

Open in a new tab
  • A, B
    Oscillating OTUs detected by the JTK_CYCLE (adjust P < 0.05, Benjamini–Hochberg q‐value < 0.05) in WT mice housed under normal light‐dark cycle (A, LD), dim light at night (A, dLAN), and constant darkness (B) conditions. Dash lines indicate the Benjamini–Hochberg q‐value < 0.05. Graphs without dash line indicate that the q‐values for all OTU were higher than 0.05. Both adjusted P‐value and q‐value were generated by JTK cycle.
  • C
    Summation of relative abundance from all oscillating OTUs from A and B.
  • D
    Relative abundance heat map of oscillating OTUs detected by the JTK_CYCLE. Each row represents 1 OTU across the day, and the graph is double plotted. Gray color indicates OTU not found. n = 4–5 for each time point.

Data information: n = 4–5 mice. *P < 0.05 with the 1‐way ANOVA Tukey post hoc test. All bar graphs are presented as mean ± SEM.

Daily oscillation of gut microbes is driven by the light‐dark cycle through ipRGCs

It has been shown that ipRGC is the primary conduit to transmit light information for many non‐image‐forming visual functions such as circadian photoentrainment and pupillary light reflex in mammals. Although the free‐running central clock could not fully drive the daily oscillation of gut microbes, ipRGCs and/or melanopsin signaling may be involved in transmitting light‐dark cycle information to shape the gut microbiota. To test this hypothesis, we analyzed the daily oscillation of gut microbe using JTK_CYCLE from control, MKO, and ipRGC‐eliminated (Opn4DTA/DTA) mice housed under normal LD cycle or dLAN conditions (Fig 2A–C). First, we compared the gut microbe oscillation between control and MKO mice. Interestingly, the pattern of daily oscillations in MKO differed from that in control mice under LD conditions, while the total oscillation percentage in MKO is still similar to control (Fig 2D and E). In control mice, peaks of oscillating OTUs were spread relatively evenly throughout the day (Fig 2F), whereas in MKO mice, peak times for many oscillating OTUs occurred between ZT8 and ZT12 (Fig 2G). These de novo microbe oscillations in MKO mice suggested that rod and/or cone signals might partially compensate for the loss of melanopsin. Next, we test whether ipRGC is the sole conduit to transmit melanopsin and rod/cone signal for gut microbe daily oscillation by comparing the rhythmicity between control and Opn4DTA/DTA mice. Because the running activities of ipRGC‐eliminated (Opn4DTA/DTA) mice were not tied to the environmental light‐dark cycle, we first collected fecal samples according to the light‐dark cycle and immediately recorded their wheel‐running activity afterward (Fig EV2). Therefore, JTK_CYCLE analysis for Opn4DTA/DTA mice was performed by arranging data points to match either circadian time (CT) predicted from the wheel‐running activity (Fig EV2) or zeitgeber time (ZT) according to the light‐dark cycle. Strikingly, in Opn4DTA/DTA mice, daily gut microbial oscillations were greatly attenuated (< 3%) under both LD and dLAN conditions when analyzed with either CT or ZT time (Fig 2H). This result indicates that ipRGC is essential to provide normal light‐dark cycle information from melanopsin and rod/cone to drive the gut microbe daily oscillation.

Figure 2. Light‐dark cycle information transmitted by ipRGC is required for gut microbe oscillation.

Figure 2

Open in a new tab
  • A–C
    Oscillating OTUs detected by the JTK_CYCLE (adjust P < 0.05, Benjamini–Hochberg q‐value < 0.05) in control (A), Opn4DTA/DTA (B), and MKO (C) mice housed under LD and dLAN conditions. Dash lines indicate the Benjamini–Hochberg q‐value < 0.05. Graphs without dash line indicate that the q‐values for all OTU were higher than 0.05. Both adjusted P‐value and q‐value were generated by JTK cycle.
  • D
    Summation of relative abundance from all oscillating OTUs in control and MKO mice.
  • E
    Venn diagram of oscillating OTUs in control and MKO mice.
  • F, G
    Relative abundance heat map of oscillating OTUs detected by the JTK_CYCLE in LD or dLAN conditions from control (F) or MKO (G) mice. Each row represents 1 OTU across the day, and the graph is double plotted.
  • H
    Summation of relative abundance from all oscillating OTUs in Opn4DTA/DTA mice according to the circadian time (CT) of the mice, or the light/dark and food cycle (ZT).

Data information: n = 4–6 mice. *P < 0.05, **P < 0.01 using the 2‐way ANOVA Bonferroni post hoc test. All bar graphs are presented as mean ± SEM.

Light signals transmitted by ipRGC influenced gut microbiota composition

In addition to modulating the daily oscillation of particular gut microbes, we next asked whether ipRGCs could modulate the overall composition of gut microbiota. We used principal coordinate analysis (PCoA) to compare the gut microbiota composition between control, MKO, and Opn4DTA/DTA mice. Analysis of mean and variation, using ANOSIM (Fierer et al, 2010) and Adonis (PERMANOVA) (Anderson, 2008), respectively, showed that gut microbiota composition is significantly different between control and MKO mice (Fig 3A), and also significantly different between control and Opn4DTA/DTA mice (Fig 3B). The alpha diversity (richness) was significantly lower for MKO and Opn4DTA/DTA mice compared with control mice (Fig 3C), while beta diversity (variation between samples) from MKO and Opn4DTA/DTA mice was both significantly higher than control mice (Fig 3D). These results indicate that melanopsin signaling and ipRGCs are important factors in maintaining normal gut microbiota composition. Next, we compared the gut microbiota from control, MKO, and Opn4DTA/DTA mice housed under LD or dLAN condition. The PCoA analysis showed a clear separation of gut microbiota from control mice housed under LD versus dLAN conditions (Fig 3E), indicating nighttime light exposure could modulate overall gut microbiota in mice. In contrast, neither MKO nor DTA animals showed separation of gut microbiota when mice were housed in LD or dLAN condition (Fig 3F and G). Moreover, neither alpha nor beta diversity of microbiota differed between LD and dLAN conditions in MKO and Opn4DTA/DTA mice, though they were significantly different in controls (Fig 3C and D). Since Opn4DTA/DTA free run under any light‐dark cycle, the activity time of mice was sometimes synchronized to the feeding schedule and then desynchronized to the feeding schedule during the course of the experiment. Interestingly, PCoA analysis resulted in a strong separation of gut microbiota between Opn4DTA/DTA mice with their circadian clock synchronized or not synchronized to the feeding schedule under any light‐dark cycle (Fig 3H). These results showed that dyssynchronization between the feeding schedule and the endogenous circadian clock could modulate the gut microbiota in ipRGC‐eliminated mice. However, aberrant light‐dark cycle‐induced dysbiosis is blocked in both melanopsin knockout and ipRGC‐eliminated mice.

