Catalytically inactive enzymes (also known as pseudoproteases, protease homologues or paralogues, non-peptidase homologues, non-enzymes and pseudoenzymes) have traditionally been hypothesized to act as regulators of their active homologues. However, those that have been characterized demonstrate that inactive enzymes have an extensive and expanding role in biological processes, including regulation, inhibition and immune modulation. With the emergence of each new genome, more inactive enzymes are being identified, and their abundance and potential as therapeutic targets has been realized. In the light of the growing interest in this emerging field the present review focuses on the classification, structure, function and mechanism of inactive enzymes. Examples of how inactivity is defined, how this is reflected in the structure, functions of inactive enzymes in biological processes and their mode of action are discussed.

INTRODUCTION

The characterization of inactive enzymes has increasingly become a subject of research as more inactive enzymes are revealed as participants in important biological processes. A study by Pils and Schultz [1] of 47 enzymatic domains across seven metazoan genomes identified inactive homologues in all species investigated. The authors concluded that the evolution of inactive enzymes was a commonly occurring event [1]. This concurred with an earlier study by Todd et al. [2]. The majority of the inactivated domains were identified as inactive signalling domains with the next largest group being inactive extracellular domains. Within the extracellular domain grouping, the trypsin-like serine proteases had the highest number of inactive enzymes identified, especially in Anopheles gambiae and Drosophila melanogaster. Pils and Schultz [1] argued that the large numbers of both inactive and active trypsin-like serine proteases, in both these species and their conservation in many other related species, indicated a gene expansion, and that the evolution of the inactive proteases and their new functions suggested that they were advantageous to insects. This conclusion can be extended to include other organisms such as humans. Ordonez et al. [3] evaluated the total number of enzyme genes in the human genome and reported that, of the 569 genes identified, 92 lacked critical catalytic residues and were therefore predicted to be inactive. What is clear from these and similar investigations are that inactive enzymes are essential, abundant and are found and shared across many organisms.

Like their active homologues, they tend not to exist in isolation but rather as members of multigene families or as components of large complex units. They have been identified across all catalytic types, exist in most enzyme families and are functionally diverse. Using examples from a variety of families, the present review focuses on the classification, structure, function and mechanism of inactive enzymes. We first discuss how inactive enzymes are initially identified as being members of a family on the basis of sequence similarity and then defined as catalytically inactive due to changes to the critical catalytic residues. Given the central role that specific amino acids play in catalysis, a change to these residues is typically highly disruptive. To demonstrate this, examples are given of how such changes contribute to the classification of inactive enzymes. Subsequent validation of enzymatic inactivity by experimental and structural evidence ultimately confirms catalytic inactivity. Typically, the addition of structural data leads to the identification of additional changes such as blocked active sites or substrate-binding pockets. These features have the dual role of confirming the classification of inactivity and of having a major influence on function. The examples discussed to demonstrate structure and function of inactive enzymes highlight how inactive enzymes have utilized these features to engage with their substrates and evolve alternative modes of action.

WHEN ARE ENZYMES CLASSIFIED AS INACTIVE?

Enzymes are classified by three important features: (i) the principal residues involved in catalysis or catalytic type, (ii) the reactions they catalyse, and (iii) their molecular structure and homology with archetypal enzymes inferring an evolutionary relationship [4]. It is the obvious disruption or change in one or more of these features that classifies an enzyme as inactive. Features common to inactive enzymes are mutations of the catalytic residues, alterations to the structure or fold, and steric changes affecting the substrate binding and active sites. Some examples are discussed below to illustrate this.

