Disruption of circadian rhythms in mice by constant light.
Mice were subjected to constant 24h light (LL) for 4 weeks, with the control group maintained under 12-hour light/dark cycles (LD). Voluntary wheel running was measured to evaluate behavioral circadian rhythms at the beginning of the study (baseline) and after two weeks of exposure to normal light or constant light. All mice showed regular running rhythms at baseline and following two weeks in normal light (LD) (Figure 1A). Following two weeks of disrupted light (LL), mice displayed consolidated voluntary running rhythms; however, the time of activity onset was altered (Figures 1A and 1B). The phase angle is a circadian parameter that describes the difference in hours between the time of lights-off (or former lights-off for LL 2nd week) which was 18:00 (zeitgeber time [ZT] 12), and the time of activity onset; corresponding to uninterrupted activity on the wheel. In the LL group at the 2nd week, activity was significantly shifted by ~30 min (Figure 1B).
In addition to wheel running, body temperature was measured as a second method to evaluate circadian rhythmicity. Following two weeks of normal light mice showed a regular pattern of body temperature, elevated during the night/active time (18:00 [ZT12] until 06:00 [ZT0]) and decreased during the day/resting time (Figure 1C). Mice exposed to circadian disruption displayed an arrhythmic pattern in body temperature throughout the day (Figure 1D).
Body mass was measured at baseline and then once a week for 4 weeks. Mice maintained in normal light exhibited a typical, moderate increase in body weight. On the other hand, mice subjected to circadian disruption showed significantly increased body mass compared with mice exposed to normal light for 4 weeks (Figure 1E).
Circadian rhythm disruption represses the expression of β-catenin and circadian clock genes in the gut.
We next analyzed the impact of circadian disruption on mRNA (Figure 2A) and protein expression (Figures 2B-2C) of circadian clock molecules, such as Clock and Bmal1, in control mice and mice with disrupted circadian cycles. These experiments included also β-catenin expression as the WNT/β-catenin pathway was demonstrated to be under a strong circadian control [26]. Alterations of circadian rhythms by LL significantly reduced intestinal Clock and Bmal1 expression at both mRNA levels (by ~70% and 50%, respectively) and protein levels (by ~55% and 65%, respectively) (Figures 2A-2C). Interestingly, we also detected a significant downregulation of ß-catenin by 65% at the mRNA level and by 35% at the protein levels in the LL group as compared to LD controls.
Circadian rhythm disruption alters intestinal permeability and TJ protein expression.
Dysfunction of intestinal integrity may be one of the main outcomes of circadian rhythm disruption. Therefore, intestinal permeability was evaluated in the ileum and colon segments in mice maintained under normal and constant light conditions. The analyses were performed ex vivo by perfusion of ileum or colon sections of the intestine with FITC-dextran 4 kDa. The circadian rhythm disrupted group (LL) experienced a significant increase in intestinal permeability compared to LD controls both in the ileum (Figure 3A) and the colon (Figure 3B). Changes in permeability measures were more pronounced and consistent in the colon sections, therefore colon cells were employed in subsequent in vitro experiments in Figures 5-7.
Alterations in intestinal permeability were accompanied by a significant downregulation of mRNA levels of TJ genes, namely, ZO-1, occludin, and tricellulin (p=0.015, p=0.023, and p=0.007, respectively) (Figure 4A). The loss of tricellulin is important because tricellulin can normally replace occludin in cell junctions [27]. We next analyzed TJ protein expression by immunoblotting (Figure 4B, representative blots, Figure 4C, quantified data). Consistent with the results on mRNA expression, disruption of circadian rhythms markedly downregulated protein levels of all TJ molecules evaluated in the present study by ~50% with at the significance level of p<0.032 for ZO-1, p<0.0001 for occludin, and p<0.0004 for tricellulin.
The expression of the clock molecules is controlled by β-catenin.
In order to investigate a potential correlation between β-catenin and clock gene expression, we returned to in vitro studies based on human colon SW480 cells. β-catenin was silenced using specific siRNA, followed by analysis of clock genes mRNA and protein expression (Figure 5). Silencing β-catenin decreased its mRNA levels by ~80% and protein expression by ~30% (Figure 5A, left and right panel, respectively). Interestingly, β-catenin silencing diminished Bmal1 mRNA and protein levels by ~40% (Figure 5B). Silencing of β-catenin also effectively decreased the clock protein level by ~40% (Figure 5C, right panel). Overall, these results suggest that β-catenin is upstream from the clock genes and proteins and may modulate their expression in intestinal cells.
Silencing of circadian clock genes increases epithelial permeability.
To evaluate the impact of clock genes on epithelial barrier function, β-catenin, Clock, or Bmal1 expression was silenced in SW480 cells with specific siRNAs, followed by measuring permeability for FITC-dextran 20 kDa in the Transwell system. Permeability across SW480 monolayers was significantly increased by 24% and 27% upon silencing with ß-catenin siRNA at 20 and 100 nM, respectively (Figure 6A). In addition, silencing of Bmal1 with specific siRNA at 10 and 50 nM increased permeability by ~28%. Silencing the Clock gene with 10 nM siRNA was ineffective, however, silencing with 50 nM significantly increased permeability by 19% (Figure 6B).
Taking into consideration the substantial impact of β-catenin on the regulation of epithelial barrier function and the role of β-catenin in circadian rhythm regulation, we next analyzed the consequence of β-catenin knockdown on TJ protein expression. Silencing of β-catenin with 20nM of specific siRNA did not alter ZO-1 mRNA levels; however, it decreased its protein levels (Figure 6C). Similarly, β-catenin silencing did not affect occludin gene expression. While there was a ~20% reduction in occludin mRNA expression as compared vehicle, similar downregulation of occludin mRNA was observed in cells transfected with scrambled siRNA, suggesting non-specific responses. As seen with ZO-1, silencing of β-catenin with 20 nM siRNA significantly decreased occludin protein expression (Figure 6D).
β-catenin regulates epithelial barrier function via upregulation of MMPs.
A decrease in TJ protein expression without changes on mRNA levels suggests post-transcriptional modification. Therefore, we evaluated a possible involvement of MMPs, which can modulate permeability by degradation of TJ proteins [28]. Transfection with β-catenin siRNA significantly upregulated the mRNA expression of MMP-2 and MMP-9 (p < 0.01) (Figures 7A and 7B, respectively). In addition, silencing of β-catenin significantly increased permeability across epithelial monolayers created by SW480 cells. In order to determine if increased production of MMPs may be involved in this effect, cells were transfected with β-catenin siRNA and co-treated with MMP-2 or MMP-9 inhibitors. Specific inhibitors for MMP-2 or MMP-9 individually did not affect β-catenin siRNA-induced elevation of epithelial permeability. However, inhibition with a dual-action blocker of both MMP-2 and MMP-9 attenuated disruption of permeability induced by β-catenin siRNA (Figures 7C), suggesting that MMPs may modulate the impact of β-catenin on epithelial barrier function in the context of circadian disruption.