Figure 3. dLAN altered composition of gut microbiota.

Figure 3

Open in a new tab
  • A
    Principal coordinate analysis (PCoA) of gut microbiota using weighted UniFrac matrix from control and MKO mice housed under LD conditions.
  • B
    PCoA of gut microbiota from control and Opn4DTA/DTA mice housed under LD conditions. The gut microbe compositions are significantly different between control and ipRGC‐manipulated mutant mice with both ANOSIM and Adonis analysis.
  • C, D
    Violin plots of the alpha diversity using Shannon index (C) and the beta diversity generated from the weighted UniFrac distance matrix (D) for gut microbiota from control, MKO, and Opn4DTA/DTA mice under LD and dLAN condition. There is no significant difference under dLAN condition.
  • E–G
    PCoA of gut microbiota from control, MKO, and Opn4DTA/DTA (with similar phase‐angle difference to the light‐dark cycle) mice housed under LD or dLAN conditions. Both ANOSIM and Adonis had significant differences for control mice housed under LD or dLAN conditions (E), but not for MKO (F) and Opn4DTA/DTA mice (G).
  • H
    PCoA of gut microbiota from Opn4DTA/DTA mice either synchronized or desynchronized to the light/dark and feeding cycle.

Data information: n = 4–5 mice. For violin plots (C and D), dot lines are 25 and 75% quartile lines. The dashed line represents the mean between groups; *P < 0.01 using the 2‐way ANOVA Bonferroni post hoc test; #P < 0.05 between same genotype using the student t‐test; n.s. not significant. ANOSIM and ADONIS tests and their P‐values were indicated in each PCoA graph.

Next, we compared the relative abundance of each gut microbe under different light‐dark cycles. The apparent gut microbe composition at the phylum level did not differ significantly among groups (Fig 4A). However, linear discriminant analysis at multiple phylogenetic levels showed that thirteen classifications, including the genus Rikenella, Odoribacter, and Lactobacillus, were significantly different in control mice housed under dLAN versus LD conditions (Fig 4B), whereas in MKO, only one classification had significant differences between dLAN and LD conditions (Fig 4C). Strikingly, there were 23 classifications with significant differences between dLAN and LD conditions in Opn4DTA/DTA mice. However, none of them were included in the 13 differential classifications in control mice housed under LD and dLAN conditions (Fig 4D). Since Opn4DTA/DTA mice were free running under our 12h:12h light‐dark cycle, the difference in microbiota pattern from Opn4DTA/DTA mice may, again, be caused by dyssynchronization between the feeding schedule and the endogenous circadian clock. Together, our results suggested that light information through activation of melanopsin and/or transmitted by ipRGCs could influence the composition and diversity of gut microbiota.

Figure 4. Enrichment of specific microbes from mice housed under dLAN and LD conditions.

Figure 4

Open in a new tab
  1. Mean relative abundance of phylum from control (Con)‐, melanopsin knockout (MKO)‐, and ipRGC‐eliminated (Opn4DTA/DTA) mice housed under conditions of normal light/dark (LD) or dim light at night (dLAN) (n = 4–5 from each group).
  2. The cladogram shows the difference in relative abundance of microbes from control mice housed under conditions of LD and dLAN. Microbes in red have significantly higher relative abundance in LD conditions. Microbes in blue have significantly higher relative abundance in dLAN conditions.
  3. The cladogram shows the difference in relative abundance of microbes from MKO mice under LD and dLAN conditions.
  4. The cladogram shows the difference in relative abundance of microbes from Opn4DTA/DTA mice under LD and dLAN conditions. For RF39, it is higher in the dLAN condition from control mice but higher in the LD condition from Opn4DTA/DTA mice.

ipRGC drive light‐evoked, daily oscillations of gut microbes via sympathetic nerves

It has been shown that light can influence peripheral organs via the retinohypothalamic tract, independent of the circadian clock through autonomic circuits (Kiessling et al, 2014). Therefore, to determine the neural circuit responsible for the light‐evoked modulation of gut microbiota, we tested whether the sympathetic nerve was involved in dLAN‐induced dysbiosis. To eliminate sympathetic nerve function, WT mice were treated with 6‐Hydroxydopamine (6‐OHDA) (Thoenen & Tranzer, 1968). Body weight of mice decreased during the first week of 6‐OHDA injection (Fig 5A), consistent with the successful elimination of sympathetic nerve function (Joseph et al, 1979). Furthermore, we did not observe any change in the number of dopaminergic amacrine cells with TH‐immunostaining, suggesting that 6‐OHDA may not penetrate across the blood‐retinal barrier (Fig EV3). After body weight stabilized, mice were housed under either dLAN or LD conditions for 2 weeks. Although the gut microbiota oscillation in the PBS control group is much higher than the LD condition due to daily injection throughout the experimental period, the daily oscillation showed a trend of decline under dLAN compared with LD conditions (Fig 5B and D). The gut microbiota composition was also clustered into 2 groups (ANOSIM P = 0.054) (Fig 5E). Surprisingly, sympathetic nerve ablation significantly dampened the daily oscillations of gut microbes under both LD and dLAN conditions (Fig 5C and D), similar to our observations in ipRGC‐ablated animals. These data suggest that light input driving daily oscillations occurs via the sympathetic nervous system. Interestingly, sympathetic nerve ablation failed to abolish the effects of dLAN‐induced dysbiosis, as dLAN still altered the composition of gut microbiota under this manipulation (Fig 5F), suggesting that these influences occur via a sympathetic nerve‐independent circuit.

Figure 5. Elimination of sympathetic nerve by 6‐OHDA inhibits daily oscillation of gut microbes.

Figure 5

Open in a new tab
  • A
    After 1 week of initial injection (baseline), about 10% of transient body weight loss was observed in the 6‐OHDA group (open circle) but not in the PBS injection group (closed circle) compared with pretreat time point. The body weight recovered after 1 week, although 100mg/kg of 6‐OHDA was injected every 2 days throughout the experiment. LD = light‐dark cycle, dLAN = dim light at night.
  • B, C
    OTU‐based JTK_CYCLE analysis from PBS (B) and 6‐OHDA (C) treated mice housed under LD and dLAN conditions. Dash lines indicate the Benjamini–Hochberg q‐value < 0.05. Both adjusted P‐value and q‐value were generated by JTK cycle.
  • D
    Gut microbes displayed significant dampened daily oscillations in 6‐OHDA‐injected mice under both LD or dLAN conditions.
  • E, F
    PCoA of weighted UniFrac matrix from PBS (E) and 6‐OHDA (F) injected mice housed under LD or dLAN conditions. The P‐value for PBS injection is 0.054, and the P‐value for 6‐OHDA injection is < 0.01 using ANOSIM.