Principal residues involved in catalysis or catalytic class

Enzymes are classically defined by a set of key functional amino acid residues that are employed in catalysis, substrate binding, structural stabilization, protein accepting or donating and nucleophilic addition. These residues contribute to the ability of an enzyme to perform catalysis, and hence any disruption to these residues tends to result in inactivity [5]. Inactive enzymes have been identified in many classes due to a variety of mutations to these archetypal residues. The most common disruptions are single or multiple substitutions of the active site catalytic residues. Examples of this are seen in SPH3 (serine protease homologue 3), haptoglobin and the scabies mite inactive cysteine protease paralogues. SPH3 protein expressed by the moth Manduca sexta is rendered inactive by the alteration of the catalytic serine residue to glycine [6]. In haptoglobin, a member of the complement control protein–serine protease family, the β chain resembles a serine protease domain with the exception of the replacement of two of the three catalytic residues His57 and Ser195 to lysine and alanine respectively [7,8]. In both of these examples, the residue substitutions have been experimentally shown to prevent the activation of the charge relay system necessary for serine protease activity. The scabies mite inactive cysteine protease paralogues are a family of inactive cysteine proteases that are homologous with the scabies mite group 1 cysteine proteases [9]. The scabies mites are unique in having both a family of proteolytically inactive cysteine and serine proteases. A phylogenetic tree of the scabies mite cysteine proteases displaying the catalytic residues indicates that the inactive cysteine proteases have substitutions of both catalytic diad residues and the glutamine residue involved in oxyanion hole formation (Figure 1). Since changes to the principal residues are readily identified when analysing the primary sequence, this feature is the main basis for an inactive classification.

Phylogenetic tree of scabies mite group 1 cysteine proteases and homologues

Figure 1
Phylogenetic tree of scabies mite group 1 cysteine proteases and homologues

House dust mite group 1 allergens (Eur m 1, Der p 1, Der f 1;), sheep scab mite (Pso o 1), scabies mite group 1 proteases (Yv4003H01, Yv5020C01, Yv6030H07, Yv5032C08, Yv9053H09) and homologous scabies mite inactive cysteine proteases (Yv4025A02, Yv6008G08, Yv4028C12, Yv6018B11, Yv5009F04). The active-site cysteine and histidine diad (CH), asparagine (N) and glutamine (Q) residues are shown for the active proteases in black and the substitutions to these catalytic residues in the inactive proteases are also shown in red. Modified from Holt et al. [9].

Figure 1
Phylogenetic tree of scabies mite group 1 cysteine proteases and homologues

House dust mite group 1 allergens (Eur m 1, Der p 1, Der f 1;), sheep scab mite (Pso o 1), scabies mite group 1 proteases (Yv4003H01, Yv5020C01, Yv6030H07, Yv5032C08, Yv9053H09) and homologous scabies mite inactive cysteine proteases (Yv4025A02, Yv6008G08, Yv4028C12, Yv6018B11, Yv5009F04). The active-site cysteine and histidine diad (CH), asparagine (N) and glutamine (Q) residues are shown for the active proteases in black and the substitutions to these catalytic residues in the inactive proteases are also shown in red. Modified from Holt et al. [9].

Reactions catalysed

An important means of grouping enzymes is based on the type of reactions that they catalyse. Inactive enzymes initially found to group with a particular family based on sequence homology must demonstrate the ability to catalyse the reaction attributed to be considered active. As described in the previous examples, a single change to critical catalytic residues can be sufficient to disrupt catalysis. However, other inactive enzymes have substitutions extending beyond the catalytic residues to ancillary residues necessary for positioning substrates, thereby resulting in a loss of functional catalysis.

Molecular structure and homology with archetypal enzymes

Outside the catalytic and ancillary residues are those involved in tertiary conformation. Mutations to these residues can result in structural rearrangements in addition to changes to the catalytic triad such as those seen in Bla g 2 and SMIPP-Ss (scabies mite inactive serine protease paralogues). Bla g 2 is an aspartic protease allergen from cockroaches and distinguishes itself from active aspartic proteases by a number of features, including residue changes within the loop referred to as the ‘flap’ which is involved in the catalytic mechanism. As a result of these changes, new hydrogen-bond networks are established that interfere with catalysis and enforce conformation changes in the flap region, resulting in a closed conformation format [10]. The SMIPP-Ss also possess a structural alteration that contributes to rendering them inactive. The SMIPP-Ss lack the cysteine residues that participate in the formation of a third disulfide bond common to proteolytically active trypsin-like proteases. This induces a conformational change that has an impact on the availability of the substrate-binding pocket [11].

