Proteinase Biology and Antibodies: A Family-Level Overview

Proteinase antibodies present a different specificity problem than antibodies to most other protein classes. Every proteinase in the human genome — there are more than 550 — shares a small number of catalytic residues with its family members, folds into a similar three-dimensional core with them, and typically retains greater than 50% sequence identity to the two or three closest paralogues. A pan-family antibody is often the accidental result of a well-intentioned reagent; a paralogue-specific antibody requires deliberate epitope design against a divergent surface region. This overview describes the proteinase families TPB works on, the biology that motivates each family, and the antibody design decisions that follow.

The organisation is by mechanistic class first (following the MEROPS convention) and then by clinical or research context. Each family section links out to the corresponding catalog page and research-area page, and at the end there is a full Pillar Resources index of the 25 spoke pages that make up this cluster.

What are proteinases — the definition and the MEROPS classification

A proteinase (or protease, or peptidase) is an enzyme that catalyses the hydrolysis of peptide bonds. Proteinases participate in essentially every physiological process — protein turnover, digestion, blood clotting, complement activation, apoptosis, tissue remodelling, and hormone processing — and are dysregulated in cancer, neurodegeneration, cardiovascular disease, autoimmune conditions, and infectious disease. Their pharmacological accessibility has made them major drug targets: renin inhibitors, ACE inhibitors, direct thrombin and factor Xa inhibitors, HIV protease inhibitors, DPP-4 inhibitors, and the recent generation of BACE1 inhibitors all target proteinases directly.

The reference classification for proteinases is MEROPS (Rawlings et al.), which organises the enzymes by catalytic mechanism and evolutionary lineage. At the top level, MEROPS assigns each proteinase to one of six mechanistic classes based on the residue or cofactor responsible for peptide-bond cleavage: aspartic (A), cysteine (C), metallo (M), serine (S), threonine (T), and glutamic (G) proteinases, plus a small group of enzymes with mixed or unknown mechanism (N, U). For research applications, four classes dominate the literature and the antibody catalog: aspartic, cysteine, metallo, and serine proteinases.

The four major mechanistic classes

Why domain-specific detection matters for proteinases

Nearly every proteinase in the four major classes is synthesised as an inactive zymogen and activated by proteolytic processing. The propeptide is cleaved off, the mature catalytic domain becomes accessible, and — for many proteinases — subsequent cleavages release soluble ectodomains from membrane-anchored forms or generate specific processed products with distinct biological activities. This creates multiple functional pools of every proteinase, each detectable at a different molecular weight and each biologically distinct.

Consider MMP-9 as an example. The full-length gene product runs at 92 kDa (proMMP-9). Autolytic or trypsin-mediated activation removes the propeptide, yielding a 82-84 kDa active MMP-9. Further processing can generate a smaller catalytic-domain-only species. In a stimulated macrophage culture, all three forms may be present simultaneously, and the biology is defined by their relative abundance. An antibody that only detects proMMP-9 will miss the active pool entirely; an antibody that recognises a shared catalytic-domain epitope will detect all three but cannot distinguish them without gel size resolution.

Domain-specific antibody design solves this: a propeptide antibody detects only the zymogen; a hinge-region antibody detects both zymogen and mature forms but not the truncated catalytic species; a hemopexin-domain antibody (for MMPs) or C-terminal antibody (for ADAMs) distinguishes membrane-anchored from shed forms. For any proteinase experiment where the functional state matters — and this is most proteinase experiments — the antibody must be selected to match the functional pool of interest. This is the principal justification for TPB's Superpooled Antibody Method: multi-domain pooling that detects all forms of a proteinase simultaneously while giving domain resolution in a multi-channel readout.

Metalloproteinases (MMPs, ADAMs, ADAMTSs)

The metalloproteinase superfamily is the largest single family of proteinases in the human genome and the most extensively drug-targeted. It includes the matrix metalloproteinases (MMPs), the a-disintegrin-and-metalloproteinase family (ADAMs), the ADAM-with-thrombospondin-motifs family (ADAMTSs), meprins, neprilysin, ACE, and ACE2. All share a zinc-dependent catalytic mechanism and a metzincin fold with a conserved HEXXHXXGXXH zinc-binding motif.

Matrix metalloproteinases (MMPs)

The MMP family comprises 23 human members (MMP-1 through MMP-28, with several vacant numbers) subdivided by substrate preference and domain organisation. Collagenases (MMP-1, -8, -13) cleave native fibrillar collagens. Gelatinases (MMP-2, -9) cleave denatured collagens and basement membrane components. Stromelysins (MMP-3, -10, -11) have broad substrate range. Membrane-type MMPs (MT-MMPs: MMP-14, -15, -16, -17, -24, -25) are membrane-anchored and activate other MMPs on the cell surface. Matrilysins (MMP-7, -26) lack the hemopexin domain and have restricted substrate specificity.