Data information: n = 4–5 for each condition. For A and D, data were presented as mean ± SEM. *P < 0.01 using the 2‐way ANOVA Bonferroni post hoc test.

Figure EV3. Immunofluorescence staining of whole‐mount retina for tyrosine hydroxylase after 2 weeks of 6‐OHDA injection.

Figure EV3

Open in a new tab
  1. Representative images of tyrosine hydroxylase immune‐positive dopaminergic amacrine cells in the retina from mice received 2 weeks of 6‐OHDA injection (A) or PBS (B).
  2. Density of dopaminergic amacrine cell in 6‐OHDA‐ and PBS‐injected mice. Density for each animal is calculated by averaging the number of DAC from 4 quadrants of the retina.

Data information: Scale bar = 100 µm. No significance is reported using t‐test. n = 3.

The immune response of intestine cells has been shown to involve in the interaction between host and gut microbes. Therefore, we would like to verify the local molecular marker that participates in light‐dependent gut microbe modulation. To see whether dLAN could modulate the gene expression in the intestine, we compared transcriptome data from mice housed under LD or dLAN conditions. We found that 2 weeks of dLAN could influence the gene expression profile in the intestine. Amount 12,901 expressing genes, 294 genes are specifically expressed under the dLAN condition and 251 genes are specifically expressed under the control condition (Fig 6A). For differential expression genes, we found that MMP10 and Iglv3 are upregulated and Nr1d1 is downregulated under the dLAN condition (Fig 6B). Since the MMP10 gene is involved in immune response and is expressed in the intestine epithelium, we collected intestine samples from mice housed under control and dLAN condition similar to previous experimental conditions. Quantitative PCR confirmed that MMP10 is upregulated under the dLAN condition in the intestine. Furthermore, the upregulation can be blocked by the elimination of ipRGC or sympathetic nerve using 6‐OHDA (Fig 6C). Together, these results indicate that environmental light signals transmitted by the ipRGC‐sympathetic nerve circuit could modulate the immune response of the digestive tract and is important to drive the daily oscillation of gut microbes. In addition ipRGC could influence the composition of gut microbes through the sympathetic nerve‐independent unidentified pathway (Fig 6D).

Figure 6. Gene expression profile of intestine under dLAN or LD condition.

Figure 6

Open in a new tab
  1. Transcriptome analysis of intestine from mice housed under dLAN or LD conditions. Venn diagram showed that 294 genes are specifically expressed under dLAN condition while 251 genes are specifically expressed under LD condition.
  2. Nr1d1 gene is significantly downregulated while MMP10 and Iglv3 are upregulated in dLAN condition compared with control. P‐values were generated by Poisson distribution estimation model with Benjamini–Hochberg false‐discovery rate estimation.
  3. Quantitative PCR showed that MMP10 gene expression is upregulated under the dLAN condition. Sympathetic nerve‐eliminated (6‐OHDA) or ipRGC‐eliminated mice (DTA) do not show a significant difference of MMP10 gene expression between dLAN and LD condition. Data were presented as mean ± SEM.
  4. Light‐dark cycle information transmitted by ipRGC through sympathetic dependent and independent pathways to influence the daily oscillation of gut microbes and the composition of gut microbiota, respectively.

Data information: n = 5–8 for qPCR and n = 3 for transcriptome. ** indicates P < 0.01 with the 2‐way ANOVA Bonferroni post hoc test.

Since a high‐fat diet could influence the gut microbiota, we would like to test whether dLAN‐induced dysbiosis is specific to our experimental setup. We perform control and dLAN treatments with ad libitum normal chew. After 2 weeks of dLAN, the total percentage of gut microbes that display daily oscillation in the dLAN group is significantly lower than in the control group (Fig 7A–C). The PCoA analysis also indicates that control and dLAN mice have distinct gut microbiota (Fig 7D). Our results also confirm with previous literature that animals fed with a high‐fat diet display a lower percentage of gut microbe daily oscillation than normal chow food. While the remaining oscillating gut microbes indicate that the circadian clock is also involved in the gut microbe daily oscillation, which agrees with previous literature. Together, our results suggest that the light‐dark cycle could influence gut microbiota under both normal and high‐fat diet conditions.

Figure 7. dLAN modulates gut microbiota under regular feeding condition.

Figure 7

Open in a new tab
  • A, B
    OTU‐based JTK_CYCLE analysis from WT mice with ad libitum feeding with normal chaw food under LD (A) or dLAN (B) conditions. Dash lines indicate the Benjamini–Hochberg q‐value < 0.05. Both adjusted P‐value and q‐value were generated by JTK cycle.
  • C
    Summation of relative abundance from all oscillating OTUs in WT mice with normal chaw food under ad libitum housed under LD or dLAN conditions.
  • D
    Principal coordinate analysis (PCoA) of gut microbiota using weighted UniFrac matrix from WT mice housed under LD and dLAN conditions with normal chaw under ad libitum.

Data information: n = 6 mice. **** indicates P‐value < 0.001 with t‐test. Bar graphs were presented with mean ± SEM. P‐values in PCoA graph were generated by ANOSIM and ADONIS.

Discussion

Our data demonstrate that light input from ipRGCs is a critical regulator of the gut microbiome. We find that gut microbe composition and oscillation are significantly altered in the absence of ipRGC signaling through a sympathetic circuit. Moreover, nighttime light exposure also alters gut microbe composition and diversity through ipRGCs and the non‐sympathetic circuit. Together, our results indicate that external light information is an important driving force to shape the gut microbiota in mice through melanopsin and ipRGC circuitry. Since dysbiosis is associated with many metabolic diseases, our results suggest that the light‐dark cycle might modulate metabolic status through ipRGCs and gut microbiota interaction.

