# Netrin and DCC: One Cue, Two Answers

> Netrin marks a place; the growth cone decides what to do there, through its receptors (DCC and UNC5) and, in some cells, its internal state. The same secreted protein draws some axons toward the midline of the embryonic spinal cord and pushes others away. You follow the evidence from the floor-plate explant assay to worm genetics, the vertebrate receptors, a cAMP flip in frog growth cones and mouse deletions that ask where the netrin that matters lies.

Course: Developmental Neuroscience (Axons find their way). Level: advanced university and graduate. About 24 minutes.
Lesson page (link this one when citing): https://stemhop.com/lessons/netrin-and-dcc/
Note: parts of this lesson are still being checked ("Under review").

Builds on: The Growth Cone: The Tip That Steers the Axon · Pioneers, Guideposts and the Logic of Guidance Cues · Ventral to Dorsal: Sonic Hedgehog and BMP in the Neural Tube

## What you'll learn

- Explain that netrin is a secreted guidance cue for axons such as commissural axons, and describe the floor-plate explant assay and how it was used to purify netrins from embryonic chick brain.
- Explain the worm genetics of unc-6, unc-40 and unc-5, and the evidence that DCC mediates attraction while UNC5 receptors mediate repulsion, and use it to state what netrin supplies (a place) and what the growth cone supplies (the sign).
- Predict, as a scoped example, how lowering cAMP or blocking PKA changes a frog spinal growth cone's turn in a netrin gradient.
- Calculate the fractional concentration difference across a growth cone in a gradient, explain why receptor binding limits the backgrounds at which it can be read, and say how adaptation helps.
- Describe how netrin acts in the mouse spinal cord, separating the contribution of floor plate netrin from progenitor-derived netrin laid down along the axons' path, and say what a culture assay can and cannot show about this.

**Put a scrap of floor plate from an embryonic rat spinal cord in a gel near a scrap of dorsal spinal cord, and axons grow out of the dorsal scrap toward the floor plate, though the two never touch.**

Something crossed the gel. That something is netrin, a secreted guidance cue. Yet the same protein draws some axons toward the midline and pushes others away, and in the living embryo much of it does not have to travel far. This lesson asks how netrin was found, how the growth cone (the moving, sensing tip of a growing axon; see [the lesson on the growth cone](https://stemhop.com/lessons/growth-cone/)) reads it, and what decides which way the growth cone turns.

## The layout

Most arguments in this lesson happen on one map, which the first film draws. Picture the embryonic spinal cord cut crosswise, like a slice through a tube, at the stage when commissural axons are growing.

- **The floor plate** is a strip of cells along the ventral midline, at the bottom of the slice.
- **The ventricular zone** lines the central canal, on the inside of the tube. Its cells are **progenitors**, dividing cells that make neurons and glia. Each sends a long fibre outward to the edge of the cord.
- **The pial surface** is that outer edge, covered by a basement membrane, a thin sheet of extracellular matrix.
- Many **commissural neurons** sit in the dorsal cord. Their axons grow ventrally, close to the pial surface, then turn inward to the floor plate and cross it to the other side.

*Figure (AI-generated illustration).* The map for this lesson: in the embryonic spinal cord, commissural axons grow ventrally close to the pial surface, then turn inward and cross the floor plate at the ventral midline. Ventricular-zone progenitors line the canal and reach the pial surface with long fibres. Image: a cross-section of the embryonic spinal cord, dorsal up. Ventricular-zone progenitors line the central canal and send fibres to the outer pial surface; the floor plate sits at the ventral midline. A commissural neuron in the dorsal cord sends its axon down near the pial surface, then inward through the floor plate to the other side.

Two more travellers, each on its own map, complete the picture.

- **Trochlear motor axons** belong to another part of the neural tube, near the junction of midbrain and hindbrain, not to this spinal slice. There they start near the ventral midline and grow dorsally, away from the floor plate, the opposite way from commissural axons.
- **In the worm *C. elegans***, cells and pioneer axons (the first axons to lay down a route) migrate around the body between its ventral and dorsal sides. The worm’s netrin, UNC-6, is made on the ventral side.

The map raises three puzzles.

1. **What crosses the gel?** The floor plate releases something. What is it?
2. **How can one cue** pull commissural axons toward the ventral midline and push trochlear axons away from it?
3. **Is the cue a gradient spreading from the midline**, or something laid down along the path the axons travel?

