Why Represent Maturation as a Chain?
A drug can change biomarker synthesis at one point in a biological pathway, while the measured biomarker changes only after material passes through several maturation or release stages. A model with only one production-and-loss compartment may then predict a response that is too immediate. Transit compartments represent the sequence of intermediate states between formation and measured output.
The key modeling choice is what the compartments mean. If each state corresponds to an identifiable biological pool, its amount and transfer rate can be interpreted as properties of that pool. If the intermediate states are not directly observed and have no separately supported biological identity, they may instead be latent mathematical stages used to reproduce a delayed response. The same equations can serve either purpose, but the biological claims are different.
Mass Balance and Transfer Rates
Let \(A_i(t)\) be the amount in stage \(i\), measured in nmol, for \(i=1,\ldots,m\). Let \(R(t)\) be the input rate into the first stage, in nmol/h. The transfer rate constant from stage \(i\) is \(k_i\), in h\(^{-1}\). Material leaves stage \(i\) at flux \(k_iA_i\), in nmol/h. After the last transit stage, material enters the measured biomarker pool \(B(t)\), which is removed at first-order rate constant \(k_{\mathrm{out}}\). The mass-balance equations are
These equations assume that transfer is irreversible, each stage is well mixed, transfer from a stage is first-order, and there is no loss or additional input within the transit stages. The measured pool has its own first-order loss. The transfer rates and input may be constant or may vary with time, but the equations must then use the appropriate time-varying values. If the measured outcome is concentration rather than amount, a fixed volume \(V\) gives \(C(t)=B(t)/V\).
The first-order flux \(k_iA_i\) is not itself the transfer rate constant: the flux has units of amount per time, while \(k_i\) has units of inverse time. Distinct \(k_i\) values are allowed. Setting all rates equal is a simplifying assumption, not a requirement of the chain structure.
To justify the result, set each derivative to zero. The first equation gives \(k_1A_1^*=R_0\). The second gives \(k_iA_i^*=k_{i-1}A_{i-1}^*\). Since the first-stage outgoing flux is \(R_0\), this equality carries the same flux through every stage. Thus \(k_iA_i^*=R_0\) for all \(i\); the last equation then gives \(k_{\mathrm{out}}B^*=R_0\). Dividing by the positive rate constants gives the stated amounts.
Worked Example: Calculating a Chain's Baseline
Suppose a hypothetical biomarker is formed at a constant rate of \(12\) nmol/h and passes through three transit stages. Let \(k_1=0.50\) h\(^{-1}\), \(k_2=0.25\) h\(^{-1}\), \(k_3=0.40\) h\(^{-1}\), and \(k_{\mathrm{out}}=0.20\) h\(^{-1}\). Calculate the steady-state amounts. Assume the measured pool has volume \(V=2.0\) L.
The steady-state equations give \(A_1^*=R_0/k_1=12/0.50=24\) nmol, \(A_2^*=12/0.25=48\) nmol, and \(A_3^*=12/0.40=30\) nmol. The measured-pool amount is \(B^*=R_0/k_{\mathrm{out}}=12/0.20=60\) nmol.
Checking the fluxes verifies the calculation: \(0.50(24)=12\) nmol/h, \(0.25(48)=12\) nmol/h, \(0.40(30)=12\) nmol/h, and \(0.20(60)=12\) nmol/h. The measured concentration is \(C^*=B^*/V=60/2.0=30\) nmol/L. The different stage amounts are consistent with equal flux because the transfer rate constants differ.
What a Change in Input Does
A chain creates a structured lag. If the system is initially at steady state and the input changes abruptly, the first stage immediately experiences the input change. Downstream stages do not all jump to new steady states: their amounts remain continuous, and their rates of change depend on the incoming and outgoing fluxes at that moment.
For example, suppose the input changes from \(R_0\) to \(R_1\) at time zero while all amounts are initially at the old steady state. For the first stage, \(\left.dA_1/dt\right|_{0^+}=R_1-k_1A_1^*=R_1-R_0\). For each later stage, the initial derivative is zero because its incoming and outgoing fluxes still equal \(R_0\). The measured pool also has initial derivative zero for the same reason. This establishes the initial propagation of a step change directly from the mass-balance equations; it does not mean downstream states will remain unchanged after time zero.
Worked Example: A Sudden Reduction in Biomarker Formation
Use the three-stage chain from the previous example. At baseline, the formation rate is \(12\) nmol/h and the amounts are \(A_1=24\), \(A_2=48\), \(A_3=30\), and \(B=60\) nmol. Suppose a hypothetical drug reduces formation immediately to \(6\) nmol/h, without changing any transfer or output rate. Find the initial rates of change and the new steady state.
Immediately after the input change, \(dA_1/dt=6-(0.50)(24)=-6\) nmol/h. For the second stage, \(dA_2/dt=(0.50)(24)-(0.25)(48)=12-12=0\) nmol/h. For the third stage, \(dA_3/dt=(0.25)(48)-(0.40)(30)=12-12=0\) nmol/h. Finally, \(dB/dt=(0.40)(30)-(0.20)(60)=12-12=0\) nmol/h. Thus the first-stage amount begins falling immediately; the downstream amounts and measured pool have zero initial slope, although their slopes can change later as upstream fluxes change.
At the new steady state, the formation rate is \(6\) nmol/h. Therefore, \(A_1^*=6/0.50=12\) nmol, \(A_2^*=6/0.25=24\) nmol, \(A_3^*=6/0.40=15\) nmol, and \(B^*=6/0.20=30\) nmol. For \(V=2.0\) L, the new steady-state concentration is \(30/2.0=15\) nmol/L. This is half the baseline concentration, but the calculation does not imply that the measured concentration reaches this value immediately.
