A Worked Model of Cell-Derived Signal Spread

Why Model the Space Around a Secreting Cell?

A cell-derived signal can affect nearby cells even when it does not enter the bloodstream or mix uniformly through a tissue. Its local concentration depends on how quickly it diffuses away from its source and how quickly it is removed. A reaction-diffusion model makes those processes explicit and lets us estimate a spatial range: the distance over which concentration remains above a chosen biological threshold.

We will construct a steady-state model for a spherical cell in an unbounded, homogeneous tissue. The cell secretes signal through its surface. In the surrounding tissue, the signal diffuses and is removed by first-order degradation and, when appropriate, an effective first-order term for receptor-mediated uptake. This is a deliberately defined model, not a claim that every tissue has spherical cells, homogeneous transport, or linear uptake.

Model and assumptions: A spherical source has radius \(a\), and \(r\) is distance from its center. The extracellular concentration is \(C(r)\), with units of molecules per volume. The tissue has constant diffusivity \(D\) and a spatially uniform, first-order loss rate \(k\). At steady state, the exterior domain \(r\geq a\) has no sources other than the cell’s prescribed outward surface flux \(J\), and \(C(r)\to 0\) as \(r\to\infty\). The model assumes spherical symmetry, constant parameters, and a linear loss process.

Deriving the Spherical Concentration Profile

For a radially symmetric concentration, the steady-state diffusion equation with first-order loss is

0=D\frac{1}{r^2}\frac{d}{dr}\left(r^2\frac{dC}{dr}\right)-kC,\qquad r>a.

Here \(D\) has units of area per time and \(k\) has units of inverse time. The balance is between diffusion, represented by the spatial derivatives, and removal, represented by \(kC\). Define the characteristic length

\lambda=\sqrt{\frac{D}{k}}.

The length \(\lambda\) combines transport and removal: larger diffusivity increases it, whereas faster loss decreases it. To solve the radial equation, set \(u(r)=rC(r)\). Substitution gives \(u''-u/\lambda^2=0\), whose solutions are exponentials. Thus

C(r)=\frac{A e^{-r/\lambda}+B e^{r/\lambda}}{r}.

The condition that concentration remain finite and approach zero at infinity requires \(B=0\). Applying the surface flux condition, with positive \(J\) denoting secretion from the cell into the tissue, gives the profile below.

-D\left.\frac{dC}{dr}\right|_{r=a}=J,\qquad C(r)=C(a)\frac{a}{r}\exp\left(-\frac{r-a}{\lambda}\right),\qquad C(a)=\frac{J}{D(1/a+1/\lambda)}.

The \(a/r\) factor reflects spherical spreading; the exponential factor reflects loss during transport. In this model, the surface concentration is not specified independently: it follows from the secretion flux and the transport and loss parameters. For a uniform flux over the cell surface, total secretion is \(4\pi a^2J\).

Worked Example: Concentration Around One Cell

Hypothetical model. Let the cell radius be \(a=10\ \mu\mathrm{m}\), the diffusivity be \(D=100\ \mu\mathrm{m}^2/\mathrm{min}\), and the total first-order loss rate be \(k=0.0800\ \mathrm{min}^{-1}\). Let the surface secretion flux be \(J=1000\) molecules/\((\mu\mathrm{m}^2\mathrm{min})\). These are illustrative parameter values, not measurements from a study. The assumed homogeneous tissue and linear loss are exactly the conditions under which the stated equation applies.

First calculate the characteristic length:

\lambda=\sqrt{\frac{100}{0.0800}}\ \mu\mathrm{m} =\sqrt{1250}\ \mu\mathrm{m} \approx35.36\ \mu\mathrm{m}.

The surface concentration is

C(a)=\frac{1000}{100(1/10+1/35.36)} =\frac{1000}{12.828} \approx77.95\ \text{molecules}/\mu\mathrm{m}^3.

At \(r=50\ \mu\mathrm{m}\), the same spherical profile gives

C(50)=77.95\frac{10}{50} \exp\left(-\frac{50-10}{35.36}\right) \approx5.03\ \text{molecules}/\mu\mathrm{m}^3.

This is the predicted steady concentration \(40\ \mu\mathrm{m}\) beyond the cell surface. The calculation assumes the observation point lies in the homogeneous exterior tissue and that the source has reached steady state.

Defining and Calculating Spatial Range

A signal’s “range” is not a unique physical constant unless a concentration threshold is specified. One useful operational definition is the distance \(R\) from the cell center at which concentration falls to a chosen threshold \(C_*\), where \(C_*<C(a)\). The distance beyond the cell surface is \(R-a\). Since the profile decreases monotonically for \(r\geq a\), there is one such threshold crossing under these conditions.

Substituting \(C(R)=C_*\) gives an equation that can be solved numerically for \(R\):

C_* = C(a)\frac{a}{R}\exp\left(-\frac{R-a}{\lambda}\right).