> **The one idea to carry through this lesson.** Netrin marks a place; the growth cone decides what to do there. Where netrin is (its source, its slope, the surface it lies on) supplies the *where*. The growth cone supplies the *sign*, attraction or repulsion: its receptors first (DCC usually draws it in; UNC5 receptors add repulsion, sometimes together with deleted in colorectal cancer (DCC)), and in some cells its internal signalling state (such as cAMP or calcium), which a matrix protein outside it, laminin, can change.

**What the rule covers.** Netrin’s steering of growth cones and migrating cells in the systems used here: the worm’s dorsal-ventral migrations, vertebrate commissural and trochlear axons, and frog spinal and retinal growth cones. It has three limits, each met below. It does not say how far the cue spreads (the mouse spinal cord is a case in point). “Decides” means receptors plus internal state, not receptor identity alone. And it says nothing about netrin’s other jobs: promoting outgrowth, acting on adhesion, and working through other receptors and partners in some tissues.

## Something crosses the gel

**Film transcript: Something Crosses the Gel**

### Film transcript: Something Crosses the Gel

A floor plate and a piece of dorsal spinal cord sit apart in a gel. How could one steer the other? Start on the map: commissural axons grow from the dorsal cord toward the floor plate. In 1988, in Thomas Jessell's laboratory, Marc Tessier-Lavigne and colleagues tested this in a dish. They cut pieces of embryonic rat spinal cord. Each piece was an explant, living tissue cultured outside the embryo. They set the explants in a collagen gel. Some held dorsal spinal cord, others floor plate. First, the control. A dorsal explant sits in the gel on its own. What should it do?

On its own, it sent out few or no axons. Now place a floor plate explant nearby. Bundles of commissural axons grew out of the dorsal explant, toward the floor plate. The explants never touched. So the effect had to cross the gel. A cue that draws axons from a distance is a chemoattractant. Steering by a diffusing cue is chemotropism. Picture smelling bread from down the street. You know which way the bakery is, but not the recipe. Unlike bread, whether netrin draws you in depends on who is smelling it. To name the molecule, you must purify it. Floor plates are far too small. So in 1994, the laboratory kept rat explants as the detector and searched embryonic chick brain.

Thousands of brains went into the work. The extract was separated over several chromatography columns. After each step, every fraction went back onto rat spinal-cord explants. The axons reported which fractions to keep. Two proteins came through. Their sequences were related to laminin, a matrix protein, and to a worm protein, UNC-6. The authors named them netrin-1 and netrin-2. Each promoted commissural outgrowth. Does promoting outgrowth prove that netrin steers axons? Think about what the assay scored.

No. The assay scored whether axons grew out, not which way they pointed. Outgrowth is not direction. A companion paper supplied direction. Floor plate cells make netrin-1 RNA. And cells engineered to secrete netrin attracted commissural axons at a distance in the gel. The takeaway: netrin is what crosses the gel, and it can attract at a distance.

From the film

- In a gel, a floor plate explant drew commissural axons from dorsal spinal cord at a distance without touching it.
- Netrin is what crosses the gel, and it can attract at a distance.
- Promoting outgrowth is not the same as setting direction, so direction needed its own test.

In 1988, in Thomas Jessell’s laboratory, Marc Tessier-Lavigne and colleagues cut pieces of embryonic rat spinal cord and embedded them in a three-dimensional gel of collagen. Each piece was an **explant**, living tissue cultured outside the embryo. Some explants held dorsal spinal cord, where the commissural neurons sit; others held floor plate.

**Predict.** A dorsal explant sits in the gel on its own, with no floor plate nearby. What should it do?

> **Check your prediction.** On its own, a dorsal explant sent out few or no axons. That is the control. With a floor plate explant nearby, bundles of commissural axons grew out of the dorsal explant toward the floor plate. The explants were apart, so the effect had to cross the gel.

*Figure (AI-generated illustration).* Dorsal spinal cord explants in collagen gel send out few or no commissural axons on their own, but with a floor plate explant nearby, commissural axon bundles grow toward it though the two never touch. Image: two collagen-gel dishes seen from above. A dorsal spinal cord explant alone sends out few or no axons; with a floor plate explant nearby but not touching, bundles of axons curve across the gel toward it.

