Deriving Vector Genome Clearance and Transgene Expression Kinetics

From Vector Entry to Measurable Expression

A vector dose does not become transgene product instantaneously. Vector material must first reach a cell and be taken up; intracellular genomes may then be lost, persist, or remain available for expression. Product accumulates and disappears according to its own synthesis and turnover. A useful kinetic model keeps these processes separate so that a delayed or short-lived expression profile is not mistaken for a single “transduction rate.”

The earlier material on vector dose, transduction, and functional copy number motivates representing successful entry as an initial or time-varying input to a cell. Here the focus is what happens kinetically after that input. The model describes a representative target cell or a homogeneous cell population; it does not resolve differences among tissues or cell types.

These states are effective model quantities, not claims that every measured intracellular genome is intact, nuclear, or transcriptionally active. A model using \(G(t)\) as an expression-capable pool assumes that its loss rate summarizes all processes that remove genomes from that pool. The simplest model also assumes constant rate coefficients, well-mixed states, no cell division or changing cell number over the modeled period, and no saturation of uptake or expression. These assumptions make the equations interpretable; data showing nonlinear uptake, multiple genome states, or changing cell populations would require a more detailed structure.

Uptake and Intracellular Genome Loss

Suppose accessible vector declines through two competing first-order processes: uptake into cells at rate \(k_u C(t)\), and loss from the accessible pool at rate \(k_c C(t)\). The intracellular genome pool is supplied by uptake and depleted by intracellular loss. With \(C(0)=C_0\) and \(G(0)=0\), the model is

Define \(a=k_u+k_c\). The accessible pool is \(C(t)=C_0e^{-at}\). To solve for \(G(t)\), multiply its equation by the integrating factor \(e^{k_gt}\). Then

For \(k_g\ne a\), integrate from 0 to \(t\), using \(G(0)=0\), and rearrange:

If \(k_g=a\), the integral instead gives \(G(t)=k_uC_0t e^{-at}\). Both expressions are nonnegative for nonnegative rates: when \(k_g>a\), both numerator difference and denominator are positive; when \(k_g<a\), both are negative. The equal-rate expression is also the limit of the unequal-rate solution as \(k_g\) approaches \(a\).

The formula also shows why an intracellular pool may rise before it falls. Initially, uptake supplies genomes while the pool is small. Later, as accessible vector is depleted, the supply term shrinks; genome loss can then dominate. Thus the time of maximum \(G(t)\) is not generally the same as the time at which the accessible vector is exhausted.

Genome-Driven Product with a Single Expression State

A compact expression model treats transcription and translation together as a net product synthesis rate per intracellular genome. Let \(k_s\) be that rate, in product units per genome per day. With product turnover at rate \(k_pP\), the equations are

This expression solution assumes no further uptake after the starting time, so \(G(0)=G_0\), and assumes \(P(0)=0\). The genome solution is \(G(t)=G_0e^{-k_gt}\). Applying an integrating factor to the product equation gives, for \(k_g\ne k_p\),

When \(k_g=k_p=k\), the continuous limiting solution is \(P(t)=k_sG_0t e^{-kt}\). The product initially rises because genome-driven synthesis exceeds product loss, then falls as the genome source diminishes. For unequal positive rates, the peak occurs at \(t=\ln(k_p/k_g)/(k_p-k_g)\), obtained by setting \(dP/dt=0\). In the equal-rate case, differentiating \(t e^{-kt}\) gives a peak at \(t=1/k\).

For unequal rates, the peak time depends on both genome disappearance and product turnover. For example, fast product turnover can make the product peak while genomes are still present; slow turnover can sustain product after genome abundance has fallen. A persistent product signal therefore does not by itself imply persistent intracellular genomes.

Making Transcription and Translation Explicit

The lumped \(k_s\) model is useful when only the overall genome-to-product relationship is needed. If mRNA measurements are available, or the delay between transcription and product accumulation matters, represent mRNA and protein as separate states. With no further uptake and initial conditions \(G(0)=G_0\), \(M(0)=P(0)=0\), use

The genome equation gives \(G(t)=G_0e^{-k_gt}\). Solving the mRNA equation by an integrating factor gives, when \(k_m\ne k_g\),

Product is driven by the time-varying mRNA pool. Substituting this \(M(t)\) into the integrating-factor solution for \(P\) yields, when the displayed denominators are nonzero,

If any two rates coincide, the corresponding quotient is evaluated by its limit; for example, \((e^{-k_at}-e^{-k_bt})/(k_b-k_a)\) tends to \(t e^{-kt}\) as both rates tend to \(k\). The expression for \(P(t)\) makes explicit that product kinetics reflect the successive delays and losses in genome, mRNA, and product states.

Diagnostics, Alternatives, and Limitations

A fitted curve should be checked against the model structure, not only against its overall visual fit. With the simple uptake model, an inferred negative genome amount or a fitted curve that rises indefinitely after accessible vector is gone contradicts the equations. In expression data, a measured product peak before any plausible transcriptional delay, a plateau despite substantial modeled source depletion, or systematic residual patterns across time may signal missing states or incorrect rate assumptions. Sampling only late time points can also make uptake, genome loss, and product turnover difficult to distinguish.

The alternatives answer different questions. A direct genome-to-product model is parsimonious when mRNA is unobserved and the objective is the overall expression profile. The explicit mRNA model is preferable when transcriptional delay or mRNA turnover is of interest, but it adds parameters that may not be identifiable from product data alone. A time-varying uptake input can replace the decaying accessible pool when the input profile is known or measured. Saturable uptake, multiple intracellular genome states, cell proliferation, or heterogeneous rates require additional model structure and data to estimate it. These equations do not establish that vector genomes cause a measured product change; they encode a mechanistic hypothesis whose assumptions must be evaluated.

The main modeling benefit is separation of input, source, and response. Uptake controls how genomes enter the intracellular pool; \(k_g\) controls how long that source remains; transcription, translation, and product turnover determine how the source is converted into an observed expression profile. Distinguishing these processes clarifies which parameter a particular observation can inform and which mechanisms remain confounded.