Both spherical spreading and loss determine the crossing. Taking logarithms gives \(\ln(C_*/C(a))=\ln(a/R)-(R-a)/\lambda\). Because \(R\) occurs both inside and outside the logarithm, one generally finds it by a numerical root solve rather than by taking a single logarithm of the threshold ratio.

Worked Example: Finding a Threshold Range

Use the hypothetical cell in the previous example and define the threshold as \(C_*=10\) molecules/\(\mu\mathrm{m}^3\). The threshold is below the surface concentration \(77.95\) molecules/\(\mu\mathrm{m}^3\), so a crossing exists outside the cell. Solve

\ln\left(\frac{10}{77.95}\right) =\ln\left(\frac{10}{R}\right)-\frac{R-10}{35.36}.

Numerical solution gives \(R\approx36.7\ \mu\mathrm{m}\) from the cell center. Substituting this value back into the concentration profile checks the result:

C(36.7)=77.95\frac{10}{36.7} \exp\left(-\frac{26.7}{35.36}\right) \approx10.0\ \text{molecules}/\mu\mathrm{m}^3.

Thus, for this specified threshold and parameter set, the modeled signal extends about \(36.7-10=26.7\ \mu\mathrm{m}\) beyond the cell surface. This is a threshold-based model prediction, not a universal measure of the cell’s biological influence.

Separating Degradation and Receptor-Mediated Uptake

If extracellular degradation is first order with rate \(k_d\), and receptor-mediated uptake can be approximated as a first-order tissue sink with rate \(k_u\), then \(k=k_d+k_u\). The profile responds to their sum: steady concentration measurements alone may not distinguish these loss mechanisms. The partition matters biologically, but requires additional information or measurements to identify.

A constant \(k_u\) is appropriate only when uptake is adequately represented as proportional to concentration over the modeled range. If uptake saturates, a term such as \(V_{\max}C/(K_m+C)\) is not first order in \(C\); the constant-\(\lambda\) solution no longer applies. Receptor binding and internalization confined to cell surfaces may instead require explicit boundary uptake conditions. Those changes should be represented in the model rather than hidden inside a constant tissue loss rate.

Worked Example: How Uptake Changes Concentration at a Fixed Distance

Consider two hypothetical models with the same \(a=10\ \mu\mathrm{m}\), \(D=100\ \mu\mathrm{m}^2/\mathrm{min}\), and secretion flux \(J=1000\) molecules/\((\mu\mathrm{m}^2\mathrm{min})\). In both, degradation is \(k_d=0.0200\ \mathrm{min}^{-1}\). The first has no additional uptake; the second has a linear effective uptake rate \(k_u=0.0600\ \mathrm{min}^{-1}\). Both models assume first-order loss, so the profiles derived above apply.

With degradation alone, \(\lambda=\sqrt{100/0.0200}=70.71\ \mu\mathrm{m}\), and \(C(a)=1000/[100(0.1+1/70.71)]=87.61\) molecules/\(\mu\mathrm{m}^3\). At \(r=100\ \mu\mathrm{m}\),

C(100)=87.61\frac{10}{100} \exp\left(-\frac{90}{70.71}\right) \approx2.45\ \text{molecules}/\mu\mathrm{m}^3.

With uptake, the total loss rate is \(0.0200+0.0600=0.0800\ \mathrm{min}^{-1}\), so \(\lambda=35.36\ \mu\mathrm{m}\) and \(C(a)=77.95\) molecules/\(\mu\mathrm{m}^3\). The predicted concentration at the same distance is

C(100)=77.95\frac{10}{100} \exp\left(-\frac{90}{35.36}\right) \approx0.612\ \text{molecules}/\mu\mathrm{m}^3.

Under these assumptions, adding uptake reduces the predicted concentration at \(100\ \mu\mathrm{m}\) from about \(2.45\) to \(0.612\) molecules/\(\mu\mathrm{m}^3\). The surface concentrations differ too, because the secretion flux—not the surface concentration—was held fixed. Comparing these models therefore shows the combined effect of uptake on profile shape and surface level.

Source Location and Multiple Cells

For a single spherical source, location enters through the distance \(r\) from its center. In a tissue coordinate system, moving the cell moves the center from which that distance is measured. With multiple sources, distances to each source can matter, but the way their contributions are combined depends on how the source and boundary conditions are defined.

The diffusion-loss equation is linear. Therefore, on a common domain with additive source forcing and compatible boundary conditions, the sum of source-specific solutions is also a solution. However, an isolated finite-cell profile has its own exterior domain and prescribed flux boundary at that cell’s surface. Simply adding such profiles does not generally satisfy the prescribed flux condition on every other cell surface: the first cell’s field can have a nonzero normal derivative at the second cell’s boundary. For finite cells with specified surface fluxes, solve the shared-domain boundary-value problem if exact boundary conditions are required.