It has been shown that clock gene(s) are essential for gut microbe daily oscillation since such oscillations are highly attenuated in circadian core gene Per1/2 knockout mice (Thaiss et al, 2014). Here we showed that many gut microbes display daily oscillation under normal light‐dark cycle similar to previous reports (Thaiss et al, 2014; Zarrinpar et al, 2014; Zhang et al, 2017). However, oscillations of gut microbes were substantially dampened from mice housed under dLAN, which showed daily circadian rhythm behaviorally and genetically. Furthermore, gut microbe oscillations were highly dampened in the ipRGC‐eliminated mice and WT mice housed under constant darkness, both have a free‐running endogenous clock. Together, our results suggested that the light‐dark cycle is also a strong force to drive the daily oscillation of gut microbes in addition to the circadian clock of the host. Here we hypothesize that distinct species of gut microbe were oscillating in different individuals of mice without ipRGC sending proper light‐dark cycle information. Therefore, the discrete oscillating microbes within different mice may be averaged out in JTK cycle analysis and showed dampened daily oscillation as groups. This hypothesis is supported by our result that both ipRGC‐eliminated and WT mice housed under dLAN conditions have higher beta diversity, which indicates a larger variance between samples, than WT mice housed under normal LD conditions. This is similar to the result that peripheral clocks will free run with their own periods in vitro without input from the SCN (Tahara et al, 2012). Alternatively, the information on external light‐dark conditions could potentially modulate the peripheral clock, which is reported recently (Husse et al, 2014; Izumo et al, 2014), to influence the daily oscillation of gut microbe. Consequently, a mismatch of the host circadian clock and external light‐dark cycle (e.g., dLAN) disrupted gut microbes, apparently through untimely activation of sympathetic nerves. Another potential hypothesis is that proper entrainment could maintain or magnify the oscillating amplitude of the host circadian clock to provide sufficient driving force for gut microbes. Therefore, the circadian central clock of the host and the light‐dark cycle signal could have an additive effect to modulate the daily oscillation of gut microbes through single or multiple neuronal circuitry including sympathetic nerves. Although we could not distinguish whether oscillation is disrupted within individual or groups from current results, our conclusion that ipRGC could transmit light‐dark cycle information to shape the coherent daily oscillation of gut microbes remains unaffected.

Interestingly, there was only a small decrease in the daily oscillation of gut microbes in MKO mice (16%, compared with 20% in WT) exposed to normal LD conditions, suggesting that an extrinsic signal from photoreceptor rods and cones through ipRGCs could provide sufficient information to drive daily oscillations of gut microbes. However, dLAN could not significantly alter the overall composition of gut microbiota in MKO mice. This result indicates that melanopsin photo‐detection was required to modulate the part composition of gut microbiota. This phenomena of melanopsin discrepancy are similar to the circadian photoentrainment, in which melanopsin is not necessary under certain conditions. Activation of ipRGCs with rods and cones in melanopsin knockout mice is sufficient to entrain animals in the external daily light‐dark cycle (Altimus et al, 2010; Calligaro et al, 2019), which is similar to what drives gut microbe oscillation. However, full activation of ipRGC with melanopsin is required for period lengthening under constant light (Panda et al, 2002; Ruby et al, 2002; Hattar et al, 2003), which is similar to systematic modulation of gut microbiota composition. In addition, we observed the de novo oscillation of gut microbiota in the MKO mice and their relative resilience to dLAN treatment. These results indicate that endogenous clock with minimum light‐dark cycle signaling and full ipRGC signaling with melanopsin photo‐detection system might provide different signaling pathways to shape the oscillation of gut microbes. Therefore, our data suggest that information transmitted by ipRGCs to the brain may utilize multiple layers of computation circuits to modulate gut microbiota. Recent studies showed that ipRGC could release both glutamate and GABA (Sonoda et al, 2020), or innervate all major types of neurons in the SCN (Fernandez et al, 2016). Furthermore, ipRGC input to the SCN and IGL could be involved in the gut microbiota modulation through food entrainment circuitry (Fernandez et al, 2020). These complicated circuits may help explain why we observed 2 distinct phenotypes of gut microbiota under dLAN condition.

In addition to daily oscillation modulation, we found that the relative abundance of two specific genera Rikenella and Odoribacter were downregulated in dLAN condition, while order RF39, family S24‐7, genus Paraprevotella, and Lactobacillus were upregulated. It has been shown that Rikenella and Odoribacter in the gut are associated with non‐infective inflammation (Zackular et al, 2013; Leiva‐Gea et al, 2018; He et al, 2019; Miranda‐Ribera et al, 2019). RF39 and S24‐7 have been shown to associate with diabetes and obesity (Krych et al, 2015; Menni et al, 2019; Robinson et al, 2019). Although our transcriptome data did not report direct upregulation of inflammation factors, pathway analysis showed that the immune signaling was different between animals housed under control LD and dLAN condition. Specifically, MMP10 that has been shown to modulate inflammation (McMahan et al, 2016) is upregulated in mice housed under dLAN. Therefore, dLAN might be a risk factor for inflammation and metabolic disorders. Indeed, a recent study showed that arrhythmic gut microbiota is linked to patients with type 2 diabetes (Reitmeier et al, 2020). Since we only house mice under dLAN for 2 weeks, further study could focus on the relationship between chronic dLAN and metabolic or inflammation‐related disease. In brief, the brain to gut neuronal circuitry including ipRGC and the sympathetic nervous system could be involved in shaping the gut microbiota through modulating host‐microbe interaction molecular pathways in the digestive tract.

In several studies, high‐fat diets or environmental stress (e.g., restricted sleep) reduced the richness (alpha diversity) of gut microbiota (Thaiss et al, 2014; Zarrinpar et al, 2014; Leone et al, 2015; Benedict et al, 2016; Zhang et al, 2017). The high‐fat diet has been shown to dampen the gut microbe daily oscillation (Zarrinpar et al, 2014). Our gut microbe daily oscillation is indeed slightly lower than previous reports, confirming that a high‐fat diet could reduce the oscillation amplitude of the gut microbe. Herein, we showed that dLAN further reduced the richness of the gut microbiota under a high‐fat diet, which indicates dLAN is an additional environmental stress for gut microbiota. Finally, the dLAN‐induced dysbiosis suggests that activation of ipRGC during the nighttime could disrupt gut microbiota. On the other hand, the gut microbiota composition, alpha, and beta diversity are different between control‐, MKO‐, and ipRGC‐eliminated mice, suggesting that the presence of melanopsin signaling during the day could also regulate the gut microbiota. Overall, our results together suggest that appropriate light‐dark cycle information transmitted by ipRGC is important to maintain a normal gut microbiota.

Materials and Methods

Mice

Care of mice and protocols used in this study were reviewed and approved by the Institutional Animal Care and Use Committee at National Taiwan University. Male littermate age‐matched controls (Opn4Cre/+ ), male melanopsin knockout mice (Opn4Cre/Cre ), and male ipRGC‐eliminated mice (Opn4DTA/DTA ), 8 weeks of age, were used. Mice were maintained on a C57BL6/J background.