A cue that draws axons toward its source from a distance is a **chemoattractant**, and steering by a diffusing cue is **chemotropism**. That fits one of the four classes of cue you met in [the lesson on guidance cues and pioneers](https://stemhop.com/lessons/guidance-cues-and-pioneers/): long-range attraction. It does not tell you what the molecule is. Picture smelling bread from down the street. You know there is a bakery, and you know which way to walk, but not the recipe. Unlike bread, though, whether netrin draws you in depends on who is smelling it.

## From an activity to a molecule

To purify a protein you need a lot of tissue, and floor plates are tiny. In 1994 Tessier-Lavigne’s laboratory kept the rat assay as a detector and looked for the same activity in a larger source. Membranes from embryonic chick brain had it. Thousands of brains went into the work. The extract was separated over several [chromatography](https://stemhop.com/methods/protein-purification/) columns, and after each step every fraction went back onto rat spinal-cord explants. The axons reported which fractions to keep.

Two proteins came through. Their sequences were related to laminin, a protein of the extracellular matrix, and to UNC-6, the product of a worm gene. The authors named them netrin-1 and netrin-2. Each promoted commissural outgrowth.

**Outgrowth is not direction.** The purification assay scored whether axons grew out of an explant, not which way they pointed, and a cue can promote growth without steering it. A companion paper supplied direction. Floor plate cells make netrin-1 RNA, and cells engineered to secrete netrin attracted commissural axons at a distance in the gel. That answers the first puzzle: netrin is a cue that can cross the gel.

## The worm already knew: one cue, two readers

**Film transcript: The Worm Already Knew**

### Film transcript: The Worm Already Knew

Before netrin had a name, a worm held the clue. How can one cue send cells in two directions? Start on the spinal cord map, then switch to the worm. There, UNC-6 is made on the ventral side. The worm C. elegans is transparent. Many of its cells can be named and found again in every animal. In 1990, Edward Hedgecock and colleagues followed migrating cells and pioneer axons. Pioneer axons are the first axons to lay down a route.

Some cells move toward the ventral side. Others move away from it. The team looked at mutants of three genes. The first was unc-6, the gene for the cue itself. Lose the cue. Which migrations should fail: toward the ventral side, away from it, or both?

Both. Without UNC-6, many migrations in both directions went the wrong way. The cells and axons still moved, but misdirected. Mutants of unc-40 failed mainly in movement toward the ventral side. Movement away was largely normal. Mutants of unc-5 showed the mirror image. They failed mainly in movement away. One cue, two readers, two directions. unc-40 reads toward. unc-5 reads away. In 1992, UNC-6 turned out to be a laminin-related protein. UNC-5 turned out to span the membrane. The table links genes to directions. To test the link, change the reader and leave the cue alone. Worm touch neurons do not normally carry UNC-5. Their axons do not grow dorsally.

Force them to make UNC-5. Which way do their axons grow? And does the new route still need UNC-6? In 1993, Michel Hamelin and colleagues did this. The axons grew dorsally, away from the ventral UNC-6. And the new route still needed UNC-6.

The cue never changed. It marked the ventral side throughout. The receptor set the direction. The takeaway: changing the receptor changed the answer to an unchanged cue.

From the film

- In the worm, one cue, UNC-6, sends cells and axons in two directions, depending on the reader.
- UNC-40 mainly reads toward and UNC-5 mainly reads away.
- Making worm touch neurons express UNC-5 turned their axons away from the unchanged cue, so the receptor set the direction.

Before vertebrate netrin had a name, worm genetics had found its logic. In 1990 Edward Hedgecock and colleagues followed cells and pioneer axons as they migrated around the body of *C. elegans*, toward or away from the ventral side, in mutants of three genes. Keep in mind that UNC-6 is made ventrally.

| Gene lost | Movement toward ventral | Movement away from ventral | Reading |
| --- | --- | --- | --- |
| *unc-6* (the cue) | disrupted | disrupted | both directions need the cue |
| *unc-40* | mainly fails | largely normal | the “toward” reader: attraction |
| *unc-5* | largely normal | mainly fails | the “away” reader: repulsion |

The asymmetry matters: *unc-40* mutants have strong defects in movement toward the ventral side and weaker defects in movement away. In the migrations compared here, *unc-5* mutants disrupt movement away from the ventral side while sparing movement toward it. Without UNC-6, cells and axons still move, but many take the wrong route. One cue, two readers, two directions. In 1992 UNC-6 turned out to be a laminin-related protein, and UNC-5 a protein that spans the membrane.

