Chapter 13: Neural Tube Formation and Patterning

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to The Deep Dive.

You know, when we think about the central nervous system, the brain, the spinal cord, we usually jump straight to the big stuff.

Consciousness, memory, you know, all the philosophical outcomes.

The glorious end products, but the actual process of building that foundation, that is infinitely more mysterious and frankly incredibly elegant.

It's the ultimate engineering project, isn't it?

It really is.

The father of modern neuroscience, Santiago Ramon Icahal, he put it so perfectly.

He talked about hunting for cells with delicate and elegant forms, the mysterious butterflies of the soul.

That's beautiful.

It really sets the stage for our mission today, which is to ask, how is this astonishingly complex organ, the central nervous system, actually built?

How is it integrated and patterned, starting, and this is the amazing part, literally from a flat sheet of cells?

Exactly.

So our deep dive today is going to focus on that very first pivotal event.

We're talking about the transformation of a simple flat layer of neuroepithelial cells into a hollow tube.

A process called neurulation.

And the structure it forms, the neural tube, that's the precursor to the entire central nervous system, the CNS.

And just to place this in time for you, this happens directly after gastrulation.

So the three primary germ layers have just been established and we're following the fate of one of them.

The ectoderm, the outermost layer.

Right.

And what's so fascinating is that this whole complex process is guided by really a few key mechanisms.

We're talking about cell growth,

changes in cell shape, and crucially, these carefully sculpted gradients of signaling molecules.

Or phogen gradients.

Exactly.

But let's unpack that initial setup because the ectoderm, I mean, it has a massive job right from the start.

It has to generate three major, very different fates all at the same time.

Okay.

So what's on its to -do list?

Well, first, it has to produce the neural plate, which is what will become the CNS.

Second, it has to make the epidermis, which is the outer layer of your skin.

And third, it has to form the presumptive neural crest.

And those cells are incredible.

They give rise to the entire peripheral nervous system, pigment cells, parts of the face.

I mean, so many other structures.

So brain, skin, and everything in between, how is that managed?

It's managed by a principle that is so potent in its simplicity.

It all comes down to the concentration of one key signal,

bone or morphogenic protein, or BMP.

So we're talking about initial specification, the cell's first job assignment.

And it's based on how much BMP it senses during gastrulation.

Precisely.

There's a rule book.

If an ectodermal cell is experiencing high levels of BMP signaling, that cell is told you will become epidermis.

You can think of high BMP as the skin signal.

Okay.

High BMP equals skin.

What's the opposite?

The opposite is very low levels of BMP.

This happens because the underlying mesoderm is actively secreting BMP inhibitors.

So if a cell finds itself in a BMP -free zone, it gets a different instruction.

It becomes the neural plate.

It's specified as the neural plate, the future CNS.

And so the third fate, the neural crest, must be in the middle somewhere.

That's the Goldilocks zone.

The cells that encounter intermediate levels of BMP right at the border, they're specified to become those presumptive neural crest cells.

They'll just sit there at the boundary between the future brain and future skin, waiting for their next signal.

So that simple gradient dictates whether you become a brain cell, a skin cell, or a pigment cell.

But I mean, how reliable is that?

If BMP levels fluctuate just a little bit, does that blue the line?

That is an excellent question.

And it gets to the heart of why development needs what you could call molecular guardians.

Cells don't just passively read the signal once.

They have to actively lock in their fate.

They need to commit.

They absolutely need to commit.

And that's where we get into the genetics of it all.

So let's get into that, section one, the two modes of neuralation.

Right.

So to make sure the neural plate stays committed, to make sure it doesn't flip back to being skin or neural crest, the embryo relies on a family of really powerful transcription factors.

And these are the SOX factors, specifically SOX 1, 2, and 3.

Exactly.

And being transcription factors, their job is to turn genes on and off.

But they have this incredibly efficient two -part function that's essential for neural identity.

OK, what's their dual mandate?

Function number one is what you'd expect.

They activate the whole suite of genes required to build neural tissue.

But function number two is the really crucial protective measure.

They inhibit the other options.

They actively inhibit the alternative fates.

They go in and block the transcription of VMPs and block their signaling pathways.

So they don't just flip the on switch for neural.

They're taking a sledgehammer to the control panel for the skin and neural crest programs.

That's a perfect analogy.

It's a classic developmental principle.

To specify one cell type, you often have to actively block the specification of all the others.

