Every time I mention this developmentally significant molecule, Sonic hedgehog, I get a volley of questions about whether it is really called that, what it does, and why it keeps cropping up in articles about everything from snake fangs to mouse penises to whale fins to worm brains. The time seems appropriate to give a brief introduction to the hedgehog family of signaling molecules.
First, a brief overview of what Sonic hedgehog, or shh, is, which will also give you an idea about why it keeps coming up in these development papers. We often compare the genome to a toolbox — a collection of tools that play various roles in the construction of an organism. If I had to say what tool Sonic hedgehog is most like (keeping in mind that metaphors should not be overstretched), it would be like a tape measure. It’s going to have multiple uses: as a straightedge, as a paperweight to hold down your blueprints, as something to fence with your coworkers on a break, and even to measure distances. It will be pulled out at multiple times during a construction job, and it’s generically useful — you don’t need one tape measure to measure windows, another to measure doors, and yet another to measure countertops. Sonic hedgehog is just like that, getting whipped out multiple times for multiple uses during development, often being used where structures need to be patterned.
Let’s dig into some of the details. I’m using the 2006 review by Ingham and Placzek for most of this summary, so if you really want to get deeper into the literature, I recommend that paper as a starting point.
How do the molecules work?
Hedgehog molecules are signaling proteins. What that means is that these are small molecules that in this case are secreted into the extracellular environment, where they can diffuse and bind to receptors on nearby cells and trigger changes in gene expression in the target. The specific properties of Hedgehog’s transport and diffusion can be modulated by the attachment of lipid molecules to it, which makes it a rather more flexible kind of signal.
What Hedgehog does is to activate an evolutionarily conserved pathway. Hedgehog binds to a receptor protein embedded in the membrane of cells competent to receive the signal; this receptor is called patched, or PTC for short. PTC then releases an inhibition of a cytoplasmic protein called smoothened (SMO), which then modifies the activity of glioma-associate oncogene homologues (GLI) that enter the nucleus and bind to various genes to switch them on and off. I know. The names don’t help at all. Students just have to memorize them. Here’s a short summary of the Rube Goldberg pathway involved here:
Hedgehog (HH) → Patched (PTC) → Smoothened (SMO) → GLI transcription factors → differential gene activity
For you visual thinkers, here’s a diagram. First, the inactive state, in which no Hedgehog is around:
See? Just like I told you, only more complicated. One difference: This is a diagram of a fly cell, and the fly homolog of GLI is a gene called Cubitus interruptus, or CI, which is modified in two different ways to form CIR and CIA in the diagram.
Below, the state when Hedgehog (HH) is bound to the receptor is illustrated, with the end result being a set of Hedgehog target genes being switched on. Also illustrated here is another set of proteins that facilitate HH binding, with the lovely names of Interference hedgehog (IHOG) and Brother of Interference hedgehog (BOI). Trust me, it gets more complicated the more closely you examine it.
(Click for larger image)
Hedgehog (HH) proteins signal by binding to the transmembrane proteins Interference hedgehog (IHOG) and Brother of IHOG (BOI), which facilitates the interaction of HH with the multipass transmembrane protein Patched (PTC). This interaction relieves the repressive effect of PTC on the serpentine protein Smoothened (SMO). Like PTC, SMO is an obligate component of the HH pathway, being required for all aspects of HH signal transduction that have so far been described. In Drosophila melanogaster, SMO becomes hyperphosphorylated in response to HH signalling and accumulates in the plasma membrane, whereas in vertebrate cells, the protein localizes to primary cilia following exposure to HH ligands. Once activated, SMO modulates the activities of members of the Glioma-associated oncogene homologue (GLI) transcription-factor family. In D. melanogaster, there is just one GLI-family protein, named Cubitus interruptus (CI), which is crucial for HH signalling. CI is a bifunctional transcription factor with both repressor and activator domains that flank a central DNA-binding zinc-finger domain. In the absence of HH signalling, CI undergoes proteolytic cleavage, which is primed by its phosphorylation by three kinases, Protein kinase A (PKA), Glycogen synthase kinase 3 (GSK3) and Casein kinase 1 (CK1), and mediated by the ubiquitin ligase pathway; this yields a truncated form of the protein that acts exclusively as a repressor of HH target-gene transcription (CIR). Activation of SMO suppresses CI cleavage and promotes the nuclear import of a full-length CI protein (CIA); the resulting depletion of the truncated form of CI relieves the repression of some HH target genes, and the full-length CI protein further enhances their transcription. CI is present in a complex with the COS2 scaffold protein, which recruits PKA, GSK3 and CK1, facilitating phosphorylation of CI on residues that are crucial for its cleavage. COS2 binds directly to the intracellular C-terminal tail of SMO, thereby providing a physical basis for the regulatory interaction between SMO and CI. Exactly how SMO activation abrogates CI processing is unclear, but one clue comes from the dependence of SMO activation on its hyperphosphorylation. Notably, the phosphorylation sites that are essential for SMO activity resemble those that are phosphorylated by PKA, GSK3 and CK1 in CI, leading to the suggestion that the phosphorylated C-terminal tail of SMO might compete with CI for a binding partner that mediates its cleavage. CI also interacts with the Fused (FU) serine threonine kinase that abrogates its sequestration in the cytoplasm by the Suppressor of fused (SU(FU)) protein. A negative-feedback loop is initiated when the BTB/rdx proteins, targets of HH signalling, degrade CI.
