Scaling Transporter Activity From In Vitro to Liver

Why Scaling Requires More Than a Protein Correction

A transporter assay may report uptake as a rate per milligram of vesicle or membrane protein. A liver model, by contrast, needs a capacity for the transporter in the whole organ, or a capacity per gram of liver. These are not interchangeable quantities: the assay protein contains a particular amount of transporter, and the liver contains a different amount. Scaling therefore requires an explicit bridge between the transporter measured in vitro and the transporter represented in the liver.

The central calculation is an abundance-adjusted activity multiplied by the amount of transporter in liver and, for a whole-organ estimate, liver mass. A recovery correction may also be needed if the assay preparation retains only a known fraction of the transporter activity. The correction must match how activity and abundance were measured; applying it automatically can introduce a bias rather than remove one.

Definition: Let $V_{\max,\mathrm{obs}}$ be the observed in vitro maximum transport rate, $Q_{\mathrm{assay}}$ the corresponding nominal transporter abundance in the assay, $Q_{\mathrm{liver}}$ the transporter abundance per gram of liver, $R_{\mathrm{rec}}$ the fraction of transporter activity recovered in the assay preparation, and $W_{\mathrm{liver}}$ liver mass. If $V_{\max,\mathrm{obs}}$ has not already been corrected for recovery, the recovery-corrected whole-liver capacity is $$ V_{\max,\mathrm{liver}} = \frac{V_{\max,\mathrm{obs}}}{R_{\mathrm{rec}}} \frac{Q_{\mathrm{liver}}}{Q_{\mathrm{assay}}} W_{\mathrm{liver}}. $$ The abundance and protein units in $Q_{\mathrm{liver}}$ and $Q_{\mathrm{assay}}$ must be compatible with the denominator used for $V_{\max,\mathrm{obs}}$.

For example, if $V_{\max,\mathrm{obs}}$ is measured in nmol/min/mg assay protein and $Q_{\mathrm{assay}}$ is measured in pmol transporter/mg assay protein, then $Q_{\mathrm{liver}}$ must be in pmol transporter/g liver. The assay-protein units cancel in the abundance ratio, leaving the rate per gram of liver before multiplication by liver mass. When transporter abundance in liver is reported per milligram of liver protein instead, convert it to abundance per gram of liver using microsomal or membrane protein per gram of liver, as appropriate to the assay and abundance measurement.

Deriving the Scaling Relationship

The relationship follows by expressing the assay rate per unit of transporter. Suppose the assay rate and nominal transporter abundance are both referenced to the same amount of assay protein. If only a fraction $R_{\mathrm{rec}}$ of the relevant activity is recovered, the estimated activity per transporter is the observed rate divided by both the recovery fraction and the nominal abundance. Multiplying that activity by the liver transporter abundance gives the rate per gram of liver.

\begin{aligned} a_{\mathrm{transporter}} &= \frac{V_{\max,\mathrm{obs}}} {R_{\mathrm{rec}}Q_{\mathrm{assay}}},\\ V_{\max,\mathrm{liver\ per\ g}} &= a_{\mathrm{transporter}}Q_{\mathrm{liver}},\\ V_{\max,\mathrm{liver}} &= a_{\mathrm{transporter}}Q_{\mathrm{liver}}W_{\mathrm{liver}}. \end{aligned}

Here $a_{\mathrm{transporter}}$ is an activity per amount of transporter. The units provide a direct check. With $V_{\max,\mathrm{obs}}$ in nmol/min/mg, $Q_{\mathrm{assay}}$ in pmol/mg, and $Q_{\mathrm{liver}}$ in pmol/g, the assay-protein units cancel, and transporter units cancel when the activity is multiplied by liver abundance. The result is nmol/min/g liver. Multiplication by $W_{\mathrm{liver}}$ in g gives nmol/min.

If $V_{\max,\mathrm{obs}}$ and $Q_{\mathrm{assay}}$ are both measured in the prepared vesicles after processing, their ratio may already describe activity per recovered transporter. In that case, dividing by $R_{\mathrm{rec}}$ again would double-correct recovery. Conversely, if the reported $Q_{\mathrm{assay}}$ is the nominal transporter amount before a preparation step that loses activity, and $V_{\max,\mathrm{obs}}$ reflects the remaining activity, the recovery correction may be appropriate. The measurement definitions determine which form to use.