It's an exclusive club.

Very exclusive.

So this StrongSOX expression establishes the neural plate cells as committed neural precursors.

Now they have their identity.

But they immediately face a huge physical challenge.

They're a flat sheet on the outside of the embryo.

They need to get inside and form a tube.

Exactly.

And this is where we see that it's not one single process along the whole body.

There are two principal modes of neurulation.

Primary neurulation and secondary neurulation.

Primary neurulation is the mechanism used for the anterior or rostral part of the body.

So this is what forms the brain and most of the spinal cord.

And that's the one we usually think of with the folding.

That's the classic folding process.

The cells surrounding the neural plate cause it to proliferate, to invaginate, fold up, and then pinch off from the surface to form that hollow tube.

OK.

So that's the front end.

What about the back end?

The mechanics change completely for secondary neurulation.

Completely.

This handles the posterior or caudal portion.

In us, that's the sacral vertebrae, the tail region.

And secondary neurulation just skips the folding part entirely.

So how does it work?

Instead, you have mesenchyme cells, which are a sort of loose collection of cells.

They aggregate together into a solid, unorganized rod.

A solid cord.

A solid cord, called the medullary cord.

Then this solid cord forms little internal cavities, or lumens.

And those little bubbles gradually connect and coalesce into the final, single, hollow neural tube.

That's fascinating.

It's like two totally different construction projects building the same structure, just using radically different techniques.

If folding works so well for the brain, why switch to this, condensation and hollowing process for the tail?

That's a really deep question in evolutionary biology.

The consensus seems to be that it's linked to how the tail itself grows.

Secondary neurulation seems to be an innovation in vertebrates tied to the formation and stretching of the tail bud.

So it's about continued growth.

It seems to be.

The embryo needs to keep growing and elongating posteriorly after the main part of gastrulation is done.

This condensation mechanism is just more effective for building those caudal structures on the fly, as it were.

So we've got a folding front end and a condensing back end.

There has to be a point where they meet.

There does, and it's a critical point.

It's called junctional neurulation.

It's the transition zone.

In humans, this spans the thoracolumbar region.

So think your mid to lower back.

Exactly, where the thoracic vertebrae meet the lumbar.

And this spot is so important because the tube isn't finished until these two independently formed sections properly join up.

The lumen from the primary tube has to connect smoothly to the lumen formed by that secondary hollowing process.

I can imagine that's a point where things could easily go wrong.

It is a sensitive area, and sometimes it involves a mix of both folding and hollowing right at that junction to make the final connection seamless.

Okay, since primary neurulation builds the most complex parts, the entire brain, let's do a really deep dive into the mechanics of that folding process, section two.

Let's do it.

We can break it down into four sequential stages, and we'll use the chick embryo as our classic model here.

So stage one.

Stage one is elongation and folding.

The first thing you notice is that the cells in the neural plate, they don't just divide randomly, they divide preferentially along the anterior -posterior axis.

So they're pushing the whole structure to get longer.

Exactly, it fuels the elongation of the entire body.

And as this happens, the edges of the plate start to thicken and rise up.

These become the neural folds.

And that creates the little U -shaped trough down the middle.

The neural groove, that's right.

And all this time, the presumptive epidermis, that high BMP territory, is right next door, and it's interacting with the neural plate.

It's actually pushing on the folds, helping to bend them inward.

So it's not just the plane itself folding, there's an external force as well.

It's a combination of forces.

Which brings us to stage two.

The bending of the neural plate.

And this is structurally impossible without the formation of specific anchor points.

The hinge regions.

The hinges.

These are the pivots the whole tissue has to curve around.

And we have the medial hinge point, the MHP, and then the dorsolateral hinge points, the DLHPs.

Correct.

The MHP forms first, right down the midline.

And these MHP cells are physically anchored to the structure directly underneath them.

The notochord.

The notochord.

This anchoring creates that initial deep furrow.

The notochord is literally pulling the center of the plate downwards.

So if the notochord is misplaced, or if it's too short, does the whole fold get skewed?

It absolutely does.

That's a testable prediction.

And we see it in experience.

Defects in the notochord lead to severe midline defects in the neural tube.

It confirms that critical anchoring rule.

Okay, so MHP is the first anchor.

Then what?

Shortly after the MHP forms, the two DLHPs appear out on the sides, laterally.

These are two separate hinges.