Enzyme databases

Traditional automated and manual methods for classifying enzymes as inactive are, of course, sequence homology with active enzymes and the omission or substitution of critical residues involved in catalysis. This initial classification was and still is then verified by manual methods of experimentation. In recent years, new methods for the identification of enzymes have emerged in an effort to improve classification and inference of function. The advent of specialized databases dedicated to enzymes has enabled researchers to BLAST-search with their novel and unannotated sequences against a database of well-characterized enzymes, holding information on sequence, active-site residues, substrate specificity, inhibitor profiles and structure. The BLAST enables a comparative analysis of the unknown enzyme against several enzyme features simultaneously. One well-established enzyme database is the peptidase database MEROPS (http://merops.sanger.ac.uk) [12].

MEROPS is a database containing extensive information on peptidases, including inactive peptidases referred to as NPHs (non-peptidase homologues). The database currently lists 54 clans, with 248 families, and within each clan a number of NPHs have been identified. All NPHs that have been assigned a MEROPS identifier (ID) have an ID number that begins with the family number, followed by a dot, the number 9 and then a sequential identifier, for example family.9XX. In addition, there are NPHs that have not been assigned an ID and have been grouped under a generic name for example: family A26 non-peptidase homologues. Using this annotation, we compiled a list of identified NPHs as they are currently listed in MEROPS (as of August 2014). Of the 248 peptidase families presently listed in the MEROPS database, 174 contain NPHs, covering all catalytic classes (Table 1). The 174 families represent 70% of the total. A full table listing all the currently identified or allocated NPHs can be found in Supplementary Table S1. What is clearly evident from the MEROPS database is that inactive peptidases are abundant and diverse and therefore indeed biologically relevant.

Table 1
Summary of the number of catalytic families containing inactive proteases listed in MEROPS (as of August 2014)
Catalytic family Total number of families Number of families with inactive proteases 
Aspartic 16 12 (75%) 
Cysteine 81 49 (60.5%) 
Glutamic 2 (100%) 
Asparagine 10 7 (70%) 
Serine 53 38 (72%) 
Metalloprotease 70 60 (86%) 
Threonine 5 (83%) 
Mixed 1 (100%) 
Unknown 1 (11%) 
Catalytic family Total number of families Number of families with inactive proteases 
Aspartic 16 12 (75%) 
Cysteine 81 49 (60.5%) 
Glutamic 2 (100%) 
Asparagine 10 7 (70%) 
Serine 53 38 (72%) 
Metalloprotease 70 60 (86%) 
Threonine 5 (83%) 
Mixed 1 (100%) 
Unknown 1 (11%) 

STRUCTURES OF INACTVE ENZYMES

The examples discussed below are chosen to demonstrate the types of structural differences seen between inactive enzymes and their active counterparts and how this impacts on catalytic inactivity. Referring to the MEROPS database, Supplementary Table S1 also lists NPHs for which a 3D structure exists.

Scabies mite inactive serine protease paralogues (SMIPP-Ss)

SMIPP-Ss are a multigene family of house dust mite allergen group 3 homologues that have been identified as members of the S1-like family (chymotrypsin-like) [13]. Recently solved X-ray crystal structures of two members, SMIPP-S-I1 and SMIPP-S-D1, revealed that, although both adopt the chymotrypsin-like serine protease fold, there are major structural rearrangements around the S1 subsite, most notably the insertion of a bulky tyrosine residue into the site [11]. This rearrangement results from a missing disulfide bond that usually occurs at loop-220 (chymotrypsin numbering) in the serine protease family. The bond is created by Cys191 and Cys220, both replaced in the SMIPP-Ss. The lack of the bond untethers the loop allowing it to shift position, affecting the position of tyrosine at position 200, conserved in the SMIPP-S family. The residue is consequently positioned to the S1 subsite, rendering the site inaccessible to substrate (shown for SMIPP-S-I1 in Figure 2). An equivalent scenario was observed in the second SMIPP-S structure (D1) and sequence analysis suggested the lack of the third disulfide bond and the presence of a large amino acid at position 200 for the majority of SMIPP-S sequences identified to date. As the interaction of the P1 residue of the substrate with the S1 subsite determines protease specificity, the observed obstruction suggested that SMIPP-Ss do not function as competitive inhibitors for substrate. The accumulated evidence of a mutated catalytic triad and an inaccessible S1 subsite led to the conclusion that these inactive proteases had not maintained canonical function and biochemical studies indicated that their altered active sites were probably not the site of interaction [11].

