Insulin-Independent Glucose Uptake and Utilization

Why Separate Insulin-Independent Uptake?

A glucose mass balance needs to account for tissues that remove glucose even when insulin is low, as well as peripheral disposal that changes with insulin. Brain and erythrocytes are important examples of the first category. In a QSP model, combining these processes into one insulin-responsive term can misattribute basal glucose use to insulin action. Conversely, treating all peripheral uptake as insulin-independent can hide an insulin-driven response.

Here, insulin-independent describes the transport or uptake component represented in the model: its rate has no acute insulin-dependent term. It does not mean that insulin can have no indirect or longer-term influence on a tissue’s metabolism. The earlier material on organ-resolved glucose mass balances provides the accounting framework; the earlier material on insulin-driven GLUT4 trafficking distinguishes the insulin-regulated peripheral component.

Definition: An insulin-independent glucose uptake flux is a rate of glucose removal from the circulating glucose pool that is not directly multiplied or regulated by the model’s insulin-dependent transport term. Denote such a tissue flux by \(R_i(C)\), where \(C\) is glucose concentration and \(R_i\) is positive when glucose leaves the circulating pool. Brain and erythrocytes can be represented by separate fluxes, \(R_{\mathrm{brain}}(C)\) and \(R_{\mathrm{RBC}}(C)\).

The word uptake concerns movement into tissue; utilization concerns what happens to glucose after it enters. Those rates may be close in a simplified model, but they are not definitions of the same process. Brain cells metabolize glucose, while erythrocytes obtain energy through glycolysis and release lactate. A model that tracks only circulating glucose may represent the net tissue removal as one flux; a model of downstream metabolites may need to represent uptake and subsequent fate separately.

Modeling Basal Uptake

Let \(C\) denote the glucose concentration accessible to tissue transport, in mmol/L. A simple saturable uptake law is

R_i(C)=\frac{V_{\max,i}C}{K_{m,i}+C}.

Here, \(V_{\max,i}\) is the tissue’s maximum modeled uptake rate, in mmol/min, and \(K_{m,i}\) is the concentration at which the modeled rate is half of that maximum, in mmol/L. This is a phenomenological saturation law: it can represent finite transport capacity without requiring the model to represent every transporter state explicitly. Under this law, uptake depends on glucose concentration but not directly on insulin.

The law assumes that the tissue’s effective transport capacity is fixed over the time interval being modeled; that \(C\) is a suitable approximation to the concentration driving transport; and that the modeled glucose removal is nonnegative and saturates as concentration rises. It also omits explicit intracellular glucose, transporter trafficking, and feedback from downstream metabolism. Those omissions may be reasonable for a whole-body glucose balance, but not for every tissue-level question.

The limiting behavior explains when simpler forms may be adequate. At low concentration relative to \(K_{m,i}\), the denominator is approximately \(K_{m,i}\), so uptake is nearly proportional to concentration. At high concentration relative to \(K_{m,i}\), the flux approaches its maximum.

C \ll K_{m,i}: \quad R_i(C)\approx \frac{V_{\max,i}}{K_{m,i}}C, \qquad C \gg K_{m,i}: \quad R_i(C)\approx V_{\max,i}.

The low-concentration approximation is a linear clearance model, \(R_i(C)=CL_iC\), where \(CL_i\) has units L/min and is approximately \(V_{\max,i}/K_{m,i}\) in that range. A constant basal flux, \(R_i(C)=R_{0,i}\), is another possible simplification, but it assumes the rate stays fixed as glucose concentration changes. It should be used only when that approximation is acceptable over the modeled conditions; unlike the saturable law, it predicts positive uptake even at \(C=0\).

Worked Example: Calculating Separate Brain and Erythrocyte Fluxes

Suppose a hypothetical whole-body model uses the saturable law for two insulin-independent tissues. Let \(C=5.0\) mmol/L. For brain, take \(V_{\max,\mathrm{brain}}=0.80\) mmol/min and \(K_{m,\mathrm{brain}}=1.5\) mmol/L. For erythrocytes, take \(V_{\max,\mathrm{RBC}}=0.30\) mmol/min and \(K_{m,\mathrm{RBC}}=2.0\) mmol/L. These values are illustrative model parameters, not measurements from a real study.

For brain:

R_{\mathrm{brain}}(5.0) =\frac{0.80\ \mathrm{mmol/min}\times 5.0\ \mathrm{mmol/L}} {1.5\ \mathrm{mmol/L}+5.0\ \mathrm{mmol/L}} =\frac{4.0}{6.5}\ \mathrm{mmol/min} =0.6154\ \mathrm{mmol/min}.

