Without eyes and ears, how does a cell know its position? One cell biologist found good design principles at work in the way cells learn about their environment for making decisions.

Arthur D. Lander, of the Department of Developmental and Cell Biology, Department of Biomedical Engineering, and Center for Complex Biological Systems at U.C. Irvine, wrote an intriguing article for Science about a road less traveled in microbiology (“How Cells Know Where They Are“).

Development, regeneration, and even day-to-day physiology require plant and animal cells to make decisions based on their locations. The principles by which cells may do this are deceptively straightforward. But when reliability needs to be high — as often occurs during development — successful strategies tend to be anything but simple. Increasingly, the challenge facing biologists is to relate the diverse diffusible molecules, control circuits, and gene regulatory networks that help cells know where they are to the varied, sometimes stringent, constraints imposed by the need for real-world precision and accuracy. (Emphasis added.)

Location determination is an engineering problem, and cells have engineering solutions to solve it. The inputs to their positioning systems are usually limited: molecular diffusion gradients. How much can you learn from a flow of molecules all around you? Imagine yourself blindfolded in a room with various scents. You might experiment to see if steps in one direction make a particular scent stronger, or another one weaker; the smell of smoke, for instance, might lead you to a campfire. However you do it, “To measure the distance from one object to another, something must traverse the intervening space,” Lander writes. Either you have to traverse it, or the molecules around you must be moving.

Cells are good at sensing gradients; “given constant production at a source, diffusion can create steady-state gradients within which concentration is a proxy for distance.” Cells are programmed to respond to this environmental information.

From this insight, it was proposed, and later demonstrated, that cells in developing animal embryos receive positional cues from diffusible molecules that indeed form stable gradients across tissues. Such molecules, dubbed morphogens, play central roles in orchestrating developmental pattern formation.

Cells’ ability to know where they are is critical to many life processes, both among microbes and human beings. Leukocytes, for instance, can follow a gradient in a bloodstream to a site of infection. Cells in a growing embryo can follow a gradient to their correct position. Tissue cells can sense their position in a layer of cells.

But diffusion gradients can vary stochastically. Unreliable information can lead to inaccuracy or imprecision. Somehow, in spite of the variations, cells succeed in turning noisy inputs to reliable outputs:

Yet the positional information that cells ultimately obtain is often exceedingly reliable, particularly during development (as evidenced by the remarkably accurate symmetries and family resemblances we encounter in our own bodies). Can steady-state diffusion gradients provide that kind of reliability?

In short, it depends. It depends on the amount and kind of variability cells face, the mechanisms by which gradients form, and how much reliability is required.

(This implies that cells know their own requirements.) The mechanisms cells use to gain robust information from noisy inputs are remarkable. One method is to alter the method by which a gradient is sensed. Cells can alter the production of morphogen relative to the input concentration, use higher levels of morphogen, or monitoring morphogen concentration before it peaks. Each of these strategies has trade-offs:

These scenarios illustrate a basic engineering principle: Strategies that improve performance in one arena typically degrade it in another….

This discussion highlights an important point: The goal of good design is not to make sensitivity go away–because of performance trade-offs every system will always be sensitive to something–but rather to shift it onto the things that are reliable and away from those that are not….

Sure enough, cells exhibit that good design. Lander gives examples of cellular processes with “numerous control mechanisms that reduce or counteract such variability,” providing robust response to environmental cues.

In addition to shifting sensitivites, cells employ another design principle used by engineers: taking multiple independent measurements and averaging them.

In this way, many short-lived disturbances are easily averaged out (for example, fluctuations in the extracellular levels of freely diffusing molecules tend to relax on a time scale of seconds). But some disturbances may have long time scales. For example, stochastic fluctuation in receptor occupancy resulting from the probabilistic nature of binding will reflect the time scales of receptor dynamics, which generally need to be long (e.g., hours) so as not to interfere with the formation of long-range diffusion gradients.