MMPs are central to tissue remodelling in wound healing, angiogenesis, cancer invasion, cardiovascular remodelling, and arthritis. Elevated MMP-9 is a marker in stroke and traumatic brain injury; MMP-13 is the primary type II collagen collagenase in cartilage and a target in osteoarthritis; MMP-2 and MMP-9 gelatinase activity is a classical readout in cancer metastasis assays.

For MMP antibody selection, the propeptide, catalytic, and hemopexin domains are the three principal epitope choices. Propeptide antibodies detect zymogen only. Catalytic antibodies detect both zymogen and active forms but cross-react across paralogues if the epitope is in the conserved zinc-binding region. Hemopexin-domain antibodies typically give the best paralogue selectivity because this C-terminal domain is the most divergent between MMPs. See the MMP family antibody selection guide for the family-specific decision framework.

ADAMs and ADAMTSs

The ADAM family (a-disintegrin-and-metalloproteinase) comprises membrane-anchored sheddases responsible for releasing soluble ectodomains from membrane precursor proteins. ADAM10 and ADAM17 are the most extensively studied — ADAM10 sheds Notch, N-cadherin, and APP (α-secretase); ADAM17 (TACE) sheds TNFα, TGFα, and multiple ErbB ligands. ADAM domain organisation adds a disintegrin domain, a cysteine-rich domain, and a transmembrane and cytoplasmic tail on top of the metalloproteinase core, so antibody choice depends heavily on whether the researcher wants to detect the shed catalytic domain (soluble supernatant) or the intact membrane-anchored form (cell surface).

ADAMTS enzymes are secreted rather than membrane-anchored and have thrombospondin-type-1 repeats after the disintegrin domain. ADAMTS1, -4, -5 are aggrecanases (cartilage); ADAMTS7, -12 are cartilage oligomeric matrix protein cleavage enzymes; ADAMTS13 cleaves von Willebrand factor and its deficiency causes thrombotic thrombocytopenic purpura.

Cysteine proteinases (cathepsins, calpains, caspases, granzymes, legumain)

Cysteine proteinases hydrolyse peptide bonds through a covalent thioester intermediate formed at the active-site cysteine. In humans, the biology splits along fold: papain-like enzymes (cathepsins), calpain-like (calpains), caspase-like (caspases), and legumain (asparagine endopeptidase, AEP).

Cathepsins

Cysteine cathepsins (B, C, F, H, K, L, O, S, V, W, X/Z) are lysosomal proteases with a papain-like fold, active at acidic pH, and involved in general protein turnover as well as specialised functions: cathepsin K in osteoclast bone resorption, cathepsin S in MHC class II processing, cathepsin L in prohormone processing. Cathepsins D and E are aspartic (not cysteine) — a common source of confusion, since "cathepsin" is a functional label (lysosomal protease) not a mechanistic one.

For cathepsin antibodies, the propeptide is the standard target for zymogen-selective detection; the mature-form N-terminus and mid-catalytic epitopes are the standard targets for total-target detection. Cathepsin K vs L discrimination and cathepsin B vs D (across mechanism classes) are common research problems addressed by the corresponding X-vs-Y selection guides in Pillar Resources.

Calpains

Calpains are calcium-activated intracellular cysteine proteases. Calpain-1 (μ-calpain, requires micromolar calcium) and calpain-2 (m-calpain, requires millimolar calcium) are the two ubiquitous isoforms; additional tissue-restricted calpains (calpain-3 in muscle, calpain-9 in intestine) exist. Calpains cleave many substrates in signalling, cytoskeletal remodelling, and apoptosis. Calpain-1 vs calpain-2 antibody discrimination is important because the two isoforms have distinct calcium sensitivity and distinct pathological roles.

Caspases

Caspases are the executioner proteases of apoptosis (caspase-3, -6, -7), initiators (caspase-8, -9, -10, -2), and inflammatory caspases (caspase-1, -4, -5, -11 in mouse). All are synthesised as zymogens and activated by cleavage. Cleaved-caspase antibodies (neoepitope antibodies that recognise the exposed N-terminus of the cleavage product) are among the most powerful tools in apoptosis research.

Granzymes

Note: granzymes A, K, and M are serine proteases (chymotrypsin fold), while granzyme B has a distinct substrate preference (aspartate specificity like caspases). All are stored in cytotoxic-lymphocyte granules and delivered to target cells for cytotoxicity induction. Granzyme B activates caspase-3 and cleaves BID directly. See the granzyme A vs B and granzyme B vs K selection guides.