Worked Example: Estimating Contributions From Two Nearby Sources

As a rough estimate, suppose two identical hypothetical cells have isolated profiles with \(a=10\ \mu\mathrm{m}\), \(\lambda=70.71\ \mu\mathrm{m}\), and \(C(a)=87.61\) molecules/\(\mu\mathrm{m}^3\). At an observation point, suppose the distances from the two cell centers are \(60\ \mu\mathrm{m}\) and \(120\ \mu\mathrm{m}\). Each distance is outside its cell, and each isolated profile uses the same degradation-only parameters as in the preceding example.

Evaluating the two isolated profiles gives

C_1=87.61\frac{10}{60}\exp\left(-\frac{50}{70.71}\right) \approx7.20\ \text{molecules}/\mu\mathrm{m}^3,

C_2=87.61\frac{10}{120}\exp\left(-\frac{110}{70.71}\right) \approx1.54\ \text{molecules}/\mu\mathrm{m}^3.

Adding these estimates gives \(7.20+1.54=8.74\) molecules/\(\mu\mathrm{m}^3\). This sum is an approximation when these are finite cells with prescribed surface fluxes; it is not automatically the exact shared-domain solution. Superposition is exact when the source terms are additive on the same domain and the boundary conditions are compatible with the sum.

Diagnostics, Alternatives, and Limitations

A calculated profile is useful only if its assumptions and implementation are credible. Check that \(D>0\), \(k>0\), and the surface flux has the intended sign and units. A positive secretion flux should produce a positive concentration that decreases with \(r\). The computed profile should also satisfy the surface-flux condition and approach zero far from the source. For a numerical PDE solution, compare it with the analytical profile in this idealized geometry and check mesh and domain-size sensitivity, as appropriate to the numerical method.

A negative concentration, a profile that grows at large \(r\), or a surface gradient with the wrong sign indicates a formulation, boundary-condition, or numerical problem. A predicted range that changes greatly with the chosen outer boundary may mean that the simulated domain is too small. Conversely, agreement with this spherical solution does not establish that real tissue is homogeneous: vessels, extracellular matrix, tissue interfaces, and spatially varying receptor abundance can all alter transport or removal.

Approach	Useful when	Main limitation
Spherical steady-state profile	One isolated source, homogeneous medium, linear loss, and a steady signal	Assumes spherical symmetry and an unbounded exterior domain
Shared-domain spatial PDE	Multiple finite cells, tissue boundaries, heterogeneous parameters, or spatially varying sources	Requires spatial discretization and numerical checks
Time-dependent reaction-diffusion model	Signal has not reached steady state or secretion and removal vary with time	Requires initial conditions and time-resolved computation

The steady-state profile also cannot describe how long a signal takes to reach a location. If the observation period is short relative to the relevant diffusion and removal times, a transient model is needed. Likewise, a linear sink cannot represent saturation, and an unbounded homogeneous domain cannot represent tissue boundaries or spatially localized uptake without modification.

Common Mistakes

Calling the threshold crossing an intrinsic range. Range depends on the threshold, parameters, and geometry; state all three when reporting it.
Using \(\sqrt{D/k}\) without checking the loss model. This length follows from a constant first-order loss rate. Saturable uptake or spatially varying loss does not generally yield the same exponential profile.
Adding finite-cell profiles and calling the result exact. Linearity is not sufficient when the isolated profiles do not satisfy the shared-domain cell-surface boundary conditions.
Holding surface concentration fixed when the model specifies fixed secretion flux. With fixed flux, changing the loss rate changes the surface concentration as well as the decay length.

Key takeaway: For a spherical cell with prescribed secretion flux and linear loss in homogeneous tissue, concentration falls as \(a/r\) times an exponential with length scale \(\lambda=\sqrt{D/k}\). The resulting threshold range is conditional on the model and threshold, and multiple finite sources require boundary conditions to be handled on their shared domain.

Check Your Understanding

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

If diffusivity increases while the first-order loss rate stays fixed, what happens to \(\lambda\), and why?
For the first worked example, what is the distance beyond the cell surface at the \(10\) molecules/\(\mu\mathrm{m}^3\) threshold?
Why does adding a first-order uptake rate to degradation shorten the characteristic length?
When is summing profiles from multiple sources exact, and why can isolated finite-cell profiles fail this test?

Solutions. (1) \(\lambda=\sqrt{D/k}\) increases because it is proportional to \(\sqrt{D}\) when \(k\) is fixed. (2) The crossing is about \(36.7\ \mu\mathrm{m}\) from the center; subtracting the \(10\ \mu\mathrm{m}\) radius gives \(26.7\ \mu\mathrm{m}\) beyond the surface. (3) The total loss rate becomes \(k_d+k_u\), so \(D/(k_d+k_u)\) is smaller than \(D/k_d\), and its square root is smaller. (4) Superposition is exact for additive forcing on a common domain with compatible boundary conditions. Isolated finite-cell profiles generally impose separate surface-flux conditions, and one cell’s field can change the normal derivative at another cell’s surface.