The table links genes to directions. To test the link, change the reader and leave the cue alone.

**Predict.** Worm touch neurons do not normally carry UNC-5, and their axons do not grow dorsally. Force them to make UNC-5. Which way do their axons grow, and does the new route still need UNC-6?

> **Check your prediction.** In 1993 Michel Hamelin and colleagues in Joseph Culotti’s laboratory did this. With UNC-5, the touch-neuron axons grew dorsally, away from the ventral UNC-6, and this still needed UNC-6. Changing the receptor changed the answer to an unchanged cue.

## The vertebrate readers: DCC and UNC5

**Film transcript: The Vertebrate Readers: DCC and UNC5**

### Film transcript: The Vertebrate Readers: DCC and UNC5

The worm found two readers for one cue. Do vertebrate growth cones use the same ones? Back on the spinal cord map, follow the commissural axons to the floor plate. The worm's toward reader, UNC-40, turned out to resemble a human protein. Its name is DCC, Deleted in Colorectal Cancer. It was first studied as a cancer gene. In 1996, Tessier-Lavigne's laboratory tested DCC in rat embryos, spinal explants and cultured cells. They showed three things. First, DCC is on commissural axons. Second, netrin-1 binds cultured cells that display DCC. Third, they added an antibody that blocks DCC to spinal explants with netrin. What should happen to the outgrowth?

The antibody selectively blocked netrin-evoked outgrowth from the spinal explants. In the same journal issue, worm work showed that UNC-40 is needed in the moving cells themselves. The mouse confirmed both partners. Mice with little or no netrin-1 had misrouted spinal commissural axons. They also lacked several forebrain commissures. Mice lacking DCC had similar defects. Remove the cue or remove the reader, and commissural axons fail in much the same way. Now repulsion. In 1995, netrin-1 repelled trochlear motor axons in explant cultures. So the second puzzle is real in vertebrates. Different growth cones, same cue, opposite turns. The worm already suggested why: different readers.

In the netrin-1 mutant, though, trochlear axons were largely normal. Other repellents also guide them. The fly shows how the repulsive reader works. Its Unc5 receptor is a repulsive netrin receptor. Unc5 alone is enough for short-range repulsion. Long-range repulsion also needs Frazzled, the fly's DCC-family receptor. So the receptor layer reads like this. DCC draws a growth cone in. UNC5 receptors add repulsion. Treat that as a starting point. DCC-family receptors also take part in some repulsion. And netrin has other receptors and partners. The takeaway: the cue is shared, and the receptors a growth cone carries set the first answer.

From the film

- DCC, the vertebrate relative of UNC-40, is on commissural axons, and blocking it blocked netrin-evoked outgrowth.
- In mice, removing netrin or removing DCC left commissural axons misrouted in much the same way.
- DCC usually draws a growth cone in and UNC5 receptors add repulsion, but that is a starting point, since other receptors and partners take part.

UNC-40 turned out to resemble a human protein, DCC (Deleted in Colorectal Cancer), found in 1990 on a stretch of chromosome 18 often lost in colorectal tumours and studied as a candidate tumour suppressor (see [the lesson on oncogenes](https://stemhop.com/lessons/oncogenes/)). In 1996 Tessier-Lavigne’s laboratory, working with rat embryos, spinal explants and cultured cells, showed three things.

1. DCC is on commissural axons.
2. Netrin-1 binds cultured cells displaying DCC.
3. An antibody to DCC selectively blocks netrin-evoked outgrowth from spinal explants.

In the same journal issue, a worm study showed that UNC-40 is needed in the moving cells themselves.

The mouse confirmed both partners. Mice with little or no netrin-1 (Serafini and colleagues, 1996) had misrouted spinal commissural axons and lacked several forebrain commissures, and mice lacking Dcc (Fazeli and colleagues, 1997) had similar defects. Remove the cue or remove the reader, and commissural axons fail in much the same way.

Repulsion has a vertebrate example too. In 1995 Sophia Colamarino and Tessier-Lavigne found that netrin-1 repels trochlear motor axons in explant cultures. So the second puzzle is real in vertebrates, and the worm suggests its answer: different readers, opposite turns. In the netrin-1 mutant, though, trochlear axons were largely normal, so other repellents also guide them.