Structural comparison of inactive serine protease SMIPP-S-I1 with trypsin

Figure 2
Structural comparison of inactive serine protease SMIPP-S-I1 with trypsin

SMIPP-S-I1 (PDB code 3H7O) in green overlaid with trypsin (PDB code 5PTP) in magenta and trypsin inhibitor (PDB code 2PTC) in red. The canonical catalytic triad is indicated by HDA and the disulfide bonds by SS1, SS2 and SS3. Tyr200 blocking the SMIPP-S-I1 subsite is indicated by Y200. The missing disulfide bond SS3 in SMIPP-S-I1 is indicated by the broken line. Modified from Fischer et al. [11].

Figure 2
Structural comparison of inactive serine protease SMIPP-S-I1 with trypsin

SMIPP-S-I1 (PDB code 3H7O) in green overlaid with trypsin (PDB code 5PTP) in magenta and trypsin inhibitor (PDB code 2PTC) in red. The canonical catalytic triad is indicated by HDA and the disulfide bonds by SS1, SS2 and SS3. Tyr200 blocking the SMIPP-S-I1 subsite is indicated by Y200. The missing disulfide bond SS3 in SMIPP-S-I1 is indicated by the broken line. Modified from Fischer et al. [11].

Bla g 2

Bla g 2 is a cockroach antigen identified as an inactive aspartic protease [14]. Here we see two major structural rearrangements resulting from the numerous residue changes to this molecule: the substitution of catalytic triad residues and the insertion/substitution of residues in the flap region involved in the catalytic mechanism. Bla g 2 comprises two domains that flank a large cleft. Within each of these domains lie two of the residues making up the catalytic triad, whereas the third is located in the cleft. In active aspartic proteases, each domain contains the DTG motif with the aspartate (D) residue located in the cleft, in Bla g 2 this is DTS and DTT. Novel hydrogen-bonding networks are established between the catalytic aspartate residues and the substituted serine and threonine. In addition, the aspartate residue also establishes a new hydrogen bond with an additional phenylalanine residue inserted into the flap region that engages the catalytic triad. The cumulative result of the novel hydrogen-bond network results in two structural changes. First, the catalytic aspartate residues are now bonded to the phenylalanine residue inserted in the flap region, which effectively closes the flap and hinders access to the catalytic aspartate residues. Secondly, the aromatic ring of the inserted phenylalanine residue is positioned to occupy the S1 substrate-binding pocket, thus occluding it from substrate binding. This mimics a similar scenario seen in the active proteases when they are in a self-inhibited transition state [10].

PfClpR

ClpR is the inactive paralogue of ClpP, the proteolytic component of ATP-dependent caseinolytic proteases. The recently solved X-ray crystal structure of ClpR in Plasmodium falciparum (PfClpR) has revealed that significant differences exist between the catalytic and non-catalytic paralogues [15]. PfClpR does resemble other ClpPs in that it forms a sevenfold symmetrical single ring with a central pore. In both ClpPs and PfClpR, each subunit has an α/β fold comprising six repeats of the α/β unit which forms the head domain. The head domain is decorated with an additional α/β unit which forms a handle region. Between these two regions exists a cleft, which is occupied by the catalytic SHD triad in an active ClpP; however, in PfClpR, this cleft contains the GND triad. The hydrophobic groove which leads to the catalytic triad and forms a substrate-binding surface in ClpP is also retained in the PfClpR structure.

