EGF-like Domain - Subtypes

Subtypes

Despite the similarities of EGF-like domains, distinct domain subtypes have been identified. The two main proposed types of EGF-like domains are the human EGF-like (hEGF) domain and the complement C1r-like (cEGF) domain, which was first identified in the human complement protease C1r. C1r is a highly specific serine protease initiating the classical pathway of complement activation during immune response. Both the hEGF- and cEGF-like domains contain three disulfides and derive from a common ancestor that carried four disulfides of which one was lost during evolution. Furthermore, cEGF-like domains can be divided in two subtypes (1 and 2) whereas all hEGF-like domains belong to one subtype.

The differentiation of cEGF-like and hEGF-like domains and their subtypes is based on structural features and the connectivity of their disulfide bonds. cEGF- and hEGF-like domains have a distinct shape and orientation of the minor sheet and one C-terminal half-cystine has a different position. The lost cysteines of the common ancestor differ between cEGF- and hEGF-like domains and hence these types differ in their disulfide linkages. The differentiation of cEGF into subtype 1 and 2, which probably occurred after its split from hEGF, is based on different residue numbers between the distinct half-cystines. An N-terminal located calcium binding motif can be found in hEGF- as well as in cEGF-like domains and is therefore not suitable to tell them apart.

hEGF- and cEGF-like domains also contain post-translational modifications, which are often unusual and differ between hEGF- and cEGF-like domains. These post-translational modifications include O-glycosylations, mostly O-fucose modifications, and β-hydroxylation of aspartate and asparagine residues. O-fucose modifications have only been detected in hEGF-like domains and they are important for the proper folding of the hEGF-like domain. β-Hydroxylation appears in hEGF- and cEGF-like domains, the former is hydroxylated on an aspartic acid while the latter is hydroxylated on an asparagine residue. The biological role of this post-translational modification is unclear, but mice with a knockout of the aspartyl-β-hydroxylation enzyme show developmental defects.

Proteins containing EGF-like domains are widespread and can be exclusively hEGF- or cEGF-like, or contain a mix of both. In many mitogenic and developmental proteins such as Notch and Delta the EGF-like domains are only of the hEGF type. Other proteins contain only cEGF such as thrombomodulin and the LDL-receptor. In mixed EGF-proteins the hEGF- and cEGF-like domains are grouped together with the hEGFs always being N-terminal of the cEGFs. Such proteins are involved in blood coagulation or are components of the extracellular matrix like fibrillin and LTBP-1 (Latent-transforming growth factor beta-binding protein 1). In addition to the aforementioned three disulfide hEGF- and cEGF-like types, there are proteins carrying a four-disulfide EGF-like domain like laminin and integrin.

The two main EGF-like domain subtypes hEGF and cEGF are not just distinct in their structure and conformation but also have different functions. This hypothesis is substantiated by research on LTBP-1. LTBP-1 anchors the transforming growth factor β (TGF-β) to the extracellular matrix. hEGF-like domains play a role in targeting the LTBP-1/TGF-β assembly to the extracellular matrix. Once attached to the extracellular matrix, TGF-β dissociates from hEGF subunits to allow its subsequent activation. cEGF-like domains seem to play an unspecific role in this activation by promoting the cleavage of LTBP-1 from TGF-β by various proteases.

In conclusion, although distinct EGF-like domains are grouped, subtypes can be clearly separated by their sequence, conformation and, most importantly, their function.