Legumain

Legumain (asparagine endopeptidase, AEP) is an unusual cysteine protease that specifically cleaves after asparagine residues. It is upregulated in tumours and in Alzheimer's disease brain, where it cleaves tau at N368 producing a fragment that seeds tau aggregation.

Serine proteinases (coagulation, complement, kallikreins, granzymes A/K/M, PCSKs)

Serine proteinases are the largest mechanistic class by human count and cover the widest functional range. The classical trypsin family (chymotrypsin fold) includes the digestive enzymes, the coagulation cascade (thrombin, factors VII, IX, X, XI, XII, kallikrein), fibrinolysis (tPA, uPA, plasmin), complement (MASP1, MASP2, factor B, C1r, C1s, C3 convertase), kallikreins (KLK1-15 including PSA/KLK3), granzymes A/K/M, and elastase. The subtilisin-fold PCSKs (proprotein convertases including furin, PC1/3, PC2, PCSK9) process prohormones and other precursor proteins.

Coagulation cascade

The coagulation cascade is a sequential activation of serine protease zymogens — factor XII → factor XI → factor IX → factor X → prothrombin (factor II) → thrombin → fibrinogen → fibrin — with parallel amplification loops and tissue-factor initiation. Every step is a serine protease activating the next. Antibody selection depends on whether the researcher needs to distinguish zymogen from active enzyme (activation state) and whether to distinguish factor X from factor Xa (cleaved vs uncleaved). See the coagulation serine proteases research page.

Complement proteases

The classical and lectin pathways of complement are protease cascades: C1r and C1s in the classical pathway; MASP1 and MASP2 in the lectin pathway. Alternative pathway proteases include factor B and factor D. MASP1 vs MASP2 discrimination is important because the two enzymes have overlapping but distinct substrate preferences and independent regulatory roles. See the MASP1 vs MASP2 selection guide.

Kallikreins

The kallikrein family (KLK1-15) is a chromosomally clustered set of 15 tissue serine proteases with tissue-restricted expression and roles in skin desquamation (KLK5, KLK7, KLK14), semen liquefaction (KLK3/PSA), and neural function. KLK3 is prostate-specific antigen, the most extensively used serum tumour marker. Selection of kallikrein antibodies depends heavily on paralogue discrimination because tissue-restricted KLKs are frequently co-expressed and the sequence identity between adjacent KLKs is high.

Granzymes A, K, M (serine)

Granzymes A, K, and M are serine proteases stored in cytotoxic-lymphocyte granules. Unlike granzyme B (which activates caspases), granzyme A and K cleave substrates in the target cell to induce caspase-independent cell death. Antibody selection distinguishes granzyme B from granzyme K because their roles diverge — granzyme B is the classical caspase-activating enzyme, granzyme K is the caspase-independent enzyme with expanding roles in inflammation.

Proprotein convertases (PCSKs)

PCSKs are subtilisin-fold proteases that process precursor proteins by cleaving at basic residues. Furin (PCSK3) is the archetypal PCSK and processes proinsulin, proBDNF, coagulation factors, and many other precursors. PCSK9 does not act as a canonical protease but binds the LDL receptor and directs it for degradation — the target of major lipid-lowering therapies.

Aspartic proteinases (cathepsins D and E, BACE1/2, renin, pepsin)

Aspartic proteinases use two aspartate residues to activate a water molecule for hydrolysis. In humans, the family is small — cathepsin D, cathepsin E, BACE1, BACE2, renin, and pepsinogens — but functionally important. BACE1 (β-secretase 1) is the initiating enzyme of amyloid precursor protein cleavage in Alzheimer's disease and has been an intensive drug target for two decades. Cathepsin D is a lysosomal aspartic protease with roles in general protein turnover and cathepsin-mediated apoptosis pathways.

BACE1 vs BACE2 antibody discrimination is a common problem — the two enzymes share 64% sequence identity in the mature catalytic domain, but BACE2 is functionally distinct (β-secretase for a subset of substrates in different tissue contexts). See the Alzheimer's BACE antibodies research page for the Alzheimer's-relevant selection framework.

Proteinase inhibitors (TIMPs, cystatins, serpins)

Proteinase inhibitors are the endogenous regulators of proteinase activity, and antibodies against them are as important as antibodies against the enzymes themselves — for both mechanistic research and biomarker studies.

See the proteinase-inhibitor pairs research page for the biology of inhibitor-enzyme pairing across the families.