Despite these shared structural features, there are three major modifications that differentiate PfClpR from ClpPs (i) an open and flatter structure, (ii) the insertion of a unique motif in the head domain, and (iii) the creation of an additional deep cleft. In the PfClpR heptamer, each subunit is twisted outwards by approximately 15° resulting in a compressed and wider ring which translates into a more open and flatter structure than in ClpP. This has the added effect of expanding the surface area of the PfClpR heptamer. The head region of PfClpR has an insertion of a ten-residue unique motif referred to as the R-motif. This motif extends from the head region, forms a β-turn and faces the internal chamber of the heptamer complex. The presence of the motif creates an additional deep cleft close to the hydrophobic groove leading to the substrate-binding surface. The positional relationship to the internal chamber does not affect the α/β fold of the subunit or interfere with access to the hydrophobic groove. The high conservation of the R-motif in the Plasmodium family suggests a role of yet unknown function for PfClpR.

FUNCTIONS OF INACTIVE ENZYMES

Traditionally, the functions of inactive enzymes were presumed to be either competitive inhibitors or regulators of their active counterparts. Although some have evolved to perform these functions, many have functions quite distinct from the active enzymes they resemble. A genomic study by Todd et al. [2] sought to determine whether function could be inferred by comparing the sequence and structure of enzyme/inactive enzyme homologues. They concluded that many of the functions of inactive enzymes that they studied were unrelated to the proteolytically active cousin [2]. The evident evolution of inactive enzymes across most families suggests that this is an advantageous expansion of enzymes. This supports the need to view inactive enzymes as a distinct group within a superfamily. Determining the functions of many inactive enzymes is ongoing and some examples are described below.

Inactive enzymes involved in immune function

A serine protease homologue identified in the crab species Scyalla paramamosain is able to bind to the bacterial crab pathogen Vibrio parahaemolyticus [16]. The protease is homologous with the catalytically inactive PPAF (prophenoloxidase-activating factor) found in three other crab species. PPAFs are initiators of the prophenoloxidase-activating system or melanization cascade, an important immune mechanism in invertebrates [17]. As essential cofactors for prophenoloxidase-activating enzyme, they constitute one of four recognized regulatory mechanisms [18]. PPAFs comprise an N-terminal clip domain and a C-terminal serine protease domain. Clip domains have been shown to regulate protease activity, engage in protein–protein interactions and perform bactericidal functions and the serine protease domain cleaves prophenoloxidase [17,19]. Inactive PPAFs differ from their active counterparts in the clip domain and in the serine protease domain, with non-synonymous substitutions of the catalytic residues in one or both domains. The serine protease in S. paramamosain has a substitution of glycine for the catalytic serine residue in the serine protease domain, rendering it catalytically inactive [16]. The inactive protease was found to be tissue-specific, being localized to the eye stalk, subcuticular epidermis, stomach, gills, haemocyte, thorax ganglion, brain and muscle. After challenge with bacterial infection, tissue expression of PPAF protein was up-regulated in the haemocytes, subcuticular epidermis and gills, all considered frontline defence tissues. The ability of this protease to recognize pathogen, its homology with other known prophenoloxidase-activation molecules and its localization suggests that it functions as an immune-recognition molecule and plays a role in crab antibacterial defences [16].

SMIPP-S-D1 and SMIPP-S-I1 have been localized to the mite gut and seen to be excreted in the mite faeces into the burrow [20]. Both of these regions represent potential sites for host immune interaction, targeting the mite for host defences. The mites have a counterdefence to one of these immune mechanisms, the complement cascade of the innate immune system. SMIPP-S-D1 and SMIPP-S-I1 are anti-complement molecules released by the mite during infection [21]. These two molecules have been studied extensively, and a further five members of the multigene family have also been shown to inhibit complement [22]. This is in complete contrast with their active protease paralogue, which does not interfere with the complement cascade and has been shown to digest skin proteins [23]. Of the 32 inactive proteases identified from this family to date, seven have been shown to have anti-complement activity, suggesting that the family is potentially specifically targeting this system. The polymorphic nature of the family would present an adaptive advantage in minimizing host opportunities to develop a specific antibody response.

ROP5 (rhoptry protein 5) is representative of a family of proteolytically inactive kinases found in Toxoplasma gondii associated with virulence and lethality in mice infections. They are expressed at the ROP5 locus as highly divergent and polymorphic isoforms. Injected into the host cell cytoplasm during infection, ROP5 localizes to the parasitophorous vacuolar membrane surface. Its location suggests that ROP5 interacts with host proteins important in protection or immunity, since mice infected with ROP5 locus-knockout parasites survived infection [24]. Similar to the SMIPP-Ss, polymorphism presumably endows the parasite with an advantage that most certainly contributes to virulence.