For erythrocytes:

R_{\mathrm{RBC}}(5.0) =\frac{0.30\ \mathrm{mmol/min}\times 5.0\ \mathrm{mmol/L}} {2.0\ \mathrm{mmol/L}+5.0\ \mathrm{mmol/L}} =\frac{1.5}{7.0}\ \mathrm{mmol/min} =0.2143\ \mathrm{mmol/min}.

The combined insulin-independent removal in this two-tissue representation is \(0.6154+0.2143=0.8297\) mmol/min. The total is the sum of the two tissue rates, not a rate that should be assigned to either tissue individually.

Distinguishing Independent Uptake from Peripheral Disposal

The insulin-independent terms should remain distinct from insulin-regulated peripheral disposal in the model structure. For example, a whole-body removal term might be written as

R_{\mathrm{total}}(C,I) = R_{\mathrm{brain}}(C) + R_{\mathrm{RBC}}(C) + R_{\mathrm{peripheral}}(C,I) + R_{\mathrm{other}}(C,I),

where \(I\) denotes insulin concentration. In this accounting, a model can make \(R_{\mathrm{peripheral}}\) insulin-responsive while leaving the specified brain and erythrocyte transport terms independent of \(I\). The terms should be defined so that a tissue’s uptake is counted once: if a peripheral term already includes its basal component, do not add that same basal component again elsewhere.

A useful model check is to hold glucose concentration fixed and vary insulin. Under the stated assumption, \(R_{\mathrm{brain}}(C)\) and \(R_{\mathrm{RBC}}(C)\) do not change directly with that intervention, whereas the insulin-regulated peripheral term may change. If the independent terms change, check whether the equations accidentally include insulin, whether glucose concentration was actually held fixed, or whether an indirect effect has intentionally been added. This distinction is about model structure, not a claim that all tissue physiology is unaffected by insulin.

Worked Example: Estimating a Saturable Uptake Curve

Suppose two hypothetical measurements for one insulin-independent tissue are \(R=0.40\) mmol/min at \(C=2.0\) mmol/L and \(R=0.80\) mmol/min at \(C=8.0\) mmol/L. Assume the saturable law \(R(C)=V_{\max}C/(K_m+C)\) applies and both measurements are exact. Find \(V_{\max}\) and \(K_m\).

At \(C=2.0\) mmol/L:

0.40=\frac{2.0V_{\max}}{K_m+2.0} \quad\Longrightarrow\quad V_{\max}=0.20K_m+0.40.

At \(C=8.0\) mmol/L:

0.80=\frac{8.0V_{\max}}{K_m+8.0} \quad\Longrightarrow\quad V_{\max}=0.10K_m+0.80.

Equating these expressions gives \(0.20K_m+0.40=0.10K_m+0.80\), so \(K_m=4.0\) mmol/L. Substitution gives \(V_{\max}=0.20(4.0)+0.40=1.20\) mmol/min. The fitted curve predicts \(R(4.0)=1.20(4.0)/(4.0+4.0)=0.60\) mmol/min. With only two exact points, the two parameters are determined algebraically; real data usually require more observations and an assessment of uncertainty and model fit.

Putting the Fluxes into a Glucose Balance

Let \(G\) be the amount of glucose in a modeled circulating pool, in mmol, and let \(V\) be its effective distribution volume, in L. If \(C=G/V\) and \(V\) is constant, then \(dC/dt=(1/V)dG/dt\). With an aggregate glucose input \(R_{\mathrm{in}}\), the balance can include insulin-independent tissue uptake and peripheral disposal as separate removals:

\frac{dG}{dt} = R_{\mathrm{in}} - R_{\mathrm{brain}}(C) - R_{\mathrm{RBC}}(C) - R_{\mathrm{peripheral}}(C,I) - R_{\mathrm{other}}(C,I).

Every term on the right has units of amount per time, such as mmol/min. A positive removal lowers \(G\); it is therefore subtracted. This equation is bookkeeping: by itself, it does not specify the physiology of glucose input or prove that each removal term is identifiable from observed concentration data.

Worked Example: Checking a Whole-Body Mass Balance

Consider a hypothetical model with \(V=12.0\) L and \(C=5.5\) mmol/L. Suppose, at this instant, the glucose input is \(1.20\) mmol/min, brain uptake is \(0.64\) mmol/min, erythrocyte uptake is \(0.21\) mmol/min, insulin-regulated peripheral disposal is \(0.55\) mmol/min, and other removal is \(0.18\) mmol/min. Treat these rates as given model inputs for this calculation.