Worked Example: Scaling Vesicle Activity to Whole Liver

Consider hypothetical vesicle data for one uptake transporter. The observed $V_{\max,\mathrm{obs}}$ is 0.024 nmol/min/mg assay protein, $Q_{\mathrm{assay}}$ is 40 pmol/mg assay protein, and the estimated activity recovery is $R_{\mathrm{rec}}=0.80$. Suppose liver abundance is $Q_{\mathrm{liver}}=12$ pmol/g liver and liver mass is 1,500 g. Assume the activity and assay abundance use the nominal pre-preparation protein basis, so the recovery correction is required.

First calculate activity per transporter:

a_{\mathrm{transporter}} = \frac{0.024\ \mathrm{nmol/min/mg}} {0.80 \times 40\ \mathrm{pmol/mg}} = 0.000750\ \mathrm{nmol/min/pmol}.

The recovered activity is corrected by dividing by 0.80; the transporter abundance is 40 pmol/mg on the stated nominal basis. The assay-protein units cancel. Next multiply by liver abundance:

V_{\max,\mathrm{liver\ per\ g}} = 0.000750\ \mathrm{nmol/min/pmol} \times 12\ \mathrm{pmol/g} = 0.00900\ \mathrm{nmol/min/g}.

Finally, scale from liver mass to whole liver:

V_{\max,\mathrm{liver}} = 0.00900\ \mathrm{nmol/min/g} \times 1{,}500\ \mathrm{g} = 13.5\ \mathrm{nmol/min}.

Under the stated assumptions, the estimated whole-liver transporter capacity is 13.5 nmol/min. This is a capacity, not a prediction that transport always proceeds at that rate; the actual rate depends on substrate concentration and the applicability of the kinetic model.

From Capacity to Low-Concentration Intrinsic Clearance

When transport follows Michaelis–Menten kinetics, its rate at concentration $C$ is $V_{\max}C/(K_m+C)$. If the same $K_m$ applies to the in vitro transporter and the scaled liver transporter, abundance scaling changes $V_{\max}$ but does not itself change $K_m$. At concentrations much smaller than $K_m$, the rate is approximately $(V_{\max}/K_m)C$. Thus, the low-concentration intrinsic clearance associated with this transporter is $V_{\max}/K_m$, after using consistent concentration and volume units.

Worked Example: Converting Scaled Capacity to Low-Concentration Clearance

Use the hypothetical whole-liver capacity of 13.5 nmol/min calculated above. Suppose the transporter’s $K_m$ is 5.00 μM, equivalent to 5.00 nmol/mL. For concentrations well below $K_m$, the low-concentration clearance is:

\frac{V_{\max,\mathrm{liver}}}{K_m} = \frac{13.5\ \mathrm{nmol/min}} {5.00\ \mathrm{nmol/mL}} = 2.70\ \mathrm{mL/min}.

The units reduce to volume per time. This value describes the linear, low-concentration limit of the specified transporter pathway under the model assumptions. It is not the rate at an arbitrary concentration. For instance, at $C=2.00$ μM, the Michaelis–Menten rate would instead be:

\frac{13.5\ \mathrm{nmol/min}\times 2.00\ \mathrm{\mu M}} {5.00\ \mathrm{\mu M}+2.00\ \mathrm{\mu M}} = 3.86\ \mathrm{nmol/min}.

The concentration units cancel in this rate calculation. Because 2.00 μM is not negligible relative to $K_m=5.00$ μM, the low-concentration approximation would not be exact here.

Tracking Uncertainty in the Scaling Factors

A scaled estimate inherits uncertainty from the measured assay activity, transporter abundances, recovery, and liver mass. For the recovery-corrected expression, the estimate is proportional to $V_{\max,\mathrm{obs}}$, $Q_{\mathrm{liver}}$, and $W_{\mathrm{liver}}$, and inversely proportional to $R_{\mathrm{rec}}$ and $Q_{\mathrm{assay}}$. Therefore, a relative error in a denominator can shift the result in the opposite direction: overestimating assay abundance, for example, lowers the scaled capacity.

For small uncertainties, if the inputs are independent, a first-order approximation to the relative standard uncertainty is the square root of the sum of squared relative standard uncertainties. This approximation is not a confidence interval, and the independence assumption matters. Correlated measurements require covariance terms or a joint uncertainty analysis.

Worked Example: Propagating Relative Uncertainty

Suppose a hypothetical calculation uses $V_{\max,\mathrm{obs}}=0.030$ nmol/min/mg with a 10% relative standard uncertainty; $Q_{\mathrm{liver}}=10.0$ pmol/g with 20%; $Q_{\mathrm{assay}}=50.0$ pmol/mg with 15%; and $R_{\mathrm{rec}}=0.750$ with an absolute standard uncertainty of 0.050. Let $W_{\mathrm{liver}}=1{,}500$ g be treated as fixed. Assume the uncertain inputs are independent.

