Only 340/565 human proteases (Nature Reviews Genetics) have known substrates, and from this incomplete knowledge, >200 proteases lack biological roles (Nature Reviews Molecular Cell Biology). This inspired Dr. Overall to invent a suite of approaches¹⁴, techniques^15-20, and software^4,16 to identify protease substrates on a system-wide scale.  Recognizing the importance of substrate-binding exosite domains on proteases, he was the first to use these as substrate ‘baits’ in a yeast two-hybrid screen¹⁴ (Science)—at a time when protein disulphide cross-linkages were predicted to exclude the yeast two-hybrid approach for extracellular proteins.  Dr. Overall showed that this was not a limitation in this yeast system¹⁴ with a paradigm shift that MMPs are tissue-protective by orchestrating neutrophil and macrophage leukocyte responses through activation/inactivation of virtually all chemoattractant cytokines known as chemokines^10,13,14, and in subsequent research, by multiple cytokines^6,8, their binding proteins, and serpin inhibitors^11,12.

Defining protease substrate cleavage-site specificity is central to protease characterization, linkage to substrates and drug development. In this way, he pioneered the first peptide library technique to simultaneously identify both the amino (P) and carboxyl (P’) amino acid residues flanking the cleavage site—Proteomic Identification of Cleavage Specificity(PICS)¹⁹ (Nature Biotechnology, Nature Protocols). With >14,000 cleavage sites identified by PICS in the protease database MEROPS, its head curator described PICS as “the new gold standard for profiling proteases.”  In one application of PICS, he reported >4,300 cleavage sites in the MMP family, leading to structural insight into their active sites. Reported in Nature Communications⁷, he identified >1000 cleavage sites by an unsuspected metallopeptidase insertion in bacterial flagellin, the monomer of bacterial flagella that propels bacteria through biofilms and tissues. These metalloproteinase-bearing flagellin molecules assemble proteolytically active flagella
(~20 µm) in >350 diverse bacterial species, e.g., the pathogen Clostridium haemolyticum.

In conventional proteomics sample preparation for mass spectrometry, 100,000s of trypsin-generated peptides of a proteome dominated by abundant proteins dilute the terminal amino (N) and carboxyl (C) terminal protein ends and protease-generated ‘neo’-termini, rendering these terminal peptides rarely detectable.  This is especially problematic for low-abundant transcription factors and signalling proteins like cytokines and antiviral interferons, impairing insight into disease-relevant proteolytic events.  Dr. Overall’s technique, Terminal Amino Isotopic Labelling of Substrates (TAILS)¹⁸, circumvents these issues in a simple but powerful high-throughput method.  TAILS purifies protein N-terminal peptides and cleaved neo-N-terminal peptides using an innovative aldehyde-polymer to simultaneously identify cleavage sites andsubstrates in native proteomes (Nature Biotechnology; Nature Protocols). Monitoring the end fragments of the cut proteins, also known as terminomics, is a powerful new way to monitor disease activity.

Protein C-termini are difficult to label chemically en route to identification that is also hampered by their absence of C-terminal lysine or arginine residues following trypsin digestion.  Dr. Overall discovered a new “trypsin mirror” protease in Archaea, LysargiNase^TM, to address this¹⁵ (Nature Methods).  Cutting before lysine or arginine, LysargiNase-digested proteins release C-terminal peptides with an N-terminal lysine or arginine, facilitating detection by shotgun proteomics or by C-terminal peptide enrichment for broad coverage using C-TAILS technique¹⁷ (Nature Methods).

To specifically analyze natural and neo-termini, Dr. Overall developed publicly available software:  WebPICS, CLIPPER, and the Termini-orientated protein Function INferred Database (TopFIND v4.1)¹⁶ (Nature Methods) with >290K termini/>33K cut-sites, receiving >4K hits p.a.  PathFINDer and TopFINDer map substrates to protease pathways, which revealed a highly connected protease network in humans¹² (PLoS Biology)—so also highlighting the problems in interpreting knockout or overexpression studies, as well as in drug targeting of proteases.

