Cell-Cycle Compartments and Phase Durations

Why Represent the Cell Cycle as Compartments?

A total cell count can show whether a population is growing or shrinking, but it does not reveal where cells spend time before division. A cell-cycle compartment model tracks cells in successive phases, such as G1, S, G2, and M. It assigns each phase its own exit rate, so the model can represent differences in phase duration and calculate how many cells are moving between phases or completing division.

The model is a simplifying account of a cycling population, not a claim that every cell follows an identical clock. Its predictions depend on assumptions about how cells leave each phase, whether all cells follow the same sequence, and what happens when a cell completes M phase. State those assumptions before interpreting rates or fluxes.

Definition: Let \(C_i(t)\) be the number of cells in phase \(i\) at time \(t\), for \(i=1,\ldots,m\). A phase-specific exit rate \(k_i\) is the per-cell rate of leaving phase \(i\), with units of inverse time. Under a constant exit hazard, the phase-exit flux is \(J_i(t)=k_iC_i(t)\), measured in cells per unit time. For the final phase, \(J_m\) is the division flux: the number of completed cell divisions per unit time.

Assumptions and Phase Durations

In the basic model, cells progress through the ordered phases \(1\to2\to\cdots\to m\). Within phase \(i\), each cell has a constant exit hazard \(k_i\), shared by cells in that phase. There is no death, migration, branching to another phase, or return to an earlier phase. At the end of the final phase, one parent cell is removed and two daughter cells enter phase 1. The compartments are well mixed: cells in a given phase are represented by the same rate, rather than by their individual histories.

A constant exit hazard implies an exponential waiting time in that phase. For \(k_i>0\), the probability that a cell remains in phase \(i\) for at least \(t\) time units is \(e^{-k_it}\). The mean time spent in the phase follows by integrating this survival probability:

E[T_i]=\int_0^\infty P(T_i>t)\,dt =\int_0^\infty e^{-k_it}\,dt =\frac{1}{k_i}.

Thus a larger exit rate corresponds to a shorter mean phase duration. The rate is not a duration: if time is measured in hours, \(k_i\) has units \(\text{h}^{-1}\), while \(1/k_i\) has units of hours. A cell's phase duration is variable under this assumption; it is not exactly \(1/k_i\) for every cell.

If cells pass through all \(m\) phases before dividing, their total cycle time is the sum of their phase times. By additivity of expectation, the mean cycle duration is the sum of the mean phase durations:

E[T_{\mathrm{cycle}}] =\sum_{i=1}^{m} E[T_i] =\sum_{i=1}^{m}\frac{1}{k_i}.

This result gives the expected time to complete the sequence under the stated model. It does not imply that every cell divides after this exact interval. The model also treats phase exits as constant-rate events; if rates change over time or depend on cell state, the reciprocal-rate calculation needs revision.

Worked Example: Converting Mean Phase Durations to Exit Rates

Suppose a hypothetical cell-cycle model assigns mean durations of 8 hours in G1, 6 hours in S, 4 hours in G2, and 2 hours in M. Under the constant-hazard model, calculate the exit rate for each phase and the mean cycle duration.

For each phase, \(k_i=1/E[T_i]\). Therefore, the rates are \(k_{\mathrm{G1}}=1/8=0.1250\ \text{h}^{-1}\), \(k_{\mathrm{S}}=1/6=0.1667\ \text{h}^{-1}\), \(k_{\mathrm{G2}}=1/4=0.2500\ \text{h}^{-1}\), and \(k_{\mathrm{M}}=1/2=0.5000\ \text{h}^{-1}\).

The expected cycle duration is the sum of the four mean phase durations: \(8+6+4+2=20\) hours. Equivalently, using the rates, \(1/0.1250+1/0.1667+1/0.2500+1/0.5000\approx20.00\) hours, with the small discrepancy possible from rounding \(1/6\).

These values describe a hypothetical model, not an observed study. They specify mean durations and constant phase-exit hazards; they do not specify a fixed 20-hour schedule for each cell.

Compartment Equations and Division Flux

For phases before the final phase, cells enter from the preceding phase and leave at rate \(J_i=k_iC_i\). In the final phase, a division removes one parent from that compartment and sends two daughters into phase 1. The resulting balance equations are:

\frac{dC_1}{dt}=2J_m-J_1, \qquad \frac{dC_i}{dt}=J_{i-1}-J_i \quad (i=2,\ldots,m), \qquad J_i=k_iC_i.

The factor of two in the first equation counts daughter cells entering phase 1. By contrast, \(J_m\) counts completed division events, not daughter cells. Summing the phase equations cancels the internal transition fluxes. The total population \(N(t)=\sum_{i=1}^{m}C_i(t)\) therefore satisfies:

\frac{dN}{dt} =2J_m-J_1+\sum_{i=2}^{m}(J_{i-1}-J_i) =J_m.

Under these assumptions, each division adds one net cell: two daughters replace one parent. Consequently, the total cell count increases at the division-event flux. This accounting would change if cells died, left the modeled population, or produced a different number of daughters.

The phase fluxes need not be equal at a particular time. For an intermediate phase, \(dC_i/dt=J_{i-1}-J_i\). If its occupancy is changing, its inflow and outflow differ. The same logic applies to phase 1, where division supplies two daughters per division event. Equality of fluxes is appropriate only under additional conditions—for example, a non-branching, one-pass sequence with constant compartment occupancies—not as a general property of a growing cell-cycle population.

Worked Example: Calculating Instantaneous Phase and Division Fluxes

Consider a hypothetical population with \(C_{\mathrm{G1}}=1200\), \(C_{\mathrm{S}}=600\), \(C_{\mathrm{G2}}=300\), and \(C_{\mathrm{M}}=100\) cells. Use the rates from the previous example: \(0.1250\), \(0.1667\), \(0.2500\), and \(0.5000\ \text{h}^{-1}\), respectively. Calculate the phase fluxes, the instantaneous change in each compartment, and the total cell-count change.