The total removal is \(0.64+0.21+0.55+0.18=1.58\) mmol/min. Thus:

\frac{dG}{dt}=1.20-1.58=-0.38\ \mathrm{mmol/min}.

At constant \(V\), the concentration rate of change is:

\frac{dC}{dt} = \frac{-0.38\ \mathrm{mmol/min}}{12.0\ \mathrm{L}} = -0.03167\ \mathrm{mmol\,L^{-1}\,min^{-1}}.

The negative sign means that glucose concentration is decreasing at that instant. The insulin-independent contribution is \(0.64+0.21=0.85\) mmol/min; it is not interchangeable with the \(0.55\) mmol/min insulin-regulated peripheral term.

Choosing a Representation and Diagnosing Problems

The appropriate form depends on the model’s purpose and the glucose range being studied. The saturable form captures concentration dependence and a finite upper rate. A linear clearance may be adequate over a narrow low-concentration range. A fixed basal flux may be convenient when concentration changes little and the aim is a local approximation, but it should not be mistaken for a transport law valid at all concentrations.

Representation	Useful feature	Main limitation
Fixed rate, \(R=R_0\)	Simple local basal removal term	Does not respond to concentration and remains positive at zero concentration
Linear clearance, \(R=CL\,C\)	Simple concentration-dependent flux with clear units	Does not saturate; may overpredict uptake outside the calibrated range
Saturable flux, \(R=V_{\max}C/(K_m+C)\)	Captures near-linear low-concentration behavior and a high-concentration limit	Requires parameters and may not describe regulation or tissue gradients omitted from the model

Several diagnostic checks follow directly from the equations. A saturable or linear uptake flux should be nonnegative for nonnegative \(C\), and the saturable flux should not exceed \(V_{\max}\). A fitted \(K_m\) or \(V_{\max}\) that is negative is incompatible with the stated interpretation. If data cover only a narrow concentration range, \(V_{\max}\) and \(K_m\) may be difficult to distinguish: many parameter pairs can yield similar fluxes there. Inspect predictions across the intended concentration range rather than relying only on fit at the observed points.

Another warning sign is a model that predicts substantial glucose removal at \(C=0\) while using a fixed-rate term. This may be an acceptable local approximation around a nonzero baseline, but it is not a globally coherent uptake law. Also check units: \(V_{\max}\) is a rate, \(K_m\) is a concentration, and \(V_{\max}/K_m\) has clearance units. A common modeling error is to add a concentration-based rate directly to an amount-based balance without converting units consistently.

The saturable law remains a simplification. It does not automatically represent differences in tissue blood flow, local concentration gradients, transporter abundance changes, intracellular metabolism, or changes in erythrocyte or brain state. Nor does a good fit establish that the chosen tissue mechanisms are uniquely identified. Use separate tissue terms when the scientific question requires them; use a pooled term only when the available data and model purpose justify that loss of detail.

Key takeaway: Represent insulin-independent uptake as its own glucose-dependent removal term, with assumptions and units stated explicitly. Keep it separate from insulin-regulated peripheral disposal, and ensure the total mass balance counts every flux once.

Check Your Understanding

Try each question, then compare your reasoning with the solutions below.

For \(R(C)=V_{\max}C/(K_m+C)\), what is the uptake rate at \(C=K_m\), and what are its units?
A tissue has \(V_{\max}=0.60\) mmol/min, \(K_m=3.0\) mmol/L, and \(C=3.0\) mmol/L. Calculate its uptake rate.
Why can a fixed basal flux be a poor representation if the model is used down to \(C=0\)?
At fixed \(C\), which terms should change directly with insulin in the example mass-balance structure, and which should not?
If \(dG/dt=0.24\) mmol/min and \(V=8.0\) L is constant, calculate \(dC/dt\) and state its units.

Solutions. 1. At \(C=K_m\), \(R=V_{\max}K_m/(K_m+K_m)=V_{\max}/2\); the rate has the same units as \(V_{\max}\), such as mmol/min. 2. \(R=0.60(3.0)/(3.0+3.0)=0.30\) mmol/min. 3. A fixed positive rate predicts glucose removal even when no glucose is available, whereas the saturable law gives \(R(0)=0\). 4. Under the stated assumptions, insulin may directly change \(R_{\mathrm{peripheral}}(C,I)\) and any explicitly insulin-dependent other term; \(R_{\mathrm{brain}}(C)\) and \(R_{\mathrm{RBC}}(C)\) do not change directly when \(C\) is fixed. 5. \(dC/dt=(1/V)dG/dt=0.24/8.0=0.030\) mmol L\(^{-1}\) min\(^{-1}\), so concentration is increasing.