Fundamental to understanding zoonotic virus pathobiology, by a One Health approach in the pandemic, Dr. Overall applied his techniques to discover >300 host cell substrates for the SARS-CoV-2 3CL^pro, main protease¹ (Cell Reports) and its cleavage site specificity (J Virology). Thereby, he discovered that the cytosolic lectin, galectin-8, is a novel intracellular sensor of SARS-CoV-2 spike protein, but galectin-8 cleavage by SARS-CoV-2 3CL^pro defeats autophagic destruction of galectin-8-tagged virus. In expanding the degradome and interactome of 3CL^{pro 20} (Nature Communications), he revealed mechanisms by which this viral protease bypasses cellular control of signal transduction by interacting with and disrupting protein assemblies in the adherens junction, cytoskeleton and centrosome of infected human lung epithelial cells. Notably, 3CL^pro drives the formation of intercellular immune-privileged conduits known as tunnel nanotubes (TNT).  Immunolocalization of virions with 3CL^pro substrates within TNTs suggests that 3CL^pro activity accelerates TNT formation for “stealthy” viral transmission, which impedes the effectiveness of immune recognition and vaccines intended to reduce viral transmission. Recently, he reported the first example of a viral protease, SARS-CoV-2 3CL^pro, being secreted with extracellular activity, including inactivation of type III interferons (Cell Reports) ²¹.

Employing degradomics, Dr. Overall has mechanistically dissected the crosstalk between proteolytic pathways in multiple systems including complement in skin inflammation⁸ (Science Signaling), lysosomal proteases in pancreatic cancer⁹ (Cell Reports), MMPs in HIV-Associated Dementia¹⁰ (Nature Neuroscience), anti-inflammatory activities of macrophage MMP12 in arthritis¹¹ (Cell Reports) and autoimmunity, e.g., lupus⁶ (Nature Communications), and MMP12’s unexpected potent antiviral roles⁵ (Nature Medicine).  In this latter work, he made a remarkable discovery that secreted MMP12 is a ‘moonlighting’ protease (Nature Reviews Drug Discovery) that re-enters cells and traffics to the nucleus as a transcription factor regulating ~200 genes.  By increasing transcription of IkBa—an inhibitor of the pro-inflammatory NFkB—he found MMP12 was indispensable for IFNa secretion.  He further identified multiple substrates of MMP12 whose gene transcription was repressed by nuclear MMP12—thereby demonstrating concerted dual-negative regulation of both protein substrates and their genes—highlighting a mechanism of rapid cellular inhibition and removal of key cellular proteins for antiviral defence.

In related work, Dr. Overall found that macrophage MMP12 first stimulates secretion and then, over time, inactivates anti-viral interferon-a⁵ (Nature Medicine).  Similarly, MMP12 inactivates interferon-g, also by removing the receptor binding site⁶ (Nature Communications), providing feedback that drives the transition from pro-inflammatory IFN-g-activated macrophages (formerly termed “M1”) to tissue-reparative immunosuppressant (“M2”) macrophages. In discovering new and major roles of MMPs in regulating signalling proteins, his studies reveal deeper complexity in regulating the extracellular matrix and the cell signalling environment than mere degradation.

In a groundbreaking example of clinical application, Professor Overall explored the roles of the intracellular protease MALT1, an essential transducer in lymphocyte antigen receptor signalling and immune activation² (Nature Communications).  He found that independent of proteolytic cleavage, non-proteolytic protein-protein interactions by MALT1 initiate NFkB activation, whereas, at late stages of NFkB signalling, MALT1 cleaves HOIL1 in the Linear Ubiquitin Assembly Complex (LUBAC) to downregulate essential linear ubiquitination of pathway mediators to halt NFkBsignalling².  By developing the algorithm GO-2-Substrates, he recently predicted and validated new substrates to double the MALT1 substrate repertoire⁴; thereby, he upended the concept that MALT1 is solely an enhancer of antigen-driven signalling—with far-reaching consequences for our understanding of immunobiology that has major implications in MALT1 drug targeting for lymphomas.  The multidisciplinary team he assembled also reported in Nature Chemical Biology³ the development of a series of potent nanomolar allosteric inhibitors of MALT1 that bind at Trp580. Remarkably, this is the same site as an immunodeficient patient’s MALT1 Trp580Ser mutation, whose markedly diminished levels of the mutant MALT1 led to 50% reduced NFkB signalling and immunodeficiency.  In an elegant use of chemical biology, Dr. Overall demonstrated that treatment with this inhibitor stabilized the mutant MALT1, restoring MALT protein levels to normal in the patient’s lymphocytes.  Moreover, inhibitor treatment also restored NFkB and JNK signalling in the patient’s B and T lymphocytes and, during treatment pauses, rescued substrate cleavage. Thus, a new low molecular weight pharmacological molecular corrector rescues an enzyme deficiency by substituting for the mutated residue, inspiring therapies to restore enzyme activity to treat similar molecularly defined disorders.