And they're induced by and anchored to the surface ectoderm that's flanking the neural plate.

So you have one central anchor pulling down and two side anchors to shape the curve.

Precisely.

Once those pivots are established, we're at stage 3, convergence of the neural folds.

The hinges act like fulcrums.

They do.

They rotate the sheets of cells.

And that continued pushing from the presumptive epidermis drives the neural folds closer and closer together up toward the dorsal midline.

And that leads us to the final step, stage 4.

Closure of the neural tube.

The folds touch, they adhere, and they fuse along the top.

And something really important happens right at that moment of closure.

The neural crest cells.

A highly specialized group of cells, at the very apex of those folds, they delaminate.

They just pull away from the epithelial layer and start migrating all over the body.

Those are the neural crest cells.

And once they've left, the tube is officially closed.

The newly formed, fully closed neural tube then separates cleanly from the overlying surface ectoderm, which will become the skin, and it sinks down into the embryo.

The cell shape change itself is, to my mind, the real mechanical marvel here.

I mean, how does a flat sheet of perfectly nice rectangular cells bend?

You can't just shrink one side of the sheet.

You can't.

You have to actively change the geometry of the cells themselves.

And the key mechanism for this is called apical constriction.

OK, break that down for us.

Imagine a row of rectangular books standing side by side on a shelf.

To get that row to curve inwards, you'd have to turn each book into a wedge shape.

Right, like a piece of a pie, wider at the bottom than the top.

Exactly.

A truncated pyramid, wider at the base than at the apex.

The cells in the hinge points do precisely that.

How do they achieve that specific wedge shape on a molecular level?

It's achieved by the localized contraction of iconomyosin complexes.

These are the same little contractile filaments that power your muscles.

But here, they are concentrated specifically at the apical border, the top surface of the MHP and DLHP cells.

So they pull on the top surface like a purse string.

That is the perfect analogy.

They contract, they shrink that apical surface area, and that forces the cell into the necessary wedge shape.

So the forces are intrinsic.

The cells are generating the curve from within.

But you mentioned they get some help.

They do.

We should definitely mention that the mechanical pushing from the presumptive epidermis helps.

And there's also evidence that faster cell division rates in the dorsal lateral parts of the plate increase the cell density there, which promotes buckling right at the DLHPs.

It's just a marvelous orchestration of forces.

But the deepest mystery, I think, is always the regulation.

What's the signal that tells the MHP cells and the DLHP cells where and when to contract their actinomyosin and become wedges?

And that brings us right back to our favorite morphogens,

particularly SH and BMP.

Let's start with the MHP, the medial hinge point.

It's on the ventral side.

Right.

And its positioning is driven by the notochord sitting right underneath it.

The signaling molecule, as you might guess, is sonic hedgehog or SH.

Of course.

She's absolutely required to induce the floor plate cells in the ventral neural plate.

And it's those floor plate cells that establish the MHP and dig that initial neural groove.

OK.

So SH from the notochord sets the center point, the ventral anchor.

But now what about the DLHPs?

They're induced dorsally closer to the surface ectoderm, which we already said is pumping out high levels of BMP.

How does a DLHP form when BMP typically suppresses folding?

This is where a BMP antagonist becomes absolutely critical.

And that molecule is called noggin.

Remember, high levels of BMP signaling normally inhibit all hinge point formation.

They stabilize cell junctions, and they keep the cells nice and rectangular.

They say stay flat.

Stay flat.

But noggin is expressed specifically in the neural folds, and its whole job is to bind to and locally inhibit those repressive BMPs.

So noggin is the relief valve.

It flies in, turns down the stay flat signal from BMP, and that allows the cells to do their thing.

Exactly.

It relaxes the cell -to -cell junctions, and that allows the intrinsic apical constriction machinery to engage and form the wedge shape needed for the DLHP.

So the key isn't zero BMP, it's just the right amount of BMP.

That is the key conclusion from a lot of experiment.

Hinge point formation relies on finely tuned, intermediate amounts of BMP signaling.

We know this because if you experimentally activate BMP receptors throughout the entire neural plate...

No folding at all.

No folding.

You just get a flat sheet of rectangular cells.

But conversely, if you totally get rid of BMP signaling everywhere, you get ectopic and exaggerated folding.

The tissue just folds all over itself.

So the embryo needs just enough BMP to keep the sides of the plate stable, but just enough noggin to alleviate that repression right where the hinges need to be.