Experimental design

Littermates were housed together under a 12h:12h light‐dark (LD) cycle and given a standard diet and water ad libitum prior to the experiment. At 8 weeks of age, mice were housed individually and randomly assigned to one of two experimental conditions: a normal light/dark cycle (LD) or dim light at night (dLAN). The LD group was exposed to a 12‐h light phase, 800 lux, and a 12‐h dark phase, 0 lux. The dLAN group was exposed to the same 12‐h light phase but a 25‐lux, 12‐h dim light phase. The light intensity for dLAN during the subjective night is 25 lux (around 2.7*1012 photon/s/cm2 at 400–500 nm wavelength range), which is above the threshold for all characterized photoreceptors in the retina including rods, cones, and ipRGCs. Mice were fed high‐fat diets (D12492, 60% fat; Research Diets, Inc., US) during the dark or dim light phase. Mice were either moved to a new cage or feed was removed during the light to limit feeding to the dark or dim light phase. Mice that hid food under bedding were eliminated from the experiment. Daily activity was recorded for 24 h using an infrared camera linked with ANY‐maze (Stoelting, Co., US). For the normal chow diet experiment, the light‐dark cycle is following the same schedule described above with the ad libitum normal chow diet (5010, LabDiet, Inc., US). After 2 weeks in experimental light conditions, fecal pellets were collected over 48 h by combining time courses from different mice and using a set period of 24 h (ZT0, ZT4, ZT8, ZT12, ZT16, ZT20) under LD and dLAN conditions for JTK analysis. Fecal pellets were flash frozen (−80°C) for DNA extraction. For experiments that involved 6‐OHDA treatments, 6‐OHDA(100 mg/kg, Tocris, UK) was given every 2 days, starting 1 weeks prior to the experiments and continuing throughout the experiment. For constant darkness gut microbe oscillation analysis, mice were housed in constant darkness for 2 weeks with temporal restriction feeding between ZT 12 to ZT 24, an additional group of constant darkness with ad libitum feeding was also carried out simultaneously. Fecal samples were collected at ZT0, ZT4, ZT8, ZT12, ZT16, and ZT20 across a 48‐h period for LD and dLAN group. For constant darkness and Opn4DTA/DTA groups, after sample collection, mice were transferred to a wheel‐running cage and the onset of the individual mouse was calculated using ClockLab (Actimetrics, US). The specific collection time was matching to the calculated circadian time for analysis.

Metagenomic library preparation and 16S sequencing

The DNA was extracted from ~25 mg of feces, using a QIAamp DNA Mini Kit (Qiagen, Germany) according to the manufacturer’s instructions and 40 ng of DNA was used in PCR amplification and 16S rRNA gene sequencing. Amplicons spanning the variable region 4/5 (V4/5) of the 16S rRNA gene were generated using the following primers:

5′‐ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGAYTGGGYDTAAAGNG‐3′ and 5′‐GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGCCGTCAATTYYTTTRAGTTT‐3′. The PCR products were cleaned and barcoded using Nextera® Index Kit (Illumina Inc, US). Final products were purified, concentrated and DNA quality verified. The library was sequenced on an Illumina MiSeq platform, with 250 bp paired‐end sequencing.

Microbiota sequence analysis

All analyses is following the procedure of Thaiss et al (2014) and Zarrinpar et al (2014). Reads were classified with QIIME (Quantitative Insight into Microbial Ecology) 1.90. In brief, operational taxonomic units (OTU) were picked using close reference methods at 97% sequence similarity against the Greengenes Database (The Greengenes Database Consortium, Version 13.8), taxonomies were assigned using the uclust consensus taxonomy assigner, and an OTU table was created. Samples with < 2,000 reads were discarded and removed from further analysis. The average number of reads for the samples was 38,361. For beta diversity analysis, unweighted and weighted UniFrac distances were calculated, and principal coordinate analysis plots were generated. Dissimilarity matrixes were also used to compare in‐group differences between microbe samples. To determine the diurnal fluctuations of each OTU, the percentage of reads was calculated for each sample and averaged for each time point. OTU data were analyzed using the JTK_CYCLE program. The significances of cyclic OTUs were achieved by both permutation‐based adjusted P‐values and Benjamini–Hochberg q‐values were < 0.05. Linear discriminant analysis effect size (LEfSe) was used to detect significant changes in the relative abundance of microbial taxa between groups via online Galaxy Browser. Significant thresholds were set as default settings: alpha < 0.05 for a factorial Kruskal–Wallis test among classes and LDA > 2.0 for logarithmic LDA score.

Quantification of circadian gene expression

Livers were homogenized using a tissue homogenizer (Minilys lyser, Bertin Corp., France) and micro‐dissected SCN tissues were lysed in lysis buffer by repeated aspiration through a syringe. Total RNAs were then extracted using Quick‐RNA Miniprep (Zymo, R1054, US) according to the manufacturer’s instructions. cDNAs were then synthesized from total RNA extracts using the iScript cDNA Synthesis Kit (Bio‐Rad, US), and cDNA products were stored at 4°C before use. Thereafter, qPCR was performed using iQ SYBR Green Supermix QPCR kit (Bio‐Rad) and detected in a CFX96 thermal cycler and detection system (Bio‐Rad). Primers for circadian gene detection are listed below. Circadian gene mRNA expressions were normalized by 18S ribosomal RNA (internal control).

Primer list for QPCR

Gene Primer pair Forward primer 5′–3′ Tm(°C)
Per2 Per2‐F ccatccaca agaagatcc tac 59.4
Per2‐r gctccacgggttgatgaagc
Bmal1 Bmal1‐F cctaattctcagggcagc agat 59.4
Bmal1‐R tccagtcttggcatcaat gagt
18S rRNA 18S rRNA‐F ttgttggttttcggaactgaggc 59–63
18S rRNA‐R ggcaaatgctttcgctctggtc
MMP10 MMP10‐F TTCAATCCCTGTATGGAGCCG 60
MMP10‐R tcaggctcgggattccaatg
Open in a new tab

Immunostaining

For immunohistochemistry staining, the mice were perfused with PBS for 2 min and 4% paraformaldehyde (PFA) for 15 min, brains or retina were removed and post‐fixed in 4% PFA for 2 h. After fixation, mouse brains were sectioned coronally with 80 µm thickness using a vibratome. Brian slices were first incubated with 3% H2O2 in PBS for 10 min. Next, brain slices or whole‐mount retina were blocked in 5% goat serum, 1% BSA, and 0.2% Triton‐100 in PBS (0.1 M) for 2 h at room temperature. Subsequently, samples were incubated overnight at 4°C with a blocking solution containing primary antibody (rabbit anti‐Per2, 1:1,000, Alpha Diagnostic, PER21‐A; mouse anti‐TH 1:1,000, ImmunoStar #22941). After washing in PBS (0.1 M), the brain was incubated with VECTASTAIN ABC KITS (PEROXIDASE, Vector Lab) following manufactory instructions. The DAB staining was performed using DAB Peroxidase (HRP) Substrate Kit (Vector Lab) following manufactory instructions. Whole‐mount retina was stained with goat anti‐mouse IGg1 secondary antibody conjugated with Alexa‐488. Samples were mounted with VECTASHIELD® (Vector Lab) for imaging. Bright‐field images were collected using a Zeiss Z1 microscope with a 20× objective. Fluorescence images were collected using Leica SP8 confocal microscope with a 20× objective.

