Ectodomain cleavage by metalloproteases, followed by gamma-secretase cleavage of the intracellular domain, generates two signaling molecules.

Ectodomain cleavage by metalloproteases, followed by gamma-secretase cleavage of the intracellular domain, generates two signaling molecules.

Only two a-disintegrin-and-metalloproteases, ADAM10 and 17, cleave most physiologically and disease-relevant substrates (>100).  A process termed ectodomain cleavage or ectodomain shedding. This explains why metalloprotease inhibitors have failed clinically in part due to negative side effects induced by broadly inhibited substrate cleavage.  Regulated substrates include e.g. the precursors of growth factors like epidermal growth factor (EGF) ligands, cytokines (TNFα), as well as adhesion molecules (e.g. CD44, ICAM-1).  Interestingly, also many of the involved growth factor or cytokine receptors undergo regulated shedding, allowing the possibility of negative regulation on the receptor level or the production of decoy receptors.  After ectodomain shedding, the subsequent release by γ-secretase of the intracellular-domains from proteins such as Notch, NRG1, Erb4 or CD44, which regulate transcriptional activity in the nucleus, provides further levels of complexity.  Ectodomain shedding therefore affects numerous cellular processes, including for example growth, adhesion and motility of cells, throughout life.  A complex interplay between cellular signaling pathways and the shedding of bioactive molecules determines the ultimately expressed phenotype of the cell.  The broad array of substrates and biological phenotypes regulated, illustrates the need for tight control.  Unfortunately it is largely unknown, how ectodomain cleavage is regulated and made substrate-specific.  Furthermore, extensive cross-talk and complexity among signaling networks, proteases, and their substrates make understanding sheddase regulation on a component-by-component basis challenging (2).  Therefore, in order to develop improved therapeutics, a strong need exists to understand how the balance of sheddase-mediated proteolysis integrates multiple layers of signaling networks to coordinately influence cell behavior in various disease contexts. 

To meet these challenges, we take a systemic view in our approach to investigate cleavage regulation.  Our research aims to delineate the signaling networks and molecular mechanisms that determine substrate specificity of cleavage by combining (1) in vitro and in vivo approaches aimed at identifying the regulated cleavage mechanism and its physiological significance in disease models, with (2) integrated biocomputational network analysis of a large dataset relevant to cleavage regulation.  This approach aims at identifying and validating specific signaling nodes that regulate the cleavage of specific substrates and relevant phenotypes.  We generated this extensive and still growing dataset with a large scale shRNA screen targeting all human kinases and phosphatases, testing their effect on the induced cleavage of the EGF ligand TGFα.  A core set of candidate regulators (activators and inhibitors), some of which regulate cleavage on the substrate level without affecting metalloprotease activity (Dang et al. PNAS 2013), was then tested for their effects on cellular proliferation, migration, invasion, apoptosis and the cleavage of a large set of relevant ADAM10 and ADAM17 substrates.  This dataset and analysis is enriched by our findings on specific cleavage mechanisms of substrates (e.g. Dang et al PNAS 2013, Hartmann et al. Mol Cancer Research 2015; Hartmann et al. JBC 2015). 

The majority of our current work focuses on breast cancer and kidney disease applications since both present a critical medical and societal need, and both critically involve the shedding of epidermal growth factor (EGF) ligands.  TNBCs are highly metastatic and contribute a large portion to breast cancer mortality due to a lack of targeted therapies.  The number of patients afflicted by acute and chronic kidney failure requiring dialysis grows steadily every year.  However, increased understanding of the regulation of shedding is also beneficial in various other diseases, e.g. in Alzheimer’s disease.


  Repair of tubular cells after kidney injury involves the activation of metalloproteases (which cleavage EGF ligands) and EGF receptor activation.

Repair of tubular cells after kidney injury involves the activation of metalloproteases (which cleavage EGF ligands) and EGF receptor activation.