Zymogens and pro-enzymes — the activation state problem

Almost every proteinase is synthesised as an inactive precursor (zymogen or pro-enzyme). The propeptide serves several functions: it blocks the active site (auto-inhibition), it can serve as an intramolecular chaperone for folding, and it can direct subcellular targeting. Activation removes the propeptide, exposing the mature catalytic surface. For research antibody selection, the propeptide vs mature-form distinction is central — an antibody raised against the propeptide detects only the zymogen; an antibody raised against a mature-form epitope may detect both.

For MMPs, activation typically involves stepwise cleavage by another protease (frequently plasmin, another MMP, or trypsin) at a defined cysteine-switch region in the propeptide. For caspases, activation involves auto-cleavage or cleavage by an initiator caspase. For serine proteases, activation is often a single cleavage after a specific arginine or lysine residue that exposes the mature N-terminus. Each mechanism generates a specific neoepitope at the newly exposed N-terminus, and neoepitope-specific antibodies (raised against the first several amino acids of the mature form, including the newly exposed residue) are among the most specific reagents for detecting activated forms. See the zymogens and pro-enzymes guide for the mechanistic detail.

Disease relevance and research contexts

Proteinase research is disease research. Every major disease category has proteinase involvement, and antibody-based detection is the primary method by which proteinase biology is measured.

How proteinase antibody research differs from other target research

Two features distinguish proteinase antibody selection from selection for most other target classes.

First, functional-state detection is nearly always required. Because every proteinase is regulated by activation and degradation, the concentration of "MMP-9" in a sample is rarely the question — the question is what fraction is zymogen, what fraction is active, and what fraction is truncated or complexed with TIMP. This requires domain-specific antibody selection or (better) multi-epitope pooled reagents that detect all forms simultaneously.

Second, paralogue discrimination is a design problem, not a selection problem. Because the catalytic core is conserved across a proteinase family, antibodies raised against the catalytic domain almost always show some pan-family reactivity. Paralogue-specific reagents require epitopes in divergent surface regions (loops, propeptides, C-terminal domains) that are typically not the highest-antigenicity sites on the protein. This is why proteinase antibody design is more difficult than generic antibody design and why TPB's 32 years of proteinase-specific epitope-mapping expertise is directly load-bearing for the catalog. See the proteinase classification overview for the family-level breakdown.

Choosing an antibody for a proteinase experiment

Bringing the biology and the antibody selection together, the recommended workflow for choosing a proteinase antibody is:

1. Identify the target functional pool. Zymogen? Mature? Cleaved? Complexed with an inhibitor? All forms? The answer determines the domain to target.
2. Identify the closest paralogues to distinguish. Which family members do you need to exclude? BLAST the target against those paralogues and identify divergent regions where the epitope must fall.
3. Match the domain and the divergent region. If the target functional pool is the mature enzyme and the paralogues to exclude are the closest family members, the epitope needs to be in a divergent surface region of the mature-form domain.
4. Check the manufacturer's validation. Does the antibody have knockout data, recombinant paralogue panel data, or both? For proteinase families, recombinant paralogue panel testing is essentially mandatory — this is a specificity claim that must be evidenced, not asserted.
5. Consider superpooled or multi-domain pools if all forms are required. If the experiment needs to detect zymogen, mature, and cleaved forms simultaneously with domain resolution, superpooled antibodies (multi-domain polyclonal pools) are typically the correct choice. See the Superpooled Antibody Method.

See the how to choose a proteinase antibody guide for the detailed decision walkthrough.

Frequently Asked Questions

Explore related topics in this cluster

• Metalloproteinases catalog— MMPs, ADAMs, ADAMTSs, meprins
• Cysteine proteinases catalog— cathepsins, calpains, caspases, legumain
• Serine proteinases catalog— coagulation, complement, kallikreins, granzymes
• Aspartic proteinases catalog— BACE, cathepsin D, renin, pepsin
• Other enzymes catalog— proteinase-adjacent enzymes
• Metalloproteinase inhibitors— TIMPs 1-4
• Cysteine proteinase inhibitors— cystatins
• Serine proteinase inhibitors— serpins
• Alzheimer's BACE antibodies— β-secretase research area
• Cancer MMP research panel— MMPs in oncology
• Coagulation serine proteases— thrombin, factor Xa, plasmin
• Inflammation ADAMs— ADAM17 and shedding
• Neurodegeneration cathepsins— cathepsins in CNS disease
• Proteinase-inhibitor pairs— endogenous inhibitor pairings
• MMP family antibody selection— matrix metalloproteinase selection guide
• Proteinase classification overview— MEROPS and mechanism classes
• Zymogens and pro-enzymes— activation and functional state detection