Fig 1.
Selected feedback-mediating microRNAs involved in the p53-MDM2 pathway.
(a) The positive feedback loop is formed by miR-192, MDM2 and p53. (b) The positive feedback loop is formed by miR-34a and p53, mediated by SIRT1 or YY1 plus MDM2. (c) The positive feedback loop is formed by miR-29a and p53, mediated by CDC42, Wip1, or Wip1 plus MDM2. Note that blue arrows represent activation while the red lines with a hammerhead represent inhibition. Note that the inhibition and activation signs here indicate the regulatory role of a regulator on its target. That is, if the regulation is of an activating (or inhibitory) nature, we use an activation (or inhibition) sign.
Fig 2.
Diagram of the p53-MDM2 oscillator under the regulation of positive feedbacks via microRNA-192, -34a and -29a.
p53 is translated from p53 mRNA and remains inactive. Phosphorylated by ATM*, p53 becomes active (p53*), and able to transcribe mdm2 mRNA. MDM2 protein, translated from mdm2 mRNA, promotes a fast degradation of p53 and a slow degradation of p53*. In addition to a basal self-degradation, MDM2 is degraded by a mechanism stimulated by ATM*. The three microRNAs, miR-192, miR-34a and miR-29a, are induced by p53*, and inhibit the mRNAs of mdm2, cdc42, wip1, sirt1 and yy1, whose protein products further regulate p53* and MDM2. Specifically, the microRNA binds with its target mRNA molecule with high affinity, forming a microRNA-mRNA complex, and subsequently dispose the complex into a degradation machinery. In other words, the microRNAs in our model are assumed to enhance the degradation of their mRNA target by complexation and subsequent disposal. In particular, CDC42, Wip1 and SIRT1 proteins deactivate p53 directly, while YY1 enhances the MDM2-dependent degradation of p53 and p53* proteins. In addition, Wip1 protein inhibits the degradation of MDM2 protein. The wip1 mRNA is also induced by p53*, whose protein product inhibits active ATM, forming a second negative feedback loop.
Fig 3.
Simulations of the time courses of p53 data shows association rates sensitive to variation.
(a) Simulations are performed using the ODE model of microRNA-p53-MDM2 network under constant ATM activation. Each microRNA is modeled by accounting for the mediating microRNA component and its associated target proteins. The association rates between the microRNAs (miR 192, 29a and 34a) and target mRNAs (kon1~5) in this in silico model were varied one at a time to predict its subsequent effect on the oscillatory p53 expression. (b) Expression levels of p53 (green) and MDM2 (red) as calculated by the simulation, with a nominal association rate value of 25 for kon1~5 (top). (c) The simulation was repeated with nominal inhibition of miR-192, miR-34a and miR-24a. (d)The simulation was repeated at varying level of inhibitor of miR-192, miR-34a and miR-24a by changing the basal induction rate of the microRNAs. The color scale from light green to dark green represents the response with a basal induction rate of microRNAs changing from zero to the respective nominal value, represented by the dotted line, and further above the nominal value.
Fig 4.
Bifurcation diagrams of the DNA double-strand-break triggered steady-state p53 response versus the association rates between the three microRNAs and their five target mRNAs (kon1, kon2, kon3, kon4, kon5) embedded in the positive feedback loops, under the wild-type (black), miR-192 repressed (purple), miR-34a repressed (green), and miR-29a repressed (blue) conditions.
Paired dots represent the bounds of p53 oscillation amplitude and the solid line represents stationary steady state. The red vertical line indicates the nominal parameter value. The inhibition of miR-192 leads to shrunk oscillating region or entirely non-oscillating region over varying parameter ranges. The inhibition of miR-34a and miR-29a only mildly affects the system behavior compared to the wild-type condition.
Fig 5.
Experimental perturbation of p53-associated microRNA affects oscillatory behavior.
(a) To examine the role of microRNA 29a, 34a and 192 in the stress-induced p53 oscillations, expression of three endogenous microRNAs known to be associated with p53 activation pathway is targeted using microRNA inhibitor. Fluorescence is captured every 10 minutes for 20 hours following transfection. Single cells are identified and their intensity is recorded to create a time series data for each cell. (b) Quantitative PCR measurements of microRNA levels before and after treatment with the microRNA inhibitor. (c) Period distribution of p53 oscillation before and after treatment with various microRNA inhibitor. Magenta bars represent period of p53 oscillations observed in wild type population, and brown bars represent the same measurement from the population treated with the specified microRNA inhibitor. (d) Heatmaps showing single cell tracks of p53 fluorescence measurement after various microRNA inhibitor treatment. Each graph has 100 rows, each row representing a single cell tracking data for 20 hours. Data shown has been reorganized to reflect the results from the classification process to identify cells with oscillatory p53 expression. Number on the left shows the number of cells identified as oscillating for each group, where the magenta bar indicates the wild-type level.