Hedgehog as a morphogen
From the diagram above, you might think Hedgehog is limited — it can switch genes on and off. What isn’t shown is that the effect of the protein is dependent on concentration and timing — it isn’t just an on/off switch, it’s an analog switch. As anyone who has done any circuit design knows, this is incredibly useful. Just by increasing the concentration of Hedgehog — turning the dial up — you can get activation of more and more target genes, and stronger expression as well. In addition, cells respond progressively to longer exposures to Hedgehog with greater activation, adding another layer of control.
This is where the comparison to a tape measure is appropriate. Organisms can set up gradients of Hedgehog, and cells along the gradient can measure off where they are, and they switch genes on and off along that gradient in a reliable pattern. What this means is that one signal can be used to generate multiple different outputs, producing the ability to generate patterned distributions of cells. The models below illustrate the multiple ways Hedgehog is able to generate different patterns of gene expression.
(Click for larger image)
Signals can specify cell fate in different ways. Blue circles represent the signal-secreting cells; other circles represent the responding cells. In a, the secreting cell establishes a spatially graded distribution of signal across the adjacent field of cells. Each cell adopts a different identity (indicated by different colours) in response to different levels of the signal. This is the classical morphogen paradigm. In b, cell number increases vertically with time. As the secreting cell divides, one of its progeny stops expressing the signal and adopts an identity that is determined by the length of time for which the parental cell has expressed the signal. The integration of signal level over time is represented by the increase in the shading intensity of the secreting cell. Note that the sequence of different cell identities that are generated in this way is the same as in a. In c, only two different identities are specified in response to the signal (arrows): initially, all cells respond to the signal (red circles), but as the cellular field expands through proliferation, some cells move out of range of the signal and adopt a different identity (white circles). In d, the signal acts over a short range to induce two different cell identities; this is accomplished by the movement of one cell away from the source, allowing a second cell that has a different potential or competence to be exposed to the same signal. In e, the competence of cells to respond to the same signal changes over time. In f, cells in different locations respond differently to the same signal.
Patterning the limb
Schematic representation of Wolpert’s spatial morphogen gradient model (4). The hypothetical threshold values postulated to specify the digits are indicated in red. Digits are numbered so that the most anterior digit (the thumb) is 1 and the most posterior one (the little finger or pinky) is 5. The ZPA in the posterior limb bud mesenchyme is indicated in dark blue, and the limb skeletal elements are shown only schematized.
One example of Hedgehog patterning is right in your hand. In early development, the limb begins as a digitless bud, a small protrusion from the body wall that only has polarity. That is, the posterior side of the bud, where the prospective pinky (digit 5) will form, has an area called the Zone of Polarizing Activity (ZPA) that secretes — you guessed it — a Hedgehog protein, specifically a vertebrate form called Sonic hedgehog. And yes, it was named after the Sega video game character. When the gene was discovered, it was identified as a member of the Hedgehog, so it was given an arbitrary name to make its affinity clear. For you Beatrix Potter fans, there is also a Tiggywinkle hedgehog.
The location of the ZPA sets up a gradient of Sonic hedgehog across the limb bud, from very high where the pinky will form, to very low where the thumb will form. Cells across the limb can read that gradient and have different patterns of gene activity turned on, which is what leads to the subsequent development of different digits.
We can also see the different ways Sonic hedgehog regulates the digits. As the diagram below illustrates, the posteriormost digits actually don’t look at concentration — their identity is set by how long they’ve been exposed to shh. The next digits are concentration dependent, and the thumb is the default state — low or no shh specifies the development of the first digit.