The fly shows how the repulsive reader works. Its Unc5 receptor is a repulsive netrin receptor: Unc5 alone is enough for short-range repulsion, and long-range repulsion also needs Frazzled, the fly’s DCC-family receptor (Keleman and Dickson, 2001). That is the receptor layer of the rule; treat it as a starting point.

## Same receptor, opposite turn

**Film transcript: Same Receptor, Opposite Turn**

### Film transcript: Same Receptor, Opposite Turn

Can a growth cone with the same receptor turn toward netrin in one condition and away in another? Start on the map, then zoom into one growth cone in a dish. Receptors are one layer. In some cells, the growth cone's internal state is a second. Mu-ming Poo's laboratory developed a turning assay. A fine pipette puffs short, regular pulses of a cue beside one growth cone. Time-lapse imaging measures which way the growth cone turns. In 1997, Guo-li Ming and colleagues used frog spinal neurons. Their growth cones turned toward netrin-1. Next they blocked cAMP signalling. They blocked PKA, the enzyme cAMP switches on. With PKA blocked, they puffed netrin again.

Toward, away, or no turn? The same growth cones turned away. Attraction had become repulsion. Now add an antibody to DCC. If the receptor stays the same, what should the antibody block? It stopped both turns, the attraction and the repulsion. Both needed DCC.

The receptor was the same. The cell's cAMP chose the direction, in this system. This is the rule's second limit at work. A growth cone carrying DCC was repelled. It is one worked example in one system. It is not a rule for every netrin-responsive axon. Calcium can also change the response. So can laminin, and the eye shows why a flip might be useful. Retinal axons run along the inner surface of the retina to the exit. That exit is the optic disk, where netrin-1 is concentrated. In mice lacking netrin-1 or DCC, the axons still reach the optic disk. But they fail to leave through it. In frog retinal growth cones, laminin-1 lowers cAMP. It turns netrin's attraction into repulsion.

Laminin lies on the retinal surface. In the authors' model, netrin repels where laminin and netrin overlap. That may help push axons down into the exit. The takeaway: the growth cone decides, with its receptors and, in some cells, its internal state.

From the film

- A growth cone with the same receptor can turn toward netrin under one condition and away under another.
- In frog spinal growth cones, blocking cAMP signaling turned DCC-dependent attraction into repulsion.
- This is one worked example in one system, not a rule for every netrin-responsive axon.

Receptors are one layer. In some cells the growth cone’s internal state is a second. Mu-ming Poo’s laboratory developed a [turning assay](https://stemhop.com/methods/axon-guidance-assays/): a fine pipette puffs short, regular pulses of a cue near one cultured growth cone, off to one side, and time-lapse imaging measures the turn. In 1997 Guo-li Ming and colleagues found that growth cones of frog (*Xenopus*) spinal neurons turn toward netrin-1.

Next they blocked cAMP signalling, in one test with Rp-cAMPS, a cAMP look-alike that blocks protein kinase A (PKA, the enzyme cAMP switches on), and in another with an inhibitor of PKA. (For how cAMP activates PKA, see [the lesson on cAMP and PKA](https://stemhop.com/lessons/camp-and-pka/).)

**Predict.** With PKA blocked, puff netrin-1 beside the growth cone. Does it turn toward, turn away, or not turn? Then add an antibody to deleted in colorectal cancer (DCC). If the receptor is unchanged, what should the antibody block?

> **Check your prediction.** The same growth cones turned away. An antibody to DCC stopped both turns, the attraction and the repulsion. The receptor was the same, and the cell’s cAMP chose the direction in this system.

This is one worked example in one system, not a rule for every netrin-responsive axon. Calcium and, in frog retinal growth cones, the matrix protein laminin, which lowers their cAMP, can also change the response (Going deeper follows both).

*Figure (AI-generated illustration).* Netrin marks the place; the growth cone sets the sign. DCC usually draws a growth cone toward netrin and UNC5 receptors add repulsion, and in frog spinal growth cones, blocking cAMP signalling turned the same DCC-dependent attraction into repulsion. Image: four growth cones beside a netrin source on the right. Left pair: a growth cone with DCC turns toward the source; one with UNC5 added turns away. Right pair, frog growth cones with DCC only: with cAMP signalling on it turns toward the source; with PKA blocked it turns away.