Transcriptome

Mice aged 8 weeks were randomly separated into two groups and housed individually under different light conditions: normal LD cycle (light 800 lux; dark 0 lux) and group in dim light at night (light 800 lux; dim light 40lux) for 2 weeks. All mice were housed with ad libitum filtered water and restricted high‐fat diet (HFD) (D12492, 60% kcal of fat) at night time. Mice were transferred to daytime cage and nighttime cage to ensure the food residue at night would not be taken in daytime. We measured the body weight of each mouse and food intake every week. Mice were anesthetized with avertin (2, 2, 2‐Tribromoethanol) at ZT 14. Small intestine tissue was moved out and dissected on ice divided into three sections depending on length. The intestinal contents were cleaned up by PBS and the small intestine was frozen immediately with liquid nitrogen. RNA was extracted from intestine tissue by RNA kit (Zymo Quick‐RNA Miniprep, R1055) for transcriptome sequencing.

Statistical analyses

Unless stated, all statistical analyses were performed using the Prism 8 software.

Author contributions

Chih‐Chan Lee: Data curation; Formal analysis; Writing—original draft. Feng Liang: Data curation; Formal analysis. I‐Chi Lee: Data curation; Formal analysis. Tsung‐Hao Lu: Data curation; Formal analysis. Yu‐Yau Shan: Data curation; Formal analysis. Chih‐Fan Jeng: Formal analysis. Yan‐Fang Zou: Data curation. Hon‐Tsen Yu: Conceptualization; Methodology. Shih‐Kuo Chen: Conceptualization; Supervision; Funding acquisition; Writing—original draft; Project administration; Writing—review & editing.

In addition to the CRediT author contributions listed above, the contributions in detail are:

H‐TY, S‐KC, and C‐CL designed the experiment and wrote the manuscript. C‐CL, FL, I‐CL, T‐HL, Y‐YS, C‐FJ, and Y‐FZ performed all experiments and analyses.

Disclosure and competing interests statement

The authors declare that they have no conflict of interest.

Supporting information

Acknowledgements

This work was supported by the Taiwan Ministry of Science and Technology grant MOST 103‐2628‐B‐002‐001‐MY3 and 111‐2636‐B‐002‐021 (to S.‐K.C.). We thank the Technology Commons, College of Life Science at National Taiwan University for technical assistance with NGS, Dr. Yi‐Juang Chern and Dr. Bon‐Chu Chung for their support and discussions, and Dr. Allan Y.‐C. Chang for writing assistance. We also thank Dr. Samer Hattar for providing transgenic mice for this study.

EMBO reports (2022) 23: e52316.

Data availability

The datasets produced in this study are available in the following databases: 16s rRNA sequence data: National Center for Biotechnology Information, U.S. National Library of Medicine, Sequence Read Archive PRJNA746721 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA746721). Transcriptome data: National Center for Biotechnology Information, U.S. National Library of Medicine, Sequence Read Archive PRJNA749005 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA749005).

