Sheddases

Any protease that cleaves a protein to release the extracellular domain can be said to have sheddase activity.2 ,4 Sheddases are also known as secretases, since they lead to the release of proteins and peptides from the cell surface. However, sheddase activity may not always be physiologically or pathologically relevant, and could be the result of bystander activity in unusual circumstances, for example, a number of the proteolytic cleavage events seen in emphysema may not be biologically significant or contribute to the pathology.

The first ‘true’ sheddase characterised was the transmembrane enzyme responsible for cleaving and releasing soluble, active tumour necrosis factor α (TNFα) from its inactive, transmembrane precursor. This TNFα-converting enzyme is also known as ADAM17. The ADAMs (a disintegrin and metalloproteinase) are a family of transmembrane proteases that are considered to be a major class of sheddases. Another family of metalloproteinases, the matrix metalloproteinases (MMPs), also play essential roles in ectodomain shedding. Some members of this family, the membrane type MMPs (MT-MMPs), directly associate with the plasma membrane as transmembrane proteins or via lipid anchors. Secreted MMPs may also reside at the cell surface through binding to other cell surface proteins.

The amino acid sequence at the cleavage site of a sheddase-susceptible protein appears not to determine the site's position. It is believed that distances from the plasma membrane and the secondary structure are the important determinants. There is also a degree of promiscuity in ectodomain shedding in that an individual sheddase may act on many proteins, and an individual membrane protein may be cleaved by a number of sheddases. For example, ADAM17 can cleave TNFα, TNFα receptors, APP, L-selectin and others, and the epithelial intercellular junction adhesion molecule, E-cadherin, can be cleaved by ADAM10, ADAM15, MMP7, MMP9 and others. In addition to the endogenous sheddases, pathogen-derived protease activity can lead to ectodomain shedding, which may play an important role in disease progression.

Consequences of ectodomain shedding

Clearly, the function of the original protein will determine the activity of shed ectodomains (figure 1).1 ,2 ,5 The ectodomain may be inert (eg, the cleaved Notch ectodomain is internalised and degraded) or have significant biological activity. Downregulation of a functional receptor or adhesion molecule at the cell surface is an effect in itself. Ectodomain shedding of active growth factors from cell surface precursors will result in increased signalling, while shed growth factor receptor ectodomains will reduce signalling by acting as decoys that bind and neutralise extracellular growth factors. Similarly, soluble cell adhesion molecule ectodomains may interfere with cell–cell adhesion by binding to intact cell adhesion molecules at the cell surface.

Fate of the remnant peptide

The short transmembrane remnant that remains after ectodomain shedding (the C-terminal fragment (CTF) for type 1 proteins) may be internalised by endocytosis (figure 1C) or may undergo further cleavage (figure 1D).1 ,5 After ectodomain cleavage, some remnant peptides become susceptible to regulated intramembranous proteolysis, mediated by intramembrane-cleaving proteases, for example, γ-secretase, to produce an intracellular domain (ICD) that is released into the cytoplasm (figure 1D). The internalised CTF and the ICD can be targeted to intracellular compartments where they influence cell function. For example, the basis of Notch signalling is that the ICD enters the nucleus and affects gene transcription. In this issue, Mimae et al3 show that the CTF of cell adhesion molecule 1 (CADM1) is targeted to, and disrupts, mitochondria, inducing apoptotic cell death.

Regulation of ectodomain shedding

Although some proteins might be continuously shed at low levels, ectodomain shedding is typically subject to regulation.6 This regulation may be at the level of expression of sheddases and their targets, and also their targeting to distinct plasma membrane domains. ADAM and MMP activities can be modulated by tissue inhibitors of MMPs (TIMPs). Sheddases themselves can be activated or degraded by other proteases. However, cell signalling events are known to alter ectodomain shedding. This may be the result of sheddase activation or a change to the target protein making it more susceptible to cleavage. The details of these control mechanisms are not clear, though Ca²⁺/calmodulin activity, protein kinase C (PKC) activation and tyrosine phosphorylation have all been implicated in inducing ectodomain cleavage, possibly by promoting specific sheddase/target interactions. In Notch signalling, it is believed that the binding to its ligand on an adjacent cell leads to exposure of a previously cryptic cleavage site. It is not clear if this type of control is used by other proteins. Some pathogens stimulate ectodomain shedding indirectly (ie, not by direct proteolysis). For example, the Pseudomonas aeruginosa virulence factor, LasA, is a protease that does not directly stimulate ectodomain shedding, but stimulates host sheddase activity that results in the cleavage of a host protein, syndecan-1. This heparan-sulfated syndecan-1 ecotodomain can bind and neutralise chemokines involved in lymphocyte cell recruitment to the lung, thereby blunting the immune response.

Ectodomain shedding in pathophysiology

Given that it is central to a number of fundamental signalling and adhesion processes, it is not surprising that ectodomain shedding is important in many pathologies. The need to dissect the molecular and cellular mechanisms of the proteolytic processing of APP that produce amyloid in Alzheimer's disease is a major driving force in understanding ectodomain shedding. Similarly, its role in immunity and inflammation implicates shedding in many pathological states. Normal pulmonary and cardiovascular development, cancer and other pathologies are strongly dependant on growth factor signalling, which is modulated by ectodomain shedding. Proteolytic cleavage of cell-surface adhesion molecules is increasingly linked with tumour progression and metastasis, and plasma levels of ectodomains can act as prognostic markers.

There are numerous examples of the importance of ectodomain shedding in the lung. In mouse models, knocking out the sheddase, ADAM17, has strikingly similar effects on lung development to knocking out TNFα, EGF or their receptors. The impaired airway branching and pulmonary epithelial defects observed emphasise the importance of ectodomain shedding in the normal control of growth factor activity. In chronic lung infection, lipoteichoic acid, released from Staphylococcus aureus, induces the activation of ADAM10, which cleaves and sheds EGF, which in turn stimulates mucus production, contributing to airway obstruction. Genome-wide association studies implicate ADAM33 in asthma and airway hyper reactivity. The mechanisms by which it acts are not clear, but ADAM33 can cleave and shed CD23, the FcɛRII low-affinity IgE receptor, and is therefore likely to influence responses to allergens. In addition, transforming growth factor β, a key factor in tissue remodelling that is upregulated in asthma, stimulates ectodomain shedding of ADAM33, which may deregulate its proteolytic activity and contribute to the remodelling process seen in asthma. MMP7 (matrilysin) is upregulated in lung fibrosis when it increases ectodomain shedding of E-cadherin from epithelia, leading to loss of epithelial coherence, and increased migration by epithelial cells, part of the epithelial repair response found in fibrosis. Emphysema is a disease for which protease activity is elevated in the lung, and it is likely that ectodomain shedding could play a significant part in its pathology. Mimae et al3 show that the ectodomain shedding of CADM1 in pulmonary epithelial cells results in apoptosis, and the elevated CADM1 shedding that they observed in emphysematous patient samples suggests that shedding might be a central mechanism in the loss of alveolar septa.