Kidney Repair: The kidney has the capacity to regenerate completely after various acute insults. Regeneration by surviving tubular epithelial cells that proliferate and migrate to reepithelialize the tubulus is the predominant mechanism of repair. Several metalloprotease cleavage-driven events (ectodomain shedding) are implicated in the underlying processes of renal regeneration. For example, in kidney proximal tubular cells, the cleavage activated epidermal growth factor receptor (EGFR) ligands EGF, heparin binding-EGF (HB-EGF), amphiregulin (AR) and epiregulin (ER) can induce regenerative proliferation, cell motility and migration. In vivo, functional EGFR is required for recovery after renal injury and its activation enhances regeneration.

Maladaptive Repair (Fibrosis): Severe or chronic kidney injury, in contrast, leads to sustained activation of the EGFR, which is pro-fibrotic, and leads to scaring of the kidney (fibrosis) and loss of function (Kidney failure). It is not known which specific EGF ligands are involved in this process and why sustained EGFR activation is pro-fibrotic.

We study this process in detail in mouse models of kidney injury or fibrosis and in tubular cells in culture.


Breast Cancer

Breast cancer ADAM17.jpg

Our research aims to provide novel solutions for the reduction of breast cancer growth and metastasis and its resultant mortality. Using already identified candidate genes (Dang et al. PNAS 2013), we aim to test key signaling components that selectively regulate EGF ligand release by ADAM17 without significantly affecting its protease activity in in vitroand in vivo mouse-models of triple negative breast cancer (TNBC) tumor growth and metastasis. Members of the ADAM17-EGF ligand-HER1 axis are often found upregulated in breast cancer and correlate with metastasis and poor clinical outcome, particularly in TNBCs, which are highly metastatic and contribute a large portion to overall breast cancer mortality. Moreover, experimental data show that these molecules can actively contribute to the development of breast cancer. Notably, upregulation of HER1 and its pathway components were identified in two of six clinical TNBC subsets (BL and MSL), represented by the TNBC cell lines we study and a subset of TNBC patients indeed responds to HER1 therapy in clinical trials.These results are important in the context of recent experiments with TNBC cells. Staged HER1 kinase inhibition with erlotinib carried out 6-24 hours prior to treatment with DNA-damaging chemotherapy (doxorubicin) (both drugs are used in breast cancer therapy in the clinic) increased apoptosis of a subset of TNBC cell lines dramatically in vitro and reduced their tumor growth in vivoin mice. This response was correlated with the basal activity state of HER1. Chemotherapy alone or simultaneous co-administration was much less effective. In summary, these results identify the ADAM17-HER1 axis as a promising target in cancer therapy and particularly in TNBC. Yet, targeting of the ADAM17-EGF ligand-HER1 axis remains difficult. Currently available HER1-directed therapies are only partially effective and prone to de novo and acquired resistance. Metalloprotease inhibition has failed as a clinical approach in part due to broad-spectrum inhibition of substrate cleavage causing side effects. Genotoxic chemotherapy provides only limited benefit. To identify new solutions and therapeutic targets, we will use combinatory inhibition of selective EGF ligand cleavage regulators (kinases and phosphatases) with clinically used genotoxic chemotherapy and/or HER1 inhibition in vitro and in vivo in TNBC cells (see preliminary data). Our research will employ strategies that allow rapid testing of therapeutic targets in vitro and in vivo in breast cancer models relevant to human disease. Using a combination of already established therapies in the clinic with genetic or chemical approaches, we envision an accelerated transfer of this knowledge into clinical application in combination therapy. Several of our studied regulators are accessible by already existing chemical inhibitor drugs that are in clinical testing for human diseases. In addition, some of our regulators do not interfere with ADAM17 or ADAM10 metalloprotease activity and do not disturb the cleavage of other relevant metalloprotease substrates, a significant advantage over clinically failed metalloprotease inhibitors. The developed approaches could also help in the avoidance of resistance to HER1 inhibitors or chemotherapy.

Plug assay.png