Perfectly said.

Specifies the MHP.

But noggin is the permissive factor that allows the DLHPs to form in the right location.

It's this beautiful subtle game of balancing molecular inhibition.

So once that bending is complete, the folds have to meet at the top, stick together, and fuse.

And you said this doesn't happen all at once.

No, it's a very dynamic event.

We know that the process generally proceeds from anterior to posterior, and the two open ends that remain are called the anterior neuropore and the posterior neuropore.

But the closure mechanism itself can vary a lot, right?

It varies by species and even by location along the axis.

In a chick, the closure starts right at the midbrain level, and it kind of zips up in both directions, rostrally and caudally.

And in mammals like us, it's way more complex.

Highly complex.

We now identify at least five distinct initiation sites for closure in the human embryo.

Five different zippers starting at once.

Where are these critical sites?

Well, closure site one is at the spinal chord hind brain junction, and that one zips in both directions like the chick.

But then you have site two at the midbrain -forebrain boundary, which is incredibly complex.

And site three is even farther up in the forebrain.

The point is, a failure at any one of these distinct sites can lead to very different and very severe birth defects.

Let's focus on the mechanics of that final zipping process.

How do two epithelial layers, which are pretty rigid, how do they find each other and adhere?

Live imaging has given us some incredible insights here, especially at that mammalian closure site two.

What we see is that the non -neural surface ectoderm cells, right at the leading edge, are incredibly active.

They're not just passive passengers.

Not at all.

They wrap around the edge of the closing neural folds and they extend these long exploratory feelers called filopodial processes.

You can picture them like tiny molecular hands reaching across the gap, searching for the opposing fold.

And they form temporary bridges.

Exactly, temporary cellular bridges.

So there's this element of dynamic cellular probing happening in mammals, literally breaking the gap before the final fusion.

What about in simpler organisms?

I know the ciona tunicate has a really clear mechanism.

It does.

It helps us see the physical forces at play.

In ciona, the mechanism is this dramatic posterior to anterior zipper,

and it's driven by a highly coordinated sequence.

What's the trigger?

It's triggered by the localized, really intense activation of actomyosin contraction in the apical membranes of the epidermal cells that are sitting just ahead of the closure point.

So a little pulse of contraction pulls it forward.

The sharp, momentary tension that pulls the junction forward, one step at a time.

This allows the transient epidermal to neural attachments to be replaced by stable, permanent epidermal to epidermal adhesion, and the neural tube closes just underneath the surface.

And once it's closed, it has to perform that crucial maneuver,

detaching from the surface ectoderm and sinking deeper.

Yes, and that separation is all about differential adhesion.

It's mediated by the famous e -cadherin to end -cadherin switch.

Ah, the molecular repulsion mechanism.

So the skin cells, the surface ectoderm, they start out expressing e -cadherin.

That's right.

But as the neural tube forms, the neural cells actively turn off the e -cadherin gene and start synthesizing end -cadherin.

And cells with different catherins don't like to stick to each other.

They generally repel each other.

They have a much higher adhesion affinity for cells that express the same catherin.

So end -cadherin cells stick to end -cadherin cells and e -cadherin to e -cadherin.

This difference provides the force needed for the neural tube to separate cleanly and sink.

It's such a beautiful, simple mechanism for sorting entire tissues.

And the experimental evidence must be pretty compelling for this.

It's rock solid.

If you experimentally force the surface ectoderm cells to express end -cadherin when they shouldn't, the separation is dramatically blocked.

The tissues just stick together like, well, like molecular Velcro.

And this is all under tight genetic control.

Of course.

We know specific transcription factors are required to orchestrate this switch.

A family called the grainy head, transcription factors are essential.

Grainy head like two in mammals is crucial for actively suppressing e -cadherin in the neural folds.

So mutations in those genes would cause big problems.

They do.

Mutations in genes like grainy head like three are strongly associated with severe neural tube closure defects.

Precisely because the cells fail to make that required adhesion switch.

And this brings us to the clinical relevance.

Because despite the elegance of all this machinery, it is incredibly sensitive to problems.

This leads to neural tube closure defects or NTDs, which are some of the most common major human birth defects.

And the location of the failure dictates the specific defect.

For instance, failure to close the posterior neuropore, usually site five, results in spina bifida.

And the severity of that can vary a lot.