The central whole-liver capacity is:

V_{\max,\mathrm{liver}} = \frac{0.030\times10.0\times1{,}500} {0.750\times50.0} = 12.0\ \mathrm{nmol/min}.

The recovery’s relative standard uncertainty is $0.050/0.750=0.0667$, or 6.67%. The first-order relative standard uncertainty in the scaled estimate is therefore:

\sqrt{0.10^2+0.20^2+0.15^2+0.0667^2} = 0.277,

or 27.7%. The corresponding approximate standard uncertainty is $12.0\times0.277=3.32$ nmol/min. This is a one-standard-uncertainty summary under the stated approximation, not a 95% interval. In particular, the calculation shows that the liver abundance uncertainty is a substantial contributor; collecting a more precise assay $V_{\max}$ alone may not greatly reduce uncertainty in the whole-liver estimate.

Choosing and Checking a Scaling Approach

There are two common ways to perform the conversion. One can normalize assay activity by transporter abundance to obtain activity per transporter, then multiply by liver abundance. Alternatively, one can use a direct abundance ratio, multiplying assay activity per assay protein by liver transporter abundance per gram and dividing by assay transporter abundance per assay protein. These are algebraically equivalent when they use the same abundance basis and recovery convention. A protein-per-gram-liver factor is useful when liver transporter abundance is reported per milligram of liver protein: multiply that abundance by the relevant milligrams of protein per gram of liver before scaling to the organ.

A generic protein scaling factor alone is not a substitute for transporter abundance scaling. It assumes, often without justification, that the transporter content per unit protein is the same in the assay preparation and liver. Direct abundance measurements make that assumption testable. Likewise, a relative activity factor estimated from functional data can be useful when direct abundance data are unavailable, but it may combine transporter expression, recovery, orientation, and assay-specific effects. Its interpretation is less specific unless those components have been characterized.

Before using the result, check that the assay and liver abundance refer to the same transporter, relevant membrane compartment, and compatible molecular or activity basis. Confirm that the substrate and assay conditions support the kinetic parameter being scaled. If activity and transporter abundance are measured in different preparations, verify whether recovery has already entered either measurement. Check the dimensions at every step and ensure liver mass is expressed in the units paired with abundance per gram.

A particularly common failure is to multiply activity per milligram of assay protein by milligrams of liver protein per gram and assume the result is transporter-specific. That calculation scales total protein, not the target transporter; it is valid only under an explicit assumption that transporter abundance per unit protein is matched. Another failure is to divide by recovery when activity has already been normalized to the recovered transporter. That double correction inflates the estimate. Sensitivity analysis is useful: recompute the scaled capacity over plausible values for abundance and recovery to see which assumptions dominate the result.

The scaled value remains an estimate for the specified transporter pathway, not proof that the transporter is the only hepatic uptake route or that its in vitro activity is fully preserved in vivo. Vesicle orientation, membrane environment, endogenous cofactors, substrate access, and assay conditions can limit transferability. A reported capacity should therefore retain its units, scaling inputs, recovery definition, and uncertainty, so that its physiological interpretation is not stronger than the evidence allows.

Key takeaway: Scale transporter activity through an explicit activity-per-transporter quantity, then apply liver transporter abundance and liver size. Correct for recovery only when the activity and abundance measurements have not already incorporated that correction.

Check Your Understanding

Use the definitions and unit logic above to answer each question. Solutions are included below the questions.

An assay reports 0.020 nmol/min/mg protein, and assay transporter abundance is 25 pmol/mg on the same nominal protein basis. What is the uncorrected activity per pmol transporter?
If liver abundance is reported per milligram of liver protein, what additional information is needed to express it per gram of liver?
An assay activity and transporter abundance are both measured in the same recovered vesicles. Should a recovery correction automatically be applied to their ratio? Explain.
Why does multiplying $V_{\max}$ by liver abundance change capacity but not, by itself, change $K_m$?

Solutions. (1) Divide the activity by abundance: $0.020/25=0.000800$ nmol/min/pmol transporter; the mg protein units cancel. (2) The relevant amount of liver protein per gram of liver is needed, with a basis compatible with the abundance measurement. Multiplying the abundance per mg protein by that factor gives abundance per gram liver. (3) No. If both quantities refer to the same recovered vesicles, their ratio may already be activity per recovered transporter. A further division by recovery could double-correct; the assay definitions must be checked. (4) Abundance scaling multiplies the number of transporter units represented, so it scales the maximum rate. $K_m$ is the concentration parameter of the kinetic relation and is not altered by multiplying the rate by transporter amount, provided the same kinetic behavior applies.