## Where is the netrin?

**Film transcript: Where Is the Netrin?**

### Film transcript: Where Is the Netrin?

Is netrin a gradient spreading from the midline, or a path laid down along the way? Return to the map: the floor plate at the bottom, the progenitors around the canal. Start in a dish. A growth cone compares the cue on its two sides. Around a small source, like a bead, concentration falls with distance. The fractional difference across a growth cone is roughly its width over its distance from the source. Close in, a growth cone sees about a ten percent difference. A millimetre away, about one percent. But a growth cone measures bound receptors, not netrin itself. The Kd is the concentration that fills half the receptors. Give it the same ten percent difference at three backgrounds: far below the Kd, at it, far above. Where is the bound difference largest?

At the Kd. Far below it, almost no receptors are bound on either side. Far above it, almost all are. Binding is noisy, because molecules arrive and leave at random. A tiny difference is easily lost. A fixed detector gets the most signal near its Kd. Adapting, by resetting its baseline, lets the growth cone respond again as the background changes. It cannot remove the binding limit. Now the embryo, where netrin-1 is made in two places. Floor plate cells make it. So do progenitors across a broad stretch of the ventricular zone. The progenitors send netrin out along their fibres. It builds up at the pial surface, along the commissural axons' path.

So there are two candidate ways to guide. One is a gradient diffusing from the floor plate. That is chemotropism. Guidance by netrin held on a surface is called haptotaxis. In 2017, two groups deleted the netrin-1 gene from one source at a time, in mice. A conditional deletion removes a gene only in chosen cells. If a floor plate gradient did most of the guiding, which deletion should cause more errors?

The floor plate deletion should. It did not. Without floor plate netrin, most commissural axons stayed on course. Without progenitor netrin, many commissural axons went off course. A third study found the same pattern in the hindbrain.

A deletion shows which source the axons need, not how they read it. A path and a short-range gradient both fit. Nor is the floor plate idle. In the spinal cord, later studies found that its netrin still contributes. So, much of the netrin that matters seems to act near where progenitors lay it down. That holds in the mouse spinal cord and hindbrain. In the spinal cord, the floor plate keeps a part. After crossing, the axons do not turn back. What changes in the reader is the next lesson. The takeaway: netrin marks a place, and the growth cone decides what to do there.

From the film

- A growth cone measures bound receptors, so a small difference across it is easiest to detect near the receptor's KD.
- In mice, removing progenitor netrin sent many commissural axons off course, while most stayed on course without floor plate netrin.
- Much of the netrin that matters seems to act near where progenitors lay it down, though the floor plate keeps a part.

The gel shows what netrin *can* do from a distance. The third puzzle asks what the embryo relies on. Two questions sit inside it: can a growth cone read a gradient at all, and where is the netrin that the embryo’s axons actually use?

### In the dish: reading a shallow gradient

A growth cone steering in a gradient compares the cue on one side with the cue on the other, a few micrometres apart. Work it in three steps.

*Step 1: the concentration difference.* For steady diffusion from a small source in three dimensions, such as a bead in a gel, with no cue lost on the way, concentration falls as 1/r (the same flow of cue spreads over ever larger spheres), and the fractional difference across a growth cone is roughly its width over its distance from the source:

> **fractional difference ≈ w / r** (for C = k / r; w = growth cone width, r = distance)

A 10 micrometre growth cone 100 micrometres away sees about a 10 percent difference. A millimetre away it sees 1 percent.

*Step 2: receptors read occupancy, not concentration.* A growth cone does not measure netrin directly. It measures how many of its receptors are bound. In the simplest binding model the fraction bound is C / (C + Kd), where Kd is the concentration that fills half the receptors.

**Predict.** Give a growth cone the same 10 percent concentration difference across its width, at three background levels: a hundredth of Kd, Kd itself, and a hundred times Kd. At which background is the difference in bound receptors between its two sides largest?

> **Check your prediction.** At Kd. The simple binding model gives these fractions bound.
>
> | Background netrin | Bound, far side | Bound, near side | Difference |
> | --- | --- | --- | --- |
> | a hundredth of Kd | 0.99% | 1.09% | 0.1 points |
> | Kd | 50.0% | 52.4% | 2.4 points |
> | a hundred times Kd | 99.01% | 99.10% | 0.1 points |
>
> Far below Kd, almost no receptors are bound on either side. Far above it, almost all are. Either way the same 10 percent concentration difference becomes a sliver of occupancy.