It depends entirely on how much of the spinal cord and spinal nerves are exposed.

But the most catastrophic failures happen up at the rostral end.

At the head.

Yes.

Failure to close those rostral sites, two or three, leaves the anterior neuropore open.

This results in anencephaly.

And because the forebrain tissue is left in direct contact with the amniotic fluid, it subsequently degenerates.

That condition is almost always lethal.

And the most severe form?

Is if the entire tube fails to close along the body axis.

That's a condition termed cranioracichesis.

So we know these defects are rooted in a combination of genetics.

You said over 300 genes are implicated.

Right.

Genes like PAX3, the grainy head factors we mentioned.

And environmental factors.

Things like maternal diabetes, certain toxins, and most famously dietary deficiencies.

Yes.

And that environmental link brings us to one of the most significant public health discoveries in all of developmental biology.

And that is the profound connection between folate or vitamin B9 deficiency and NTDs.

This isn't just a correlation.

Oh, no.

We know that folate supplementation dramatically reduces the occurrence of NTDs.

But the hypothesis explaining why it works has really shifted the entire field of NTD research.

Because historically you might think, okay, folate is needed for making DNA for cell division.

A lack of it just means cells can't divide fast enough for closure.

Right.

That was the old model.

But the current focus is much broader and it's moved into the realm of epigenetics.

Okay.

So how does that work?

The current leading hypothesis is that environmental perturbations, including a lack of folate, directly modify the embryo's epigenome.

And the epigenome is that layer of regulation on top of the DNA sequence itself.

Exactly.

It's the set of chemical marks like DNA methylation and chromatin modifications that tells the cellular machinery when and how much a gene should be expressed.

So not having enough folate misses with the gene regulation.

Precisely.

Folate is a crucial cofactor for the enzymes involved in the methylation cycle.

It provides the little methyl groups that are attached to DNA.

If you have insufficient folate, you disrupt the highly sensitive timing and level of expression for all those hundreds of NTD causative genes.

The adhesion molecules, the hinge point regulators, everything.

This genetic variability, which is caused by an environmental factor modifying the epigenome, is what we believe leads directly to closure failure.

The recommendation for all women of childbearing age to take folate supplements is a direct public health response to this fundamental developmental vulnerability.

Okay.

We've spent a lot of time on the front ends folding failures.

Let's shift gears completely and go back to the tail end.

Secondary neuralation.

Right.

A completely different strategy.

So we're focused now on the tail bed region during caudal elongation.

And what you see is that mesenchame cells, which are derived from both ectoderm and mesoderm precursors, first condense under the surface ectoderm.

They just clump together.

Imagine a bunch of loose cells just aggregating tightly into a solid cylindrical rod.

This is the medullary cord.

And then the hollowing happens from the inside out.

Exactly.

The central portion of that solid cord undergoes cavitation.

Multiple small fluid -filled hollow spaces,

lumens,

just spontaneously appear inside the tissue.

Like bubbles forming.

Like bubbles forming.

And then those little cavities coalesce.

They merge together into a single continuous central cavity, which becomes a lumen of the spinal cord in the tail region.

So the surface ectoderm and the neural ectoderm, which were so intimately linked during the folding of primary neuralation, they're essentially uncoupled here.

That's a fundamental structural difference.

But what's really fascinating genetically about this posterior zone is the shared lineage of the cells.

The precursor cells in the tail can give rise to both the neural ectoderm, which is defined by expressing SOX2.

The neural marker.

And the paraxial mesoderm, which will form muscle and bone, defined by expressing TBI6.

So you have these two competing fates driven by two key transcription factors.

And they are fiercely antagonistic.

The gene TBI6 actively represses SOX2.

TBI6 is there specifically to ensure that the cells destined to become mesoderm do not accidentally adopt a neural fate.

So what happens if you take that repression away?

How quickly do the cells default to becoming neural tissue?

Incredibly fast.

The classic experiment is the TBI6 knockout mouse.

When TBI6 is deleted, the repression on SOX2 is lifted, and those cells just default to the highly stable neural fate.

And what's the result?

What does the embryo look like?

The result is this bizarre and amazing phenotype.

Instead of a single neural tube in the posterior, these mice develop three neural tubes.

Three.

The two rods of cells that should have become paraxial mesoderm, they lack the TBI6 repression, and they aberrantly become ectopic neural tubes right alongside the central intended one.

Wow.