Biological Stages or Empirical Delay States?
A transit-chain model can be mechanistic at different levels. When stages correspond to known precursor, maturation, or release pools, the \(A_i\) may represent biological amounts and the \(k_i\) may represent their transfer rates. Those interpretations require evidence about the pathway and compatible units, and can be strengthened by measurements of intermediate pools or independent knowledge of transfer processes.
In an empirical transit model, intermediate states are latent: they help describe how a change in input reaches the measured biomarker over time, but they are not asserted to be separately observed pools. A good fit to the output time course alone does not establish that the proposed stages exist biologically. The earlier material on sequential precursor maturation and quasi-steady-state reduction addresses explicit precursor pools and approximations that eliminate fast states; a transit chain should not be treated as proof of either biological structure or a particular approximation.
Worked Example: A Drug Changes a Transfer Rate
Consider a hypothetical two-stage chain with constant formation \(R_0=8\) nmol/h, \(k_1=0.40\) h\(^{-1}\), \(k_2=0.20\) h\(^{-1}\), \(k_{\mathrm{out}}=0.10\) h\(^{-1}\), and measured-pool volume \(V=4.0\) L. The chain is at baseline steady state when a drug instantaneously reduces \(k_2\) to \(0.10\) h\(^{-1}\), leaving formation and other rates unchanged. Calculate the initial changes and the new steady state.
Before the change, \(A_1^*=8/0.40=20\) nmol, \(A_2^*=8/0.20=40\) nmol, and \(B^*=8/0.10=80\) nmol. Immediately after the change, the incoming flux to stage 2 is still \((0.40)(20)=8\) nmol/h, while its outgoing flux is now \((0.10)(40)=4\) nmol/h. Hence \(dA_2/dt=8-4=4\) nmol/h. The measured pool initially has \(dB/dt=(0.10)(40)-(0.10)(80)=4-8=-4\) nmol/h.
At the new steady state, \(A_1^*=8/0.40=20\) nmol, \(A_2^*=8/0.10=80\) nmol, and \(B^*=8/0.10=80\) nmol. The steady-state measured concentration remains \(80/4.0=20\) nmol/L, even though stage 2 holds twice as much material. Under the stated no-loss assumptions, changing an internal transfer rate changes the amount required to carry the steady flux; it does not change that flux or the eventual output amount. It does change the time course, as the initial negative slope of \(B\) shows.
Assumptions, Diagnostics, and Alternatives
Transit compartments are useful when the timing of a biomarker response matters: the model separates the site of formation from the later measured output and allows each stage to have its own transfer rate. They are especially interpretable when the proposed pathway is supported by biology. Before fitting or interpreting one, check whether the assumed sequence, irreversibility, first-order transfer, and absence of intermediate loss are plausible for the application.
Signs of trouble include systematic residual patterns around the onset or recovery of a response, predictions that miss observed intermediate-pool measurements, or fitted rates that imply implausible transfer times under a biological interpretation. A chain with no intermediate losses also predicts equal steady-state flux throughout. If the observed system has branching, accumulation with degradation, reversible exchange, or changing volume, the simple equations do not represent that process without modification. Model fit alone cannot resolve whether an unobserved transit state is a real pool.
| Approach | What it represents | Important distinction |
|---|---|---|
| Transit-compartment chain | Sequential first-order states between formation and measured output | Intermediate states may be biological pools or latent delay states; the interpretation must be justified separately. |
| Single turnover compartment | One measured pool with input and loss | Does not explicitly represent multiple sequential maturation stages. |
| Explicit precursor-pool model | Named biochemical pools with specified formation, conversion, or loss | Requires a defensible mass-balance structure for those pools; it is not merely a relabeling of latent transit states. |
A common mistake is to interpret every transit compartment as a measured biological pool because its equation resembles a compartment mass balance. The correction is to distinguish the mathematical role from the biological claim: the equations specify sequential states, while evidence determines whether those states correspond to actual pools. Another frequent error is to assume that slowing an internal transfer must lower the eventual output at steady state. In a chain with constant input and no intermediate loss, the steady flux remains the input rate; the transfer rate changes the amount held in a stage and the transient response.
Check Your Understanding
Answer each question, then compare with the solutions.
- A two-stage chain has \(R_0=10\) nmol/h, \(k_1=0.50\) h\(^{-1}\), \(k_2=0.25\) h\(^{-1}\), and \(k_{\mathrm{out}}=0.20\) h\(^{-1}\). Find all three steady-state amounts.
- At the old steady state, formation falls abruptly from \(R_0\) to \(R_1\). Which compartment has a nonzero initial derivative under the simple chain assumptions?
- If the transfer rate out of an intermediate stage decreases while input and all other rates remain fixed, must the eventual steady-state output amount decrease? State the assumptions behind your answer.
- What evidence would be needed to describe a latent transit state as a biological maturation pool?
Solutions. (1) \(A_1^*=10/0.50=20\) nmol, \(A_2^*=10/0.25=40\) nmol, and \(B^*=10/0.20=50\) nmol. Each outgoing flux is \(10\) nmol/h. (2) Only the first stage has a nonzero initial derivative: \(dA_1/dt=R_1-R_0\). Later stages and the measured pool initially have equal incoming and outgoing fluxes, so their initial derivatives are zero. (3) No. With constant input, positive rates, no intermediate loss, and the stated chain structure, the steady output amount is \(B^*=R_0/k_{\mathrm{out}}\); reducing an internal transfer rate changes the relevant stage amount, not the steady flux. (4) The proposed stage should be supported by pathway knowledge and, where available, compatible measurements of the intermediate pool or independent evidence about its transfer process. A good fit to the output curve alone is not sufficient to establish that biological identity.