Circadian rhythms of fetal liver transcription persist in the absence of canonical circadian clock gene expression rhythms in vivo

Abstract

The cellular circadian clock and systemic cues drive rhythmicity in the transcriptome of adult peripheral tissues. However, the oscillating status of the circadian clocks in fetal tissues, and their response to maternal cues, are less clear. Most clock genes do not cycle in fetal livers from mice and rats, although tissue level rhythms rapidly emerge when fetal mouse liver explants are cultured in vitro. Thus, in the fetal mouse liver, the circadian clock does not oscillate at the cellular level (but is induced to oscillate in culture). To gain a comprehensive overview of the clock status in the fetal liver during late gestation, we performed microarray analyses on fetal liver tissues. In the fetal liver we did not observe circadian rhythms of clock gene expression or many other transcripts known to be rhythmically expressed in the adult liver. Nevertheless, JTK_CYCLE analysis identified some transcripts in the fetal liver that were rhythmically expressed, albeit at low amplitudes. Upon data filtering by coefficient of variation, the expression levels for transcripts related to pancreatic exocrine enzymes and zymogen secretion were found to undergo synchronized daily fluctuations at high amplitudes. These results suggest that maternal cues influence the fetal liver, despite the fact that we did not detect circadian rhythms of canonical clock gene expression in the fetal liver. These results raise important questions on the role of the circadian clock, or lack thereof, during ontogeny.

Figure 1. Heatmap of 66 probe sets representing clock and known rhythmic genes.

Natural scale expression values from series 2 data for core clock genes and genes known to be rhythmically expressed in the adult liver were plotted with Heatmap builder. Some genes were represented by multiple probe sets. Phases were sorted by the “lag” values given by JTK_CYCLE.

Figure 2. Heatmap of the 145 rhythmic probe sets in series 2 microarray data.

Series 2 natural scale expression values for the 145 probe sets that were rhythmic in both series of data (p<0.1 by JTK_CYCLE) were plotted. Phases were sorted by the “lag” values given by JTK_CYCLE.

Figure 3. Heatmap of pancreatic exocrine enzyme-related probe sets in series 2 data.

Normalized log2 scale expression values for the probe sets representing transcripts related to pancreatic exocrine enzymes and zymogen secretion from series 2 data were plotted. Log2 scale data were used to facilitate viewing of all expression peaks. More dramatic variations in expression peaks in natural scale data led to improper viewing of those peaks on heatmap.

Figure 4. Semi-quantitative RT-PCR analyses of pancreatic exocrine enzymes-related transcripts in fetal liver tissues.

Equal amounts of starting RNA from the 12 time points of series 2 fetal liver tissues were reverse transcribed and subject to semi-quantitative PCR. PCR cycles were adjusted depending on transcripts abundance. Products were visualized by agarose gel electrophoresis.