Ultradian oscillations of Stat, Smad, and Hes1 expression in response to serum

Abstract

Serum response has been used as a model for studying signaling transduction for many biological events such as cell proliferation and survival. Although expression of many genes is up- or down-regulated after serum stimulation, the Notch effector Hes1 displays oscillatory response. However, the precise mechanism and biological significance of this oscillation remain to be determined. Here, we identified serum-induced ultradian oscillators, including molecules in Stat and Smad signaling. Stat and Smad oscillations involve activation of Stat3 and Smad1 and delayed negative feedback by their inhibitors Socs3 and Smad6, respectively. Moreover, Stat oscillations induce oscillatory expression of Hes1 by regulating its half-life, and loss of Hes1 oscillations leads to G(1) phase retardation of the cell cycle. These results indicate that coupled Stat and Hes1 oscillations are important for efficient cell proliferation and provide evidence that expression modes of signaling molecules affect downstream cellular events.

Fig. 1.

Identification of serum-induced ultradian oscillators by microarray analysis. (A) Microarray analysis of serum-induced ultradian oscillators. Seven candidate genes are found. (B) Time course of Socs3, Hes1, Gse1, and Smad6 expression in real-time PCR analysis. Means with SE of three independent experiments are shown. The patterns of microarray and real-time PCR experiments are very similar to each other.

Fig. 2.

Oscillations in Stat signaling. Cells were treated with serum at t = 0 in the absence (control) or presence of AG490, dnStat3, Socs3, or siSocs3. (A) Expression profiles of Socs3 mRNA were analyzed at 10-min intervals by real-time PCR. The peaks of Socs3 mRNA appear at 50 and 170 min after serum stimulation. (B) Expression profiles of Socs3 mRNA were analyzed by real-time PCR. Socs3 mRNA oscillation is abolished in the presence of AG490, dnStat3, or siSocs3. (C) Expression profiles of Socs3 protein in the absence (control) or presence of AG490, dnStat3, or siSocs3 were analyzed by Western blotting. Socs3 protein oscillation is abolished in the presence of AG490, dnStat3, or siSocs3. (D) Expression profiles of p-Stat3 in the absence (control) or presence of AG490, Socs3, or siSocs3 were analyzed by Western blotting. p-Stat3 oscillation is abolished in the presence of AG490, Socs3, or siSocs3. Means with SE of three independent experiments are shown in all graphs.

Fig. 3.

Oscillations in Smad signaling. Cells were treated with serum at t = 0 in the absence (control) or presence of dnSmad1 or Smad6, and mRNA and protein levels were quantified by real-time PCR and Western blots, respectively. (A) Expression profiles of Smad6 mRNA were analyzed at 10-min intervals. The peaks of Smad6 mRNA appear at 120 and 230 min after serum stimulation. (B) Expression profiles of Smad6 mRNA in the absence (control) or presence of dnSmad1 were analyzed. Smad6 mRNA oscillation is abolished by dnSmad1. (C) Expression profiles of Smad6 protein in the absence (control) or presence of dnSmad1 were analyzed. Smad6 protein oscillation is abolished by dnSmad1. (D) Expression profiles of p-Smad1/5/8 in the absence (control) or presence of Smad6 were analyzed. p-Smad1/5/8 oscillation is abolished by Smad6. Means with SE of three independent experiments are shown in all graphs.

Fig. 4.

Regulation of Hes1 oscillation by Stat signaling. (A) Expression of Hes1 mRNA oscillates after serum stimulation (t = 0), but this oscillation is abolished by AG490, dnStat3, or Socs3. (B) Expression of Hes1 protein oscillates after serum stimulation (t = 0) (control), but this oscillation is abolished by AG490, dnStat3, or Socs3. (C) Measurement of Hes1 protein half-life. Hes1 protein levels were measured in the presence of cycloheximide (100 μM), which blocks new protein synthesis. Hes1 protein is degraded with the half-life of 22.4 ± 0.9 min (n = 3). This half-life is elongated by AG490, dnStat3, or Socs3 and shortened by siSocs3. Thus, Stat signaling regulates Hes1 protein stability. (D) Hes1 mRNA oscillation is abolished by siSocs3. (E) Hes1 protein oscillation is abolished by siSocs3. All values are the average of three independent experiments with SE.

Fig. 5.

Hes1 oscillation in cell cycles. (A) Comparison of growth curves. Cells that express Hes1 protein persistently and cells that lose Hes1 activity do not proliferate extensively, compared with the control, where Hes1 expression oscillates. (B) Regulation of cell cycle regulators by Hes1. Expression was examined by real-time PCR. Both sustained Hes1 expression (+Hes1) and knockdown of Hes1 activities (+dnHes1) increase expression of p21 and p27, G₁ phase genes, compared with oscillatory Hes1 expression (control). Transfected cells with the control vector and nontransfected cells gave the same results (data not shown). (C) Expression of cyclin D1, a G₁-specific marker. Both Hes1 and dnHes1 increase cyclin D1 expression, compared with control. All values are the average of three independent experiments with SE.

Fig. 6.

Mathematical simulation of Hes1 oscillations. Hes1 oscillations are mathematically simulated using a negative autorepression model. (A) When Hes1 protein half-life is changing, Hes1 expression shows stable oscillation. (B and C) In contrast, when the Hes1 protein degradation rate is fixed at the value b(t) = 0.8, which represents the absence of p-Stat3 formation (B), and b(t) = 2.77, which represents persistent formation of p-Stat3 (C), Hes1 oscillation becomes damped. This simulation well mimics Stat3-dependent Hes1 oscillations.