The mysterious biochemical network that determines how one cell in the developing embryo becomes muscle whereas an adjacent cell becomes skin or bone, is beginning to be revealed. In the 31 August issue of Science, Emerson and colleagues describe how the main function of the enzyme QSulf1 is to modify the signalling co-factors, heparan sulphate proteoglycans (HSPGs). This allows a cell to respond to molecular signals, such as Wnt, and transform into muscle instead of skin or bone.

The researchers highlight how the discovery of QSulf1 sheds light on mechanisms that can regulate responses to developmental signals for embryo patterning. Although the developmental signalling molecules that control embryo patterning for body-plan specification are well known, the mechanisms that regulate the spatially localized responses to these signals within the developing embryos are less understood.

One candidate for embryo-patterning regulation was HSPGs, as they are localized to the cell surface where they influence diverse developmental signals. In addition, the sulphation states of N-acetyl glucosamine residues in heparan sulphate moieties of HSPGs influence their activities in fibroblast growth factor (FGF) and Wnt signalling, indicating that HSPG sulphation might regulate developmental signalling. However, the mechanism through which this could be achieved was unclear.

The researchers first identified the enzyme QSulf1 by screening quail embryos for genes that are expressed when Sonic hedgehog (Shh) signalling is activated — precisely at the time when presomitic mesoderm epithelializes to form somites. QSulf1 is part of an evolutionarily conserved protein family that is related to heparin-specific N-acetyl glucosamine sulphatases. In situ hybridization studies showed QSulf1 was coexpressed with the muscle-specification genes Myf5 and MyoD in the Shh-responsive epaxial muscle progenitors — which give rise to the deep-back and intercostal muscles of the adult — of newly formed somites.

The function of QSulf1 was investigated by antisense inactivation of QSulf1. This resulted in the specific inhibition of MyoD, but not Myf5, in the epaxial somite muscle progenitors and did not disrupt expression of Pax3 or Pax1 in the ventral somite. As Myf5, Pax3 and Pax1 are Shh-response genes, the researchers concluded that QSulf1 does not function in Shh signalling.

However, MyoD is Wnt-inducible, indicating QSulf1 might be involved in Wnt signalling, which is controlled by HSPGs — the likely substrate of QSulf1 activity. Co-transfection studies showed that QSulf1 is localized on the cell surface, and reporter gene assays in C2C12 myogenic progenitor cells found that QSulf1 regulates HSPG-dependant Wnt signalling through a mechanism that requires its catalytic activity, providing evidence that QSulf1 regulates Wnt signalling through desulphation of cell-surface HSPGs.

The researchers propose that QSulf1 could function in a two-step mechanism to regulate HSPG-dependent Wnt signalling. Wnts in the extracellular matrix could bind widely to heparan sulphate moities on cell-surface HSPGs, but only cells expressing QSulf1 on their surface would desulphate heparan sulphate to locally release HSPG-bound Wnts. This would allow Wnt to activate regulatory genes, such as MyoD, which then instructs these cells to become muscle progenitor cells instead of skin or bone progenitor cells.