Schematic representation outlining the mechanisms by which Shh contributes to each individual digit in the mouse limb. Digit 1 is not reliant on Shh, at least in the hindlimb. Cells comprising digit 2 never expresses Shh itself, though they are reliant on secretion of Shh from producing cells (paracrine signaling). Approximately half of digit 3 and all of digits 4 and 5 are composed of cells that have expressed Shh (autocrine Shh signaling), and differential identity is thought to be shaped dependent on the length of time cells are exposed to high-level autocrine signaling. Both paracrine and autocrine Shh signaling contributes to digit 3 formation.
Patterning the nervous system
The classic place to find Sonic hedgehog expressed is the central nervous system (CNS) — this was where the importance of the gene product in vertebrate development was first identified. A structure called the floorplate, a band of cells that extends the length of the nervous system on the ventral floor, secretes Sonic hedgehog. Another band of cells, the roofplate, on top of the CNS, secretes another set of signaling molecules, the bone morphogenetic proteins (BMPs) (which have many more roles than simply controlling bone development — it’s too bad they weren’t named after a Nintendo sprite). These two complementary gradients of Shh and BMP generate unique molecular addresses for any cell along the dorsoventral axis, and enable the differentiation of many different cell types. This is a case of two simple inputs generating a multitude of outputs.
(Click for larger image)
Control of Commissural Axon Guid- ance by Patterning Molecules (Left) At early stages, BMPs (purple) from the roof plate (RP) control the generation of distinct neuronal subtypes in the dorsal spinal cord (DL1-DL6), whereas Shh (green) from the floor plate (FP) acts to pattern the ventral spinal cord (V0-V3, MN). (Right) At later stages, commissural axon guidance is influenced sequentially by (1) BMPs (purple) from the RP to direct axons ventrally, (2) combinatorial activities of Netrin-1 and Shh (green) to attract axons to the FP, and (3) Wnt4 (yellow) expression in a caudal-low rostral-high gradient to control rostral turning toward the brain. BMP and Shh signals are shown only on the cross- sectional face of the three-dimensional schematic drawing of the spinal cord but are expressed along the entire rostrocaudal length of the spinal cord. Wnt1 signals (blue) derived from the roof plate act to instruct neural crest (NC) cells to acquire DRG sensory neuron identity. Dorsoventral and rostrocaudal axes are indicated by black arrows.
Hedgehog is a morphogen, a molecule that defines cell identity by its concentration, but it is also a mitogen. Mitogens regulate cell activity, so Shh can also signal adjacent cells to proliferate.
Sonic hedgehog downregulation has also been associated with some cancers, especially of the skin and brain. As you might expect, a gene that can regulate cell type and division might be implicated in errors of cell growth and identity…cancers. This gene is not just of concern in embryos, but is important in you right now for maintaining your cellular organization.
The developmental pathways in your immune system are regulated in part by Shh.
Shh is one of the key signaling molecules in generating left-right asymmetries in development. Your heart is skewed to the left and your appendix is on the right because of a pattern modulated by Hedgehog signaling.
In a role that has to be useful in getting funding from aging, balding Republicans, Shh is also important in regulating the organization of smooth muscle and endothelium in a structure that is a complicated maze of spongy tubes and smooth muscle — the penis. Abnormal Shh expression has also been found in men with erectile dysfunction.
Expect to see Sonic hedgehog and other Hedgehog family members to come up frequently in any discussion of the molecular genetics of animals — they are ubiquitous (except, strangely enough, they are absent in C. elegans) and they have utility in many roles. As I discussed yesterday, they appear in the most surprising roles, as in patterning the distribution of teeth, and seem to be expressed in almost any place where the organism is measuring out and partitioning and organizing a pattern of cell fates.
Arsic D, Beasley SW, Sullivan MJ. (2007) Switched-on Sonic hedgehog: a gene whose activity extends beyond fetal development—to oncogenesis. J Paediatr Child Health 43(6):421-3.
Crompton T, Outram SV, Hager-Theodorides AL (2007) Sonic hedgehog signalling in T-cell development and activation. Nat Rev Immunol 7(9):726-35.
Ingham PW, Placzek M (2006) Orchestrating ontogenesis: variations on a theme by sonic hedgehog. Nat Rev Genet 7(11):841-50.
McGlinn E, Tabin CJ. (2006) Mechanistic insight into how Shh patterns the vertebrate limb. Curr Opin Genet Dev 16(4):426-32.
Podlasek CA, Meroz CL, Korolis H, Tang Y, McKenna KE, McVary KT (2005) Sonic hedgehog, the penis and erectile dysfunction: a review of sonic hedgehog signaling in the penis. Curr Pharm Des 11(31):4011-27.
Salie R, Niederkofler V, Arber S. (2005) Patterning molecules; multitasking in the nervous system. Neuron 45(2):189-92.