References

  1. Altimus CM, Guler AD, Alam NM, Arman AC, Prusky GT, Sampath AP, Hattar S (2010) Rod photoreceptors drive circadian photoentrainment across a wide range of light intensities. Nat Neurosci 13: 1107–1112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson MJ (2008) A new method for non‐parametric multivariate analysis of variance. Austral Ecol 26: 32–46 [Google Scholar]
  3. Baver SB, Pickard GE, Sollars PJ, Pickard GE (2008) Two types of melanopsin retinal ganglion cell differentially innervate the hypothalamic suprachiasmatic nucleus and the olivary pretectal nucleus. Eur J Neurosci 27: 1763–1770 [DOI] [PubMed] [Google Scholar]
  4. Benedict C, Vogel H, Jonas W, Woting A, Blaut M, Schurmann A, Cedernaes J (2016) Gut microbiota and glucometabolic alterations in response to recurrent partial sleep deprivation in normal‐weight young individuals. Mol Metab 5: 1175–1186 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bercik P, Denou E, Collins J, Jackson W, Lu J, Jury J, Deng Y, Blennerhassett P, Macri J, McCoy KD et al (2011) The intestinal microbiota affect central levels of brain‐derived neurotropic factor and behavior in mice. Gastroenterology 141: 599–609.e3 [DOI] [PubMed] [Google Scholar]
  6. Berson DM, Dunn FA, Takao M (2002) Phototransduction by retinal ganglion cells that set the circadian clock. Science 295: 1070–1073 [DOI] [PubMed] [Google Scholar]
  7. Buffington SA, Di Prisco GV, Auchtung TA, Ajami NJ, Petrosino JF, Costa‐Mattioli M (2016) Microbial reconstitution reverses maternal diet‐induced social and synaptic deficits in offspring. Cell 165: 1762–1775 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Calligaro H, Coutanson C, Najjar RP, Mazzaro N, Cooper HM, Haddjeri N, Felder‐Schmittbuhl MP, Dkhissi‐Benyahya O (2019) Rods contribute to the light‐induced phase shift of the retinal clock in mammals. PLoS Biol 17: e2006211 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen SK, Badea TC, Hattar S (2011) Photoentrainment and pupillary light reflex are mediated by distinct populations of ipRGCs. Nature 476: 92–95 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chew KS, Renna JM, McNeill DS, Fernandez DC, Keenan WT, Thomsen MB, Ecker JL, Loevinsohn GS, VanDunk C, Vicarel DC et al (2017) A subset of ipRGCs regulates both maturation of the circadian clock and segregation of retinogeniculate projections in mice. Elife 6: e22861 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Diaz Heijtz R, Wang S, Anuar F, Qian Y, Bjorkholm B, Samuelsson A, Hibberd ML, Forssberg H, Pettersson S (2011) Normal gut microbiota modulates brain development and behavior. Proc Natl Acad Sci USA 108: 3047–3052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fan S‐Y, Chang Y‐T, Chen C‐L, Wang W‐H, Pan M‐K, Chen W‐P, Huang W‐Y, Xu Z, Huang H‐E, Chen T et al (2018) External light activates hair follicle stem cells through eyes via an ipRGC‐SCN‐sympathetic neural pathway. Proc Natl Acad Sci USA 115: E6880–E6889 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fernandez DC, Chang YT, Hattar S, Chen SK (2016) Architecture of retinal projections to the central circadian pacemaker. Proc Natl Acad Sci USA 113: 6047–6052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fernandez DC, Fogerson PM, Lazzerini Ospri L, Thomsen MB, Layne RM, Severin D, Zhan J, Singer JH, Kirkwood A, Zhao H et al (2018) Light affects mood and learning through distinct retina‐brain pathways. Cell 175: 71–84.e18 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fernandez DC, Komal R, Langel J, Ma J, Duy PQ, Penzo MA, Zhao H, Hattar S (2020) Retinal innervation tunes circuits that drive nonphotic entrainment to food. Nature 581: 194–198 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fierer N, Lauber CL, Zhou N, McDonald D, Costello EK, Knight R (2010) Forensic identification using skin bacterial communities. Proc Natl Acad Sci USA 107: 6477–6481 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fonken LK, Workman JL, Walton JC, Weil ZM, Morris JS, Haim A, Nelson RJ (2010) Light at night increases body mass by shifting the time of food intake. Proc Natl Acad Sci USA 107: 18664–18669 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Godinho‐Silva C, Domingues RG, Rendas M, Raposo B, Ribeiro H, da Silva JA, Vieira A, Costa RM, Barbosa‐Morais NL, Carvalho T et al (2019) Light‐entrained and brain‐tuned circadian circuits regulate ILC3s and gut homeostasis. Nature 574: 254–258 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Güler AD, Ecker JL, Lall GS, Haq S, Altimus CM, Liao H‐W, Barnard AR, Cahill H, Badea TC, Zhao H et al (2008) Melanopsin cells are the principal conduits for rod‐cone input to non‐image‐forming vision. Nature 453: 102–105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hatori M, Le H, Vollmers C, Keding SR, Tanaka N, Buch T, Waisman A, Schmedt C, Jegla T, Panda S (2008) Inducible ablation of melanopsin‐expressing retinal ganglion cells reveals their central role in non‐image forming visual responses. PLoS One 3: e2451 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hattar S, Kumar M, Park A, Tong P, Tung J, Yau KW, Berson DM (2006) Central projections of melanopsin‐expressing retinal ganglion cells in the mouse. J Comp Neurol 497: 326–349 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hattar S, Liao HW, Takao M, Berson DM, Yau KW (2002) Melanopsin‐containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 295: 1065–1070 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hattar S, Lucas RJ, Mrosovsky N, Thompson S, Douglas RH, Hankins MW, Lem J, Biel M, Hofmann F, Foster RG et al (2003) Melanopsin and rod‐cone photoreceptive systems account for all major accessory visual functions in mice. Nature 424: 76–81 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. He Z, Kong X, Shao T, Zhang Y, Wen C (2019) Alterations of the gut microbiota associated with promoting efficacy of prednisone by bromofuranone in MRL/lpr mice. Front Microbiol 10: 978 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hughes ME, Hogenesch JB, Kornacker K (2010) JTK_CYCLE: an efficient nonparametric algorithm for detecting rhythmic components in genome‐scale data sets. J Biol Rhythms 25: 372–380 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Husse J, Leliavski A, Tsang AH, Oster H, Eichele G (2014) The light‐dark cycle controls peripheral rhythmicity in mice with a genetically ablated suprachiasmatic nucleus clock. FASEB J 28: 4950–4960 [DOI] [PubMed] [Google Scholar]
  27. Izumo M, Pejchal M, Schook AC, Lange RP, Walisser JA, Sato TR, Wang X, Bradfield CA, Takahashi JS (2014) Differential effects of light and feeding on circadian organization of peripheral clocks in a forebrain Bmal1 mutant. Elife 3: e4617 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Joseph J, Boggan W, Powell DA (1979) Adrenal demedullation and peripheral 6‐OHDA administration in the rabbit: effects on body weight, general activity and cardiovascular responsivity. Pharmacol Biochem Behav 10: 875–881 [DOI] [PubMed] [Google Scholar]
  29. Kiessling S, Sollars PJ, Pickard GE (2014) Light stimulates the mouse adrenal through a retinohypothalamic pathway independent of an effect on the clock in the suprachiasmatic nucleus. PLoS One 9: e92959 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Krych L, Nielsen DS, Hansen AK, Hansen CH (2015) Gut microbial markers are associated with diabetes onset, regulatory imbalance, and IFN‐gamma level in NOD mice. Gut Microbes 6: 101–109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kuang Z, Wang Y, Li Y, Ye C, Ruhn KA, Behrendt CL, Olson EN, Hooper LV (2019) The intestinal microbiota programs diurnal rhythms in host metabolism through histone deacetylase 3. Science 365: 1428–1434 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Leiva‐Gea I, Sanchez‐Alcoholado L, Martin‐Tejedor B, Castellano‐Castillo D, Moreno‐Indias I, Urda‐Cardona A, Tinahones FJ, Fernandez‐Garcia JC, Queipo‐Ortuno MI (2018) Gut microbiota differs in composition and functionality between children with type 1 diabetes and MODY2 and healthy control subjects: a case‐control study. Diabetes Care 41: 2385–2395 [DOI] [PubMed] [Google Scholar]