Inactive enzymes as regulators

Caspases are cysteine proteases that have an essential role in apoptosis and inflammation, and as such must be tightly regulated. In mammals, nematodes and arthropods, a number of caspase homologues have been identified and shown to have a role in the regulation of active caspases. An example of one such caspase homologue is CASPS18, found in the mosquito Aedes aegypti. CASPS18 is a caspase-like decoy protease that lacks two critical catalytic residues, the cysteine residue of the catalytic diad and a conserved arginine residue [25], and is a positive regulator of its active paralogue CASPS19. This was determined in vitro by Bryant et al. [26] who demonstrated that co-expression of CASPS18 and CASPS19 results in an increase in CASPS19 activity and a reduction in apoptosis of cells expressing CASPS19. Another well-studied caspase is the mammalian caspase 8 homologue, c-FLIPL {cellular FLICE [FADD (Fas-associated death domain)-like interleukin 1β-converting enzyme]-inhibitory protein (long form)} whose mutated protease domain lacks catalytic activity. c-FLIPL is a regulator of the extrinsic apoptotic pathway through its interaction with the pathway initiators caspase 8 and 10. The best described of these is caspase 8 [27].

Caspase 8 is expressed as a monomer that requires dimerization for the formation of the active-site dyad and substrate-binding pocket. Dimerization enables a structural rearrangement of the four loops that stabilize the catalytic site into an active conformation. It has been demonstrated that dimerization is a critical requirement for activation, whereas cleavage of the interdomain linker in the protease domain is not. c-FLIPL is able to form a heterodimeric complex with caspase 8 and activate it [28,29].

Heterodimerization of the caspase prodomains is facilitated in a stable protein platform called the DISC (death-inducing signalling complex). In the DISC, a monomer of caspase 8 preferentially cleaves the interdomain linker in the protease-like domain of c-FLIPL over itself. This cleavage promotes dimerization with caspase 8, which in turn activates caspase 8. Processed c-FLIPL also increases the recruitment of caspase 8 to the DISC [30]. A heterodimer containing processed caspase 8 is capable of cleaving and activating downstream apoptosis pathway targets. When cleavage occurs, the heterodimer becomes stabilized and caspase activity is increased [31].

In a heterodimer containing unprocessed caspase 8, the interdomain linker in the catalytic domain of caspase 8 occupies its own active site. However, activity is still evident in these dimers. Although caspase 8 is active, the substrate specificity is sufficiently narrowed so that caspase 8 is unable to cleave the downstream pro-apoptosis targets Bid and caspase 3. Processed c-FLIPL also mobilizes additional pro-survival proteins to the complex [32,33]. c-FLIPL has also been shown to be anti-apoptotic when at concentrations that exceed that of caspase 8. In this context, c-FLIPL competes with caspase 8 for recruitment to the DISC. At high concentrations, c-FLIPL occupies available binding sites preventing caspase 8 from binding [34].

In these interactions with caspase 8, c-FLIPL has been described as a dual regulator, with the ability to either inhibit or activate apoptosis. Caspase 8 has been reported to have additional functions outside apoptosis, and evidence suggests that it is the heterodimer with c-FLIPL that facilitates these additional functions [31].

Multifunctional inactive enzymes

In humans, one of the best described inactive serine proteases is human HBP (heparin-binding protein) also known as azurocidin or CAP37 (cationic antimicrobial protein of 37 kDa). It resembles neutrophil elastase, but substitutions of the catalytic histidine and serine residues render it inactive [35]. HBP is a multifunctional protein involved in host defence and inflammation [36]. In addition to its heparin-binding abilities [37], HBP has been shown to display antibiotic activity against Gram-negative bacteria [38] and the ability to chemoattract and activate monocytes and T-cells [39].