That is just a striking demonstration of how critical gene repression is for maintaining boundaries.

If you lose the molecular guardrails, the default fate, which appears to be neural, just takes over.

It really does.

And this ability to generate both neural and mesodermal tissue from a common precursor pool is what really defines caudal development.

Okay, so while all that is happening in the back, let's switch our focus to the anterior tube.

You said even before the posterior tube closes, the anterior section starts to balloon out, establishing the brain.

Yes, this is the beginning of patterning along the anterior -posterior axis.

This ballooning forms the three primary brain vesicles.

Way up front, rostrally, you get the present cephalon.

That will become the massive forebrain, eventually the cerebral hemispheres.

In the middle.

In the middle is the mesencephalon, the midbrain.

It's involved in things like movement coordination, motivation.

And the most caudal of the three is the rhombencephalon, the hindbrain.

And that becomes the more primitive parts of the brain.

The core centers, yeah.

It gives rise to the cerebellum, the pons, the medulla oblongata.

These control all the involuntary activities, like breathing and heart rate.

And the forebrain, the prosencephalon, it quickly subdivides even further.

It does.

It splits into two secondary vesicles, increasing the complexity.

It divides into the telencephalon, which forms the cerebral hemispheres, and the deencephalon.

And the deencephalon is critical.

Incredibly critical.

It generates structures like the optic vesicles, which become your eyes, the pineal gland, and the really important sensory relay centers, the thalamus and the hypothalamus.

The hindbrain, the rhombencephalon, it undergoes a different kind of segmentation.

It forms these periodic swellings.

Yes, these are called rhombomeres.

And you can think of them as functionally separate territories or compartments.

While cells can move around freely inside their own rhombomere, they absolutely do not cross the boundaries into the next one.

And this physical segmentation must translate directly into functional specialization.

It does, through the expression of unique combinations of transcription factors in each one.

This segmentation is what ensures the precise wiring of the head.

How so?

Well, we see the very first motor neurons appearing specifically in the even -numbered rhombomeres, so R2, R4, and R6.

And this dictates the precise origin point for the cranial nerves.

Can you give an example?

Sure.

The neurons that originate in the ganglia of rhombomere 2, or R2, they form the trigeminal nerve, which is cranial nerve V.

That handles facial sensation and your shoeing muscles.

Neurons from R4 form the facial and vestibuloacoustic nerves.

This structural segmentation is how the nervous system guarantees the precise placement of motor neurons for all the structures of the head and face.

Okay, so we've established the AP axis, which gives us brain vesicles and spinal cord.

But the neural tube also needs to be polarized along the dorsal ventral, or DV, axis.

Absolutely.

Functionally, this axis defines where sensory information comes in and where motor commands go out.

Let's use the spinal cord as the model here.

How is it organized?

If you look at a cross -section, the dorsal region is where sensory neurons will eventually connect.

The ventral region is where the motor neurons live, the ones that project out to your muscles.

And the middle.

The middle is populated by various classes of inner neurons that relay information between them.

And all of these different specialized cells arise from precursor populations, progenitors, in the ventricular zone.

And these progenitors are defined by unique transcriptional ZIP codes.

Like Pax3 and Pax7 dorsally and NK by 6 .1 ventrally?

Exactly.

So the burning question is, how does a cell know if it should turn on Pax7 and become a dorsal progenitor, or turn on NK by 6 .1 and become a ventral one?

It has to be another gradient.

It's a powerful interaction of two opposing extrinsic morphogen gradients, which then set up intrinsic secondary signaling centers inside the neural tube.

Let's start on the ventral side, from the bottom up.

The ventral pattern is initiated, once again, by the notochord.

It's secreting high levels of sonic hedgehog.

This sheesh signal induces the adjacent neural tube cells to differentiate and become the floorplate.

And then the floorplate takes over.

The floorplate becomes the secondary signaling center.

It starts pumping out its own shh, creating a high to low ventral -to -dorsal gradient that permeates the entire ventral half of the tube.

So the notochord is the initiator, and the floorplate is the amplifier that maintains that ventral signal.

What about the dorsal side?

The dorsal pattern is initiated by TGFS proteins, specifically BMP4 and BMP7.

And these are secreted by the overlying surface epidermis, the skin.

So the signal comes from the tissue that just separated from the tube.

Exactly.

These BMPs induce a formation of the roof plate in the dorphel -most part of the neural tube.

