Frequency control of cell cycle oscillators

Abstract

The cell cycle oscillator, based on a core negative feedback loop and modified extensively by positive feedback, cycles with a frequency that is regulated by environmental and developmental programs to encompass a wide range of cell cycle times. We discuss how positive feedback allows frequency tuning, how size and morphogenetic checkpoints regulate oscillator frequency, and how extrinsic oscillators such as the circadian clock gate cell cycle frequency. The master cell cycle regulatory oscillator in turn controls the frequency of peripheral oscillators controlling essential events. A recently proposed phase-locking model accounts for this coupling.

Figure 1

Positive and negative feedback loops in the cyclin-CDK oscillator. A Inset: a negative feedback loop which can give rise to oscillations. Such an oscillator is thought to form the core of eukaryotic cell cycles, with cyclin-Cyclin Dependent Kinase (cyclin-CDK) acting as activator, Anaphase Promoting Complex-Cdc20 (APC-Cdc20) acting as repressor, and non-linearity in APC-Cdc20 activation preventing the system from settling into a steady state. Below is shown the cyclin-CDK machinery in eukaryotic cell cycles. CDKs, present throughout the cell cycle, require the binding of a cyclin subunit for activity. These cyclin partners can also determine the localization of the complex and its specificity for targets. At the beginning of the cell cycle, cyclin-CDK activity is low, and ramps up over most of the cycle. Early cyclins trigger production of later cyclins and these later cyclins then turn off the earlier cyclins, so that control is passed from one set of cyclin-CDKs to the next. The last set of cyclins to be activated, the G2/M-phase cyclins, initiate mitosis, and also initiate their own destruction by activating the APC-Cdc20 negative feedback loop. APC-Cdc20 targets the G2/M-phase cyclins for destruction, resetting the cell to a low-CDK activity state, ready for the next cycle. B Positive feedback is added to the oscillator in multiple ways. Left: a highly-conserved but non-essential mechanism consists of “handoff” of cyclin proteolysis from APC-Cdc20 to APC-Cdh1. Cdh1 is a relative of Cdc20 which activates the APC late in mitosis and into the ensuing G1. Cdh1 is inhibited by cyclin-CDK activity, resulting in mutual inhibition (which is logically equivalent to positive feedback). Middle: antagonism between cyclin-CDK and stoichiometric CDK inhibitors (CKIs) results in positive feedback. These loops stabilize high- and low-CDK activity states. Right: a double positive feedback loop comprising CDK-mediated inhibition of the Wee1 kinase (which inhibits CDK) and activation of the Cdc25 phosphatase (which activates CDK by removing the phosphorylation added by Wee1) is proposed to stabilize intermediate CDK activity found in mid-cycle, and an alternative stable state of high mitotic CDK activity.

Figure 2

Size control mechanisms in budding and fission yeasts. Top: in S. cerevisiae, size control operates in G1. Transcription of many genes, including the G1/S cyclins (Cln1,2) is controlled by the SBF and MBF transcription factors (Swi4/Swi6 and Swi4/Mbp1, respectively). The Whi5 repressor inhibits this transcription until it is exported from the nucleus by the most upstream G1 cyclin, Cln3, in response to sufficient cell size. Cln3 thus relieves transcriptional inhibition, promoting Cln1,2 expression and subsequent cell cycle Start. Actual size “measurement” was recently proposed to operate through direct binding of Cln3 to the SCB target sequences of Swi4/Swi6, with Start occurring upon titration of these sites by Cln3. Bottom: in S. pombe, size control operates in G2. Pom1, localized to cell poles, indirectly inhibits CDK activity (through inhibition of Cdr2, which inhibits Wee1, which in turn inhibits CDK). As the cell elongates, the concentration of Pom1 at the center of the cell (where the nucleus is located) drops, allowing CDK activation leading to mitosis.

Figure 3

A phase-locking model for entrainment of peripheral oscillators to the cyclin-CDK ocillator. A Molecular mechanism of the Cdc14 release oscillator. The mitotic phosphatase Cdc14 is activated upon release from sequestration in the nucleolus. This release is controlled by a negative feedback loop in which Cdc14 release, promoted by the polo kinase Cdc5, activates APC-Cdh1, which then promotes Cdc5 degradation, allowing Cdc14 resequestration. This negative feedback oscillator is entrained to the cyclin-CDK cycle at multiple points: by cyclin-CDK promotion of CDC5 transcription and Cdc5 kinase activation, and by cyclin-CDK inhibition of Cdh1 activity. B Schematic of multiple peripheral oscillators coupled to the CDK oscillator in budding yeast. As described above, coupling entrains such peripheral oscillators to cell cycle progression; peripheral oscillators also feed back on the cyclin-CDK oscillator itself. For example, major genes in the periodic transcription program include most cyclins, CDC20, and CDC5; Cdc14 directly promotes establishment of the low-cyclin-CDK positive feedback loop by activating Cdh1 and Sic1 as well as more indirectly antagonizing cyclin-CDK activity by dephosphorylating cyclin-CDK targets; the centrosome and budding cycles could communicate with the cyclin-CDK cycle via the spindle integrity and morphogenesis checkpoints. C Oscillator coupling ensures once-per-cell-cycle occurrence of events. Three hypothetical oscillators are shown: a master cycle in black, a faster peripheral cycle in blue, and a slower peripheral cycle in red. In the absence of phase-locking (top), the oscillators trigger events (colored circles) without a coherent phase relationship. In the presence of oscillator coupling (bottom), the peripheral oscillators are slowed or accelerated within their critical periods to produce a locked phase relationship, with events occurring once and only once within each master cycle.