Another well-studied inactive enzyme is the acute-phase reactant protein haptoglobin. It is an inactive serine protease and comprises a complement control protein domain and a serine protease domain. However, the serine protease domain lacks the residues required for a functional catalytic triad and has several distinguishing surface loop regions that differ from other serine proteases [40]. Haptoglobin has a role in restoring systemic homoeostasis through anti-inflammatory activities. Its main function is to bind free haemoglobin, thereby removing it from the circulation and preventing oxidative tissue damage [41]. The binding of free haemoglobin to haptoglobin enables the ligation of the scavenger receptor CD163, a signal-inducing protein found on the surface of macrophages and monocytes. The ligation signals the release of anti-inflammatory cytokines [42]. The binding of haptoglobin to CD163 is mediated through one of the unique loop regions on the surface of the serine protease-like domain. The binding site for haemoglobin is in the region which dictates substrate specificity in active serine proteases [43,44]. The combination of haemoglobin removal and a role in triggering the release of anti-inflammatory cytokines makes haptoglobin an important anti-oxidant and anti-inflammatory protein with a pivotal role in maintaining host haemostasis. Haptoglobin has also been associated with the regulation of epidermal Langerhans cell maturation [45].

Multifunctional inactive enzymes have also been found in protozoan species such Trypanosoma brucei. Of the five metacaspases (MCA1–MCA5) expressed by T. brucei, two, MCA1 and MCA4, contain substitutions in the active site. MCA1 has both the histidine and cysteine residue of the catalytic diad replaced, whereas MCA4 has a single alteration of the histidine residue to serine. MCA4 has been experimentally shown to lack activity and to be incapable of autocatalysis, but instead is processed by MCA3. Like many of its caspase relatives, MCA4 was found to be multifunctional with roles in blood-stage parasite cytokinesis and virulence during mammalian infection. Processing of MCA4 by MCA3 also suggests that MCA4 itself is part of the catalytic cascade, with MCA4 being a substrate of MCA3. This scenario suggests that the inactive MCA4 is regulated by the active MCA3. The biological relevance and exact mechanism of this regulation are yet to be determined [46].

THE MODE OF ACTION OF INACTIVE ENZYMES

The principal site of interaction in an enzyme is the active site. Typically comprising a groove or cleft built from loops, it houses the residues that facilitate the global binding of a substrate, interacts with substrate residues (subsites) and is responsible for the catalytic reaction. Historically, this site and its interactions were considered unwaveringly specific with regard to the catalytic reaction and the substrate specificity. However, enzymes have been found to display both catalytic promiscuity, i.e. performing reactions other than those for which they evolved, and substrate promiscuity (ambiguity), i.e. binding structurally related substrates. This is distinct from having broad specificity or being multi-specific. The promiscuity features are commonly found in enzyme families and are increasingly considered to be the rule rather than an exception [47]. Substrate promiscuity is also found in non-catalytic molecules [48]. Evolutionary biochemists consider promiscuity to be important in both catalytic and non-catalytic molecules in the evolution of new mechanisms and functions that enhance fitness of a molecule. An extensive discussion on the inherent characteristics of enzymes that facilitate promiscuity can be found in [4951].

Two points in the literature regarding enzyme promiscuity worth mentioning for the present discussion are: (i) the conformationally dynamic active site, and (ii) the relevance of individual subsites in substrate specificity. First, it is widely agreed that the active site is a conformationally flexible structure and that this trait is a major contributing factor that enables substrate promiscuity. This is clearly evident from the fact that the binding of multiple substrates by a single enzyme is not unusual [5254]. Secondly, efforts to quantify the influence individual subsites have on specificity demonstrates that not all subsites have equal value. That is, some are more critical for specificity than others in the substrate-binding pocket [55]. What this infers is that the active site and subsites constitute a dynamic space with the potential for degrees of specificity.