The roof plate then becomes the dorsal secondary signaling center, and it secretes its own mix of TGF family members, BMPs, dorsalin, active and creating a gradient that runs high to low from dorsal to ventral.

So you have this elegant push -pull system, shh, coming from the bottom, BMPs coming from the top.

And a cell's position, and therefore its fate, is determined by the specific combination of those two opposing signals it receives.

Precisely.

And we can test this.

The experiments are classic.

If you remove the notochord from an embryo, no floor plate ever forms, and the neural tube completely lacks ventral cell types, like motor neurons.

And you add an extra one?

If you transplant an extra piece of notochord, or even just a little pellet of screening cells, and stick it to the side of the neural tube, you induce a brand new secondary floor plate, and a whole new cluster of ectopic motor neurons right there on the side.

The signal is clearly sufficient.

Now, here is where it gets, I think, really, really interesting.

Because a progenitor cell's identity isn't just determined by the concentration of the morphogen.

It's not just about its distance from the source.

It's also about the duration of the exposure.

This is a critical point.

The relationship is incredibly dynamic.

Let's first map out the concentration zones for shh.

The highest concentrations, right next to the floor plate, induce the most ventral cells.

The floor plate itself, V3 interneurons, which express NKiIs 2 .2.

Okay, a little further away.

A moderate level of shush induces the motor neurons, which express a gene called Olig2.

And then progressively lower and lower concentrations induce the increasingly dorsal interneurons, and ultimately, where there's basically no shh, you get the dorsal progenitors expressing PEC7.

That seems straightforward enough, a simple distance map.

But let's get into that key experiment that demonstrated the temporal dependence, the time value of Sonic Hedgehog.

This really challenges the simplicity of that concentration map.

It does.

It's a beautiful piece of work.

Researchers took neural tube explants, little pieces of intermediate neural tissue, and they exposed them to shh and dish.

First, they just confirmed the concentration effect.

More shh made them switch from a dorsal fate to a ventral fate.

Standard stuff.

Right.

But then they tested time.

They took these explants and exposed them to a constant intermediate concentration of shh over several days.

So if the concentration was constant, shouldn't the cell just pick one fate and stick with it?

That's what you'd think.

But it did not.

What they saw was that the cells first turned on the gene for the intermediate ventral fate, Olig2, which marks motor neuron progenitors.

But then with continuous longer exposure to the exact same concentration of shh, they switched off Olig2 and transitioned to expressing the most ventral marker, NKI2 .2.

Whoa.

So the cell's final fate wasn't fixed by the concentration it saw at one moment.

It was a cascade, and it was driven by how long that exposure lasted.

More time at the same concentration meant a more ventral fate.

That's it.

It's like the cell needs to bank a certain number of signaling hours before it commits to that deepest ventral identity.

The cell's internal machinery must be able to measure both the concentration and the duration of the shh signal.

That just adds this whole other layer of robustness to the patterning.

It does.

But it also raises a major mechanical issue.

We know that the down -frame effectors of shh signaling, the Glee activators, they actually decrease in concentration over time in the neural tube.

So if the primary signal is fading,

how do the cells maintain these incredibly sharp, refined boundaries between the different progenitor domains?

Logically, the boundary should blur as the signal fades out.

How do they avoid that blurring?

They avoid it through the genius of the intrinsic gene regulatory network they've established, specifically through transcriptional cross -repression.

Okay, what does that mean?

It means that the adjacent progenitor domains, for instance the OLED2 -expressing domain right next to the NKX2 .2 -expressing domain, they express transcription factors that are mutually inhibitory.

Like a molecular stalemate, a tug of war at the border.

Precisely.

OLED2 actively represses the NKX2 .2 gene, and NKX2 .2 actively represses the OLED2 gene.

This mutual antagonism creates and maintains a razor -sharp, distinct border between them.

Even as the initial signal that set up these domains begins to fade,

this cross -repression network stabilizes the pattern.

It's how the cell remembers the positional information it received, both spatially and temporally.

It locks it in.

It locks it in, maintaining that intricate DV organization long after the initial morphogen wave has passed.

So we've established the AP axis, which gives us the big structures like brain vesicles, and the DV axis, which gives us the functional cell types.

But these two axes have to be coordinated.

How does the body make sure DV patterning only kicks in when the tissue is structurally ready for it?

This coordination is most evident in the caudal region during secondary neuralation.