Each strategy, Lander emphasizes, has trade-offs, so cells are designed not to rely too heavily on any one approach. Another strategy is to use switches: committing to an action when the inputs line up.

There are other reasons why switches are a useful thing to include in the machinery with which cells read position: Diffusible carriers of positional information (morphogens) are usually smoothly graded in space, but cells often need to make binary decisions (e.g., to differentiate or not). Response circuits that are switchlike, that is, ultrasensitive, are thus essential.

Another strategy is to pool different kinds of data: e.g., getting “positional information from more than one morphogen at the same time.” Lander gives an example of this in the developing vertebrate brain.

There are even more strategies cells use to deduce where they are. For instance, “Measurements can also be pooled over space instead of time; that is, neighboring cells can share information.” Another strategy: to “fix errors before they happen” (called “disturbance compensation”). This strategy is particularly interesting because it has the advantage of canceling out errors:

This becomes available whenever multiple measurements are affected by a common disturbance and there is a priori knowledge of the nature of their mutual dependence. A particularly simple example is ratiometric measurement: For instance, if a morphogen’s rate of production rises in a certain way with temperature, then measuring its abundance relative to that of some other molecule that rises in the same way with temperature will produce a temperature-corrected reading.

Ratiometric measurement can be implemented in a surprising number of ways. For example, cell-to-cell variation in receptor occupancy resulting from noisy receptor expression can be nullified by measuring the ratio of occupied to unoccupied receptors rather than receptor occupancy per se. Indeed, amounts of the morphogen Hedgehog appear to be measured in just this way. Similarly, when morphogens with opposing actions are produced at opposite ends of a field of cells, the net signal that cells receive can cancel out perturbations that affect both morphogens equally.

Surprisingly, “Even the noise in a signal can drive disturbance compensation, provided there is a known correlation between noise strength and signal strength,” Lander writes. Yes, noise can be a kind of signal, if one knows how to factor it with other information.

One more thing cells can do is use community information to sense position. The arrival of a spreading wave of cell-to-cell signaling can modify the diffusion gradient, providing more information. This intercellular signal “can itself be used to measure distances in much the same way that the time of arrival of thunder allows us to judge distance from a lightening strike.”

There are, in fact, many ways in which short-range cell-cell interactions can make long-range things happen reliably. Systems that do this are called spatially self-organizing, because spatial order emerges directly from collective, or pooled, interactions.….

All in all, we see a complex answer to a simple question: how does a cell know where it is? Here we have seen multiple interacting mechanisms for gathering information from a noisy environment, refining it, and making decisions reliably. This is a form of irreducible complexity — not so much of physical parts interacting, but strategies interacting, much like a software engineer would use multiple strategies to provide robustness for high-reliability software. Cells are so good at it, they gain “exceedingly reliable” information even from noisy, unreliable inputs.

Lander ends with questions that will require more research.

In biology, simple questions rarely have simple answers, and “how do cells know where they are?” is no exception. I have focused here on the problem that, despite the existence of straightforward ways for cells to measure position, making measurements sufficiently accurate and precise is inherently challenging.

Although recent years have seen considerable progress in identifying mechanisms for encoding and detecting positional information, as well as for achieving reliability, there is much we still do not know.

ID advocates can point to this article as an example of design-theoretic approaches leading to fruitful research. There seems no reason to doubt Arthur Lander’s commitment to conventional evolutionary theory, yet his article focuses on design: how a cell employs good engineering strategies to solve a problem. Lander says nothing about how these sensory strategies might have evolved by a Darwinian process. Indeed, Darwinian theory is essentially useless to the entire discussion. Instead, Lander thinks like an engineer. This is clear from his sentence that begins, “This discussion highlights an important point: The goal of good design….”
Keep this in mind when you hear Darwinists arguing that ID is a “show stopper” that somehow brings science to a halt. Here we see how Darwinian evolutionists, when they put a design hat on, ask questions that propel science forward.