Binding is also noisy, since ligand molecules arrive and leave at random, so one snapshot of occupied receptors cannot be trusted and a difference of a tenth of a point is easily lost.

*Step 3: what adaptation can and cannot do.* Fixed receptors give the most signal in a band of backgrounds around Kd, yet growth cones meet netrin at very different levels, near a source and far from it. A growth cone that resets its baseline, **adapting** to the background, keeps its internal response from sticking at one level, so it can respond again as the background changes. It cannot undo the binding limit: if Kd is unchanged, the occupancy difference is still small and noisy far from Kd. Both halves have been seen, in different systems: sensory axons in controlled gradients of nerve growth factor were extremely sensitive, but only within a narrow range of concentrations (Rosoff and colleagues, 2004), and frog growth cones in a raised background of netrin-1 lost sensitivity and then regained it (Ming and colleagues, 2002).

### In the embryo: which source do the axons need?

Go back to the map. Netrin-1 is made in two of its places: by floor plate cells, and by ventricular-zone progenitors. The progenitors send their netrin-1 out along their fibres, and it builds up at the pial surface, along the stretch the commissural axons travel. That gives two models to test in the embryo: a gradient diffusing from the floor plate, as in the gel, or a path of netrin held on the surface the axons grow along. Guidance by a cue bound to a surface, unlike chemotropism, is called **haptotaxis**.

**Predict.** In 2017 two groups, Dominici and colleagues and Varadarajan and colleagues, deleted the netrin-1 gene from one source at a time in mice, by conditional deletion, which removes a gene only in chosen cells. If a gradient from the floor plate did most of the guiding, which deletion should cause more errors?

> **Check your prediction.** The floor plate deletion should, and it did not. Without floor plate netrin, most spinal commissural axons stayed on course and reached the midline. Without progenitor netrin, many commissural axons went off course. A third study found the same pattern in the hindbrain.

**What this shows, and what it does not.** A deletion shows which source the axons need, not how they read it. A netrin path laid along the pia fits the result, and so does a short-range gradient near it; the experiment does not separate the two. Nor is the floor plate idle. In 2019 two further studies found that floor plate netrin still contributes, working with progenitor netrin and, in one study, with Sonic hedgehog, another floor plate cue.

*Figure (AI-generated illustration).* Netrin-1 comes from two sources: the floor plate, and ventricular-zone progenitors whose netrin collects at the pial surface along the commissural axons' path. In mice, removing progenitor netrin sent many commissural axons off course, while most stayed on course without floor plate netrin; which source matters does not by itself say how axons read it. Image: a cross-section of the embryonic mouse spinal cord, dorsal up. Netrin dots sit at the floor plate and travel out along progenitor fibres from the cells lining the canal, collecting in a line along the pial surface, where a commissural axon grows down to the floor plate. An inset compares two deletions: without floor plate netrin, three of four axons still reach the floor plate; without progenitor netrin, only one does and three wander.

So, for commissural axons in the mouse spinal cord and hindbrain, much of the netrin that matters seems to act close to where progenitors lay it down, with the floor plate keeping a part in the spinal cord. Elsewhere netrin’s range has to be tested place by place. That answers the third puzzle, for this system.

### Design problem

How would you experimentally test whether netrin acts as a long-range diffusible gradient versus a short-range contact cue? Design a simple in vitro experiment.

> **One way to do it.** Use commissural explants from the dorsal spinal cord in a three-dimensional collagen gel. In one set, place aggregates of cells that secrete netrin at measured distances, say 100, 300 and 600 micrometres, with a porous filter or a clear gap so that no cell can touch the explant. Score outgrowth and turning toward the source at each distance. In a second set, offer netrin bound to the substrate in a stripe or spot, with no secreting cells, and ask whether axons grow only on the netrin or turn toward it from off the stripe.
>
> Turning at a distance, falling off as the source moves away, supports a diffusible cue. Growth only on bound netrin, with no turning from a distance, supports a contact cue. A cue could pass both tests, and netrin can act in both forms in vitro.

## Putting it together

The three puzzles share one answer: netrin marks a place; the growth cone decides what to do there. The gel showed the cue can travel; the worm showed that changing the reader changes the sign, and vertebrate studies named the readers; the frog growth cone showed the reader’s state can flip it; and the mouse showed where the netrin that matters lies.