  33. Leone V, Gibbons S, Martinez K, Hutchison A, Huang E, Cham C, Pierre J, Heneghan A, Nadimpalli A, Hubert N et al (2015) Effects of diurnal variation of gut microbes and high‐fat feeding on host circadian clock function and metabolism. Cell Host Microbe 17: 681–689 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Liang X, Bushman FD, FitzGerald GA (2015) Rhythmicity of the intestinal microbiota is regulated by gender and the host circadian clock. Proc Natl Acad Sci USA 112: 10479–10484 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Lucas RJ, Hattar S, Takao M, Berson DM, Foster RG, Yau KW (2003) Diminished pupillary light reflex at high irradiances in melanopsin‐knockout mice. Science 299: 245–247 [DOI] [PubMed] [Google Scholar]
  36. McMahan RS, Birkland TP, Smigiel KS, Vandivort TC, Rohani MG, Manicone AM, McGuire JK, Gharib SA, Parks WC (2016) Stromelysin‐2 (MMP10) moderates inflammation by controlling macrophage activation. J Immunol 197: 899–909 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Menni C, Hernandez MM, Vital M, Mohney RP, Spector TD, Valdes AM (2019) Circulating levels of the anti‐oxidant indoleproprionic acid are associated with higher gut microbiome diversity. Gut Microbes 10: 688–695 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Miranda‐Ribera A, Ennamorati M, Serena G, Cetinbas M, Lan J, Sadreyev RI, Jain N, Fasano A, Fiorentino M (2019) Exploiting the zonulin mouse model to establish the role of primary impaired gut barrier function on microbiota composition and immune profiles. Front Immunol 10: 2233 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mukherji A, Kobiita A, Ye T, Chambon P (2013) Homeostasis in intestinal epithelium is orchestrated by the circadian clock and microbiota cues transduced by TLRs. Cell 153: 812–827 [DOI] [PubMed] [Google Scholar]
  40. Neufeld KM, Kang N, Bienenstock J, Foster JA (2011) Reduced anxiety‐like behavior and central neurochemical change in germ‐free mice. Neurogastroenterol Motil 23: 255–e119 [DOI] [PubMed] [Google Scholar]
  41. Panda S, Sato TK, Castrucci AM, Rollag MD, DeGrip WJ, Hogenesch JB, Provencio I, Kay SA (2002) Melanopsin (Opn4) requirement for normal light‐induced circadian phase shifting. Science 298: 2213–2216 [DOI] [PubMed] [Google Scholar]
  42. Prigge CL, Yeh P‐T, Liou N‐F, Lee C‐C, You S‐F, Liu L‐L, McNeill DS, Chew KS, Hattar S, Chen S‐K et al (2016) M1 ipRGCs influence visual function through retrograde signaling in the retina. J Neurosci 36: 7184–7197 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Provencio I, Rollag MD, Castrucci AM (2002) Photoreceptive net in the mammalian retina. This mesh of cells may explain how some blind mice can still tell day from night. Nature 415: 493 [DOI] [PubMed] [Google Scholar]
  44. Reitmeier S, Kiessling S, Clavel T, List M, Almeida EL, Ghosh TS, Neuhaus K, Grallert H, Linseisen J, Skurk T et al (2020) Arrhythmic gut microbiome signatures predict risk of type 2 diabetes. Cell Host Microbe 28: 258–272.e6 [DOI] [PubMed] [Google Scholar]
  45. Robinson H, Barrett H, Gomez‐Arango L, McIntyre HD, Callaway L, Dekker Nitert M (2019) Ketonuria Is associated with changes to the abundance of Roseburia in the gut microbiota of overweight and obese women at 16 weeks gestation: a cross‐sectional observational study. Nutrients 11: 1836 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Ruby NF, Brennan TJ, Xie X, Cao V, Franken P, Heller HC, O'Hara BF (2002) Role of melanopsin in circadian responses to light. Science 298: 2211–2213 [DOI] [PubMed] [Google Scholar]
  47. Rupp AC, Ren M, Altimus CM, Fernandez DC, Richardson M, Turek F, Hattar S, Schmidt TM (2019) Distinct ipRGC subpopulations mediate light's acute and circadian effects on body temperature and sleep. Elife 8: e44358 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Sampson TR, Debelius JW, Thron T, Janssen S, Shastri GG, Ilhan ZE, Challis C, Schretter CE, Rocha S, Gradinaru V et al (2016) Gut microbiota regulate motor deficits and neuroinflammation in a model of Parkinson's disease. Cell 167: 1469–1480.e12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Schmidt TM, Kofuji P (2010) Differential cone pathway influence on intrinsically photosensitive retinal ganglion cell subtypes. J Neurosci 30: 16262–16271 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sonoda T, Li JY, Hayes NW, Chan JC, Okabe Y, Belin S, Nawabi H, Schmidt TM (2020) A noncanonical inhibitory circuit dampens behavioral sensitivity to light. Science 368: 527–531 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Sudo N, Chida Y, Aiba Y, Sonoda J, Oyama N, Yu XN, Kubo C, Koga Y (2004) Postnatal microbial colonization programs the hypothalamic‐pituitary‐adrenal system for stress response in mice. J Physiol 558: 263–275 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Tahara Y, Kuroda H, Saito K, Nakajima Y, Kubo Y, Ohnishi N, Seo Y, Otsuka M, Fuse Y, Ohura Y et al (2012) In vivo monitoring of peripheral circadian clocks in the mouse. Curr Biol 22: 1029–1034 [DOI] [PubMed] [Google Scholar]
  53. Thaiss C, Zeevi D, Levy M, Zilberman‐Schapira G, Suez J, Tengeler A, Abramson L, Katz M, Korem T, Zmora N et al (2014) Transkingdom control of microbiota diurnal oscillations promotes metabolic homeostasis. Cell 159: 514–529 [DOI] [PubMed] [Google Scholar]
  54. Thoenen H, Tranzer JP (1968) Chemical sympathectomy by selective destruction of adrenergic nerve endings with 6‐Hydroxydopamine. Naunyn Schmiedebergs Arch Exp Pathol Pharmakol 261: 271–288 [DOI] [PubMed] [Google Scholar]
  55. Voigt RM, Forsyth CB, Green SJ, Mutlu E, Engen P, Vitaterna MH, Turek FW, Keshavarzian A (2014) Circadian disorganization alters intestinal microbiota. PLoS One 9: e97500 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wu G, Tang W, He Y, Hu J, Gong S, He Z, Wei G, Lv L, Jiang Y, Zhou H et al (2018) Light exposure influences the diurnal oscillation of gut microbiota in mice. Biochem Biophys Res Commun 501: 16–23 [DOI] [PubMed] [Google Scholar]
  57. Zackular JP, Baxter NT, Iverson KD, Sadler WD, Petrosino JF, Chen GY, Schloss PD (2013) The gut microbiome modulates colon tumorigenesis. MBio 4: e00692–e613 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zarrinpar A, Chaix A, Yooseph S, Panda S (2014) Diet and feeding pattern affect the diurnal dynamics of the gut microbiome. Cell Metab 20: 1006–1017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zhang SL, Bai L, Goel N, Bailey A, Jang CJ, Bushman FD, Meerlo P, Dinges DF, Sehgal A (2017) Human and rat gut microbiome composition is maintained following sleep restriction. Proc Natl Acad Sci USA 114: E1564–E1571 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    Data Availability Statement

    The datasets produced in this study are available in the following databases: 16s rRNA sequence data: National Center for Biotechnology Information, U.S. National Library of Medicine, Sequence Read Archive PRJNA746721 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA746721). Transcriptome data: National Center for Biotechnology Information, U.S. National Library of Medicine, Sequence Read Archive PRJNA749005 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA749005).

    Articles from EMBO Reports are provided here courtesy of Nature Publishing Group

    RESOURCES