How a sparsely covered cell captures molecules

A worked extension of Lecture 2's absorbing sphere: small dispersed receptor patches can collect a substantial fraction of the ideal arrival flux, even when most of the surface reflects approaching molecules.

CCBS · EST 5053 · Lecture 2 extension mini-essay · lecturer's worked example

1Replace the absorbing surface with receptors

The core's reference counts every arrival at the sphere; a receptor model must specify which parts of the surface can capture.

Perfect absorption provides a useful reference flux. For a stationary sphere of radius a, diffusion coefficient D and maintained far-field number concentration $c_\infty$, the steady arrival flux is $J_{\max}=4\pi Da c_\infty$. This is the boundary-value problem derived in core section 8.

Now put N small circular absorbing patches of radius s on the sphere, with $s\ll a$. The remaining surface reflects molecules. Assume dilute independent diffusion, receptors available for capture and no depletion of the distant reservoir. We want the total capture flux relative to the absorbing-sphere reference, not the accuracy of a concentration estimate or the rate of downstream signaling.

Define the capture geometry and output before using the absorbing sphere as a biological benchmark.

2The mixed boundary problem has a useful approximation

The relevant geometrical comparison involves summed patch radii rather than just covered area.

Outside the cell, the steady concentration still satisfies the diffusion equation. Concentration is zero on each absorbing patch and its normal gradient is zero on the reflecting surface. The total flux through enclosing shells is constant, although the local concentration is no longer spherically symmetric.

Berg and Purcell used an electrostatic analogy and an approximation for small dispersed receptor patches to obtain

Their equation 8 interpolates between two useful limits.¹ At small $Ns/(\pi a)$, the flux approaches $4NDs c_\infty$, the sum of isolated small-disk capture contributions. At large $Ns/(\pi a)$, it approaches the whole-sphere reference. These checks explain the formula's behavior; they do not turn an approximation into an exact solution for arbitrary receptor placement.

Use $Ns/(\pi a)$ to compare dispersed-patch capture with the ideal sphere while retaining the small-patch and spatial-distribution assumptions.

3Half the flux needs little surface coverage

A numerical example shows why fractional area alone is a poor capture estimate.

Half capture occurs when the two lengths in the denominator are equal. Setting $J/J_{\max}=1/2$ gives $N_{1/2}\approx\pi a/s$. Take the example radii used by Berg and Purcell: $a=1\,\mu\mathrm m=1000\,\mathrm{nm}$ and $s=1\,\mathrm{nm}$. Then

Thus about three thousand available absorbing patches collect half the ideal flux while covering only 0.079% of the surface. The area-per-patch spacing estimate $\sqrt{4\pi a^2/N}$ is about 63 patch radii. Achieving 90% in the same approximation requires $N\approx9\pi a/s\approx28{,}300$, or 0.71% coverage.

Calculate capture from the spatial search model; a surface that is less than one-thousandth absorbing can still collect half the ideal flux in this geometry.

4Nearby molecules have repeated opportunities to return

Diffusion near a reflecting surface does not consist of a single independent draw from its area.

A molecule just outside a sphere is likely to return. Let $u(r)$ be the probability of eventually hitting the sphere from radius r before escaping to infinity. It solves the spherical Laplace equation with $u(a)=1$ and $u(\infty)=0$. The solution is $u(r)=a/r$. Starting at $r=a+s$ therefore gives

Berg and Purcell used an excursion argument to explain the many opportunities for capture.¹ In an idealized repeated-excursion model that restarts each trial a distance s outside the surface, the expected number of successful returns before escape is $P_{\mathrm{return}}/(1-P_{\mathrm{return}})=a/s\approx1000$. This is a coarse-grained count of excursions, not a literal count of every contact of a continuous Brownian path.

Repeated nearby exploration explains why the first-arrival area fraction underestimates capture. The excursion argument gives the scale of the enhancement. The mixed-boundary calculation is still needed for the spatial correlations and the coefficient in equation (1).

Distinguish return opportunities at a specified distance scale from independent surface contacts when interpreting diffusive search.

5Clustering the same absorbing area reduces capture

The large capture gain depends on dispersal, so total receptor area is insufficient to predict it.

Combine the same N disks into one small patch. Equal area gives a combined radius $s_{\mathrm{one}}=\sqrt N\,s$. Its small-disk capture estimate is $J_{\mathrm{one}}\approx4D\sqrt N\,s\,c_\infty$. At $N=\pi a/s$, division by the ideal flux gives the dimensionless ratio

The dispersed patches give about one half of the ideal flux, so their capture is about 28 times larger than that of the equal-area single patch in these approximations. Here the combined patch radius is about 56 nm, still small compared with the 1 µm cell radius.

Dispersal need not be cost-free or biologically optimal. Receptors can cluster for cooperative signaling, organization or other functions. The calculation isolates the capture consequence of placement; it does not price those other functions.

Compare receptor arrangements at fixed absorbing area to expose the contribution of spatial distribution.

6Turn the capture result into a testable biological question

The model predicts a physical capability, while a biological design claim needs additional evidence.

The sphere reference can remain informative with a mostly reflecting surface. Many dispersed patches can recover much of its flux. A single small enzyme-binding patch is a different geometry, and actual complex formation can additionally require orientation or conformational change. An enzyme may also have multiple sites; the appropriate number and arrangement must come from its structure.

Capture alone does not establish a sensing or signaling bottleneck. Finite receptor reactivity, occupancy, reset time, background binding and downstream processing can all matter. Berg and Purcell also studied concentration-measurement accuracy, a question distinct from the total absorption flux calculated here.

Use the predicted capture curve to choose measurements of receptor geometry and availability before drawing conclusions about cellular design.

†Reference