Commissural axons pose one more puzzle. Once across the floor plate, they do not turn back toward it. What changes in the reader after it crosses, with Slit and Robo signalling, is the subject of [crossing the midline](https://stemhop.com/lessons/crossing-the-midline/).

## Going deeper

### Binding sites and a retracted paper

A 2014 [crystal structure](https://stemhop.com/methods/x-ray-crystallography/) showed that one netrin can bind deleted in colorectal cancer (DCC) at two sites, a DCC-specific one and a second, generic one, and that UNC5A can take DCC’s place at the generic site, which the authors proposed switches the response from attraction to repulsion (Finci and colleagues). A 1999 frog-neuron paper that had reported this switch with chimeric receptors (Hong and colleagues) was retracted in 2023, so this lesson does not rely on it.

### Calcium, laminin and the retinal exit

Calcium is part of the readout. In 2003 Makoto Nishiyama and colleagues reported that the ratio of cAMP to cGMP in these growth cones sets the direction of the netrin-1 turn by changing the activity of L-type calcium channels (see [the lesson on calcium signaling](https://stemhop.com/lessons/calcium-signaling/)). The eye shows why a flip might be useful. Retinal axons run along the inner surface of the retina to the optic disk (optic nerve head), the exit, where netrin-1 is concentrated. In mice lacking netrin-1 or DCC the axons still reach the optic disk but fail to leave through it, so netrin acts locally there. In 1999 Vivian Höpker and colleagues found that laminin-1 turns netrin’s attraction into repulsion in frog retinal growth cones, and that laminin-1 lowers their cAMP. Laminin-1 lies on the retinal surface. Their model: where laminin and netrin overlap, netrin repels, and this pushes axons down into the exit, where there is netrin without laminin.

### Real gradients are not ideal

Departures from the assumptions of Step 1 change the profile. The floor plate is a strip, not a point. In a long-line approximation with an outer boundary R, C is proportional to ln(R/r) and the fractional difference is about w / [r ln(R/r)], which is shallower than w/r once R is more than about 2.7 times r. Netrin also binds the extracellular matrix. Reversible binding holds a bound pool and slows how fast the profile forms, but does not by itself steepen it. Returning to a steady small point source in homogeneous three-dimensional space, suppose netrin is irreversibly removed at a uniform first-order rate kC, where k has units of inverse time. The concentration outside the source is then C ∝ e−r/λ/r, with λ = √(D/k). For w much smaller than both r and λ, the fractional difference is approximately w(1/r + 1/λ), so far from the source it does not fall as low as w/r. These equations describe the point-source model; a line or sheet source has a different profile.

For timing, if netrin’s effective diffusion coefficient D is near 70 µm²/s (an illustrative value, not a measurement), the diffusion time scale r²/D over half a millimetre is about an hour. That is a time scale, not the time to a steady gradient: for a source switched on in an open gel, the concentration 500 µm out is only about half its final value after an hour and takes more than a day to reach 90 percent of it. Binding slows this further.

How much noise matters depends on numbers that are not measured for netrin. In an illustrative model with 10,000 independent receptors split between the two halves of a growth cone, the gradient whose signal just equals one standard deviation of the snapshot noise is about 4 percent at Kd and about 20 percent at a hundredth or a hundred times Kd. Reliable detection needs a larger signal than that. Averaging M independent snapshots over time cuts the noise by about √M.

### Who did the experiments

The 1988 gel assay was by Marc Tessier-Lavigne, Marysia Placzek, Andrew Lumsden, Jane Dodd and Thomas Jessell. The 1994 purification had Tito Serafini as first author and Timothy Kennedy as second, and the companion paper on direction was by Kennedy, Serafini, José de la Torre and Tessier-Lavigne. The 1990 worm study was by Edward Hedgecock, Joseph Culotti and David Hall. The 1996 DCC study was led by Kazuko Keino-Masu and Masayuki Masu in Tessier-Lavigne’s laboratory, and the worm UNC-40 study in the same issue by Shirley Chan and colleagues. The 1997 turning study was by Guo-li Ming, Hong-jun Song and colleagues in Poo’s laboratory.

---

From STEMhop (https://stemhop.com), free lessons by Christopher Deppmann. Lesson text © the authors. Practice questions: https://stemhop.com/lessons/netrin-and-dcc/