The phase-exit fluxes are rate multiplied by phase count: \(J_{\mathrm{G1}}=0.1250(1200)=150\) cells/h, \(J_{\mathrm{S}}=0.1667(600)\approx100\) cells/h, \(J_{\mathrm{G2}}=0.2500(300)=75\) cells/h, and \(J_{\mathrm{M}}=0.5000(100)=50\) divisions/h.

The compartment equations give \(dC_{\mathrm{G1}}/dt=2(50)-150=-50\) cells/h, \(dC_{\mathrm{S}}/dt=150-100=50\) cells/h, \(dC_{\mathrm{G2}}/dt=100-75=25\) cells/h, and \(dC_{\mathrm{M}}/dt=75-50=25\) cells/h.

Adding these changes gives \(-50+50+25+25=50\) cells/h, which equals the division flux of 50 divisions/h because each division adds one net cell. The unequal phase fluxes are consistent with compartments changing in size at this instant; they do not violate cell accounting.

Interpreting Changes in Phase Duration

A phase-specific change in \(k_i\) changes the mean time spent in that phase. If the other rates remain fixed, it also changes the mean time to complete the sequence. This provides a direct way to express a model assumption about a perturbation: for example, a reduced exit rate in S phase implies a longer mean S-phase duration. It does not, by itself, establish why the rate changed or prove a biological mechanism.

Worked Example: A Reduced G1 Exit Rate

Use the hypothetical phase durations of 8, 6, 4, and 2 hours from the first example. Suppose a model scenario halves the G1 exit rate from \(0.1250\ \text{h}^{-1}\) to \(0.06250\ \text{h}^{-1}\), while leaving the other phase rates unchanged. Calculate the new mean G1 duration and mean cycle duration.

The new mean G1 duration is the reciprocal rate: \(1/0.06250=16.00\) hours. The other mean durations remain 6, 4, and 2 hours, so the new mean cycle duration is \(16+6+4+2=28.00\) hours, compared with 20.00 hours initially. The modeled mean has increased by 8.00 hours, or \(8/20=0.4000\), a 40.00% increase.

This calculation compares expected durations under the assumed rates. It does not determine the division flux without phase counts, because the flux from the final phase is \(k_{\mathrm{M}}C_{\mathrm{M}}\), not simply the reciprocal of the total mean cycle duration multiplied by an unspecified population size.

Diagnostics, Alternatives, and Limitations

When using this model, check that every exit rate is nonnegative and has units consistent with the time scale. Check that each flux has units of cells per time and that every daughter-producing transition is counted correctly. A useful accounting check is to sum the compartment equations: with no loss other than division, the total cell-count derivative must equal the number of division events per time. If it instead equals twice the division flux, daughter production may have been confused with net population increase.

The constant-hazard assumption also deserves scrutiny. If observed phase durations are tightly concentrated around a fixed value, an exponential waiting-time assumption may be a poor description. If cells pause, die, leave the cycle, or enter from outside the modeled population, the basic sequence omits relevant pathways. Time-varying conditions can make a single constant \(k_i\) inadequate. In those situations, model predictions should not be interpreted as if the omitted processes were absent in reality.

An aggregate model with one cycling compartment can be simpler when only total population change matters, but it cannot distinguish phase occupancy or phase-specific delays. A fixed-duration or delay-based representation can encode a more scheduled progression, but it does not use the same memoryless phase-exit assumption. The compartment model is useful when phase-level counts or transition mechanisms matter and a constant-rate approximation is defensible; it is not universally preferable to these alternatives.

A common mistake is to assume that phase counts multiplied by their exit rates must give the same flux in every phase. That equality would require the relevant compartment sizes to be constant in a one-cell-in/one-cell-out sequence. In the division model, phase occupancies can change, and completion of M phase supplies two cells to phase 1. Use the balance equations to relate inflow and outflow rather than imposing equal fluxes.

Key takeaway: In a sequential cell-cycle compartment model with constant phase-exit hazards, the mean duration of phase \(i\) is \(1/k_i\), the mean cycle duration is the sum of phase means, and the division flux is \(k_mC_m\). With two daughters entering phase 1 and no other losses, the total population grows at one net cell per division event.

Check Your Understanding

Answer each question, then compare your reasoning with the solutions.

A phase has a constant exit rate of \(0.20\ \text{h}^{-1}\). What is its mean duration?
A hypothetical three-phase sequence has mean durations of 5, 3, and 2 hours. What is its mean cycle duration under the sequential model?
If \(C_m=80\) cells and \(k_m=0.25\ \text{h}^{-1}\), what is the division flux? How many daughter cells enter phase 1 per hour?
In a two-phase model, the first phase has inflow 90 cells/h and outflow 70 cells/h. Is its occupancy increasing or decreasing, and at what rate?
Why is it generally incorrect to set all phase fluxes equal in a growing population with division?

Solutions. (1) The mean duration is \(1/0.20=5.00\) hours. (2) The mean cycle duration is \(5+3+2=10.00\) hours. (3) The division flux is \(J_m=k_mC_m=0.25(80)=20\) divisions/h; two daughters per division means 40 daughter cells enter phase 1 per hour. (4) The phase changes at inflow minus outflow, \(90-70=20\) cells/h, so its occupancy is increasing at 20 cells/h. (5) A phase with changing occupancy has unequal inflow and outflow, and the division event supplies two daughter cells to phase 1. The balance equations, rather than an assumption of equal flux, determine the relationships.