And it's all about the unique behavior of a cell population called the neuromesodermal progenitors, or NMPs.

Right.

These are those bi -potential cells we met earlier.

They can become either neural tissue or mesoderm.

Correct.

They live in the caudal lateral epiblast, right in the tail bud.

And their decision to either keep proliferating or to mature and differentiate is governed by another set of antagonist ingredients, this time established along the rostral caudal axis.

So what are the competing signals in the tail bud?

Well, the proliferative immature state, the maintenance of that NMP stem cell pool, is driven by high levels of FGF8 and Wnt signaling.

These are concentrated way back in the caudal end of the tail bud.

And what's opposing them?

Opposing them is redenoic acid, or RA.

RA is secreted by the already maturing semitic mesoderm, which is more rostral, more anterior,

and RA actively inhibits FGFW signaling.

So you have a high FGFWMT zone at the very tail, keeping things immature and dividing, and a high RA zone farther up, signaling maturity.

It creates a functional roadmap.

Exactly.

Or a competence gradient.

As those NMP cells effectively move rostrally, which means they're moving away from the high FGFWMT source and into a higher concentration of RA, they undergo crucial change in their state of competence.

And what does competence mean in this context?

It means the cell gains the necessary internal machinery, the right receptors, the right cofactors, the right transcriptional environment, to actually read and respond to the next set of signals.

When they're highly proliferative NMPs, they might not even have the right receptors to fully interpret a shh or a TGS signal.

But as they travel along that rostral caudal gradient, they mature and they gain that competence.

Yes.

This repositioning, which is triggered by the drop in FGFWMT and the rise on RA, it switches these preneural progenitors into a state where they are finally competent to respond to the NGGF signals.

So DV patterning, the thing that defines motor neurons versus sensory neurons, it only gets established correctly once the tissue has been properly matured by the rostral caudal signaling axis.

That's the beautiful integration of the two axes.

The rostral caudal axis dictates the timing and maturity, the when, and it sets the stage.

That allows the dorsal ventral signals to then come in and dictate the final identity, the what.

It's a perfectly controlled cascade that ensures structural development precedes functional specialization.

That really does bring the whole organization together.

An incredible process.

Which brings us to the close of our deep dive into neurulation.

Let's maybe quickly recap the central tenets we've uncovered today.

Let's do it.

First lesson is mechanical.

Building the CNS requires two totally different strategies.

Primary neurulation with its folding for the brain, and secondary neurulation with its condensation and hollowing for the tail.

And that folding relies entirely on intrinsic cell shape change.

The apical constriction happening in those key hinge points, which are positioned by signals from the notochord and the ectoderm.

Secondly, we saw that closure is this incredibly delicate process that demands the precise E -cadherin to N -cadherin switch so the tube can separate from the skin.

And we highlighted that critical link between this process and the epigenome, and how things like folate deficiency can modify DNA methylation and lead to failures like spina bifida.

Third, the overall blueprint.

It's established by the AP axis, creating the brain's primary vesicles and the segmented hindbrain, the rhombomers, which guarantee that precise placement of the cranial nerves.

And fourth, the functional organization.

It all relies on that beautifully balanced battle of opposing morphogen gradients along the DV axis.

SH from the floor plate on the bottom and TGS from the roof plate on the top.

And maybe the most sophisticated finding of all was that time dimension.

Cell identity isn't just about the concentration of SHU, but about the crucial duration of exposure, a pattern that's then locked in by that transcriptional cross repression between the progenitor domains.

It's a system built on complexity, memory, and spatial dynamics.

Which leads to a final provocative thought.

We've mostly assumed that the cells are static, that they're just sitting still in the gradient waiting for their signal.

But some of the newer research, especially in organisms like zebrafish, suggests that these progenitor cells are actually highly dynamic.

They're moving and sorting themselves within that morphogen field.

So the cell isn't just sitting still.

It might be actively surfing the waves of SH and BMP.

How does that physical movement, that fourth dimension, change how a cell interprets its positional identity?

Exactly.

It makes the self -assembly of the central nervous system not just elegant,

but perpetually in motion.

It brings us right back to those mysterious butterflies of the soul.

Recognizing that the construction of this incredibly complex machine starts with movements and signals of really profound simplicity.

Well said.

Thank you for joining us on this deep dive into the fundamental building blocks of the central nervous system.

We'll catch you next time.