Given that these mechanisms exist in enzymes and non-catalytic molecules, is it plausible that they also exist in inactive enzymes? And what impact would they have on the mode of action? A loss in catalytic activity simply means a loss of the ability to catalyse a chemical reaction. It does not necessarily infer that the inactive enzyme has also lost its ability to utilize the remaining active-site apparatus to facilitate substrate interactions, mode of action and function. Unless there are considerable changes that occlude the active and/or substrate sites, then the use of either is not inconceivable. Should they be unavailable then the evolution of an exosite would be logical. So what evidence exists? Many inactive enzymes whose structure and mode of action have been characterized to date use the pseudo-active site or an alternative exosite. Although the use of the canonical substrate-binding sites, if they are available, appears to be possible, there is very little evidence in the literature to indicate that this is a mode of action utilized by inactive proteases. As further inactive enzymes are characterized and large-scale comparative analysis of the mechanisms employed by inactive enzymes becomes possible, perhaps additional modes of action will become evident. Below we discuss some examples of the binding by the pseudo-active site and the exosite.

Binding via the pseudo-active site

Inactive domains can be utilized as a means of regulating activation, inhibition or binding affinity. In the metalloprotease-like protein Sonic Hedgehog, it is the pseudo-active site that facilitates binding to regulatory proteins of the Hedgehog pathway. Sonic Hedgehog is an important signalling molecule in the Hedgehog pathway involved in embryogenesis and tissue regeneration. The protein is composed of two domains: the N-terminal signalling domain and the C-terminal intein-like domain. The N-terminal domain of Sonic Hedgehog is responsible for short- and long-range signalling and contains a pseudo-active site. Although the pseudo-active site has been shown to be incapable of catalytic activity, it still acts as a ligand for membrane-bound receptors [56]. Hedgehog regulatory proteins such as Hedgehog-interacting protein and Patched 1 bind to the pseudo-active site groove. In this manner, Sonic Hedgehog is able to affect the regulation of Hedgehog pathway signalling [57,58].

Alternative binding sites or exosites

Another possibility is the use of an alternative binding site. An investigation by Pils and Schultz [59] on the evolution of the PTP (protein tyrosine phosphatase) family found that there was a loss of evolutionary pressure around the catalytic centre of a subclass of inactive domains, resulting in a high rate of evolutionary change. They questioned whether this site could still be responsible for the observed regulatory function or whether it had evolved a novel site. They found several sites of high conservation undergoing low rates of evolution on the opposite side of the active site of the domain structure. Pils and Schultz [59] suggest that this could indicate a newly evolving functional centre for these domains.

PZ (Protein Z) is a plasma protein that is a co-factor for the serpin PZI (Protein Z inhibitor), an inhibitor of the coagulation factor fXa (Factor Xa) [60,61]. PZ is structurally related to the coagulation cascade serine protease factors fVII (Factor VII), fIX (Factor IX) and fX (Factor X), but is catalytically inactive due to the replacement of two of the three catalytic residues histidine and serine by lysine and aspartate [62]. The N-terminus of PZ is composed of a Gla (γ-carboxyglutamic acid) domain followed by two epidermal growth factor-like domains and a C-terminal inactive catalytic domain [63]. The C-terminal region containing the inactive catalytic domain has a trypsin serine protease fold [64]. The region adjacent to the inactive site pocket is the site for PZ binding to PZI, an interaction facilitated through ionic and polar interactions [65]. Mutagenesis studies of this region demonstrate the importance of this site for the interaction between PZ and PZI. The Gla domain of PZ is used to anchor PZI when complexed with PZ to membrane surfaces to orientate the complex for efficient interaction with fXa. This factor has been shown to accelerate the inhibitory activity of PZI [65].

The structural studies of SMIPP-Ss led researchers to conclude that these inactive enzymes mediate their unique biological activity via an exosite [11]. Patches of highly conserved residues on the face opposite the active site were identified and postulated to be good candidate exosites for protein–protein interactions. Mutagenesis studies targeting the conserved surface-exposed residues at the exosite enabled the research team to narrow this region down to a smaller patch of residues as the potential protein interaction site [21] (Figure 3).

Defining the protein interaction site of the inactive serine protease SMIPP-S-I1

Figure 3
Defining the protein interaction site of the inactive serine protease SMIPP-S-I1

Structure of SMIPP-S-I1 (PDB code 3H7O) with regions of conservation (red) and conserved surface-exposed residues indicated. Conserved residues in the exosite targeted in mutagenesis studies are boxed and the associated regions that were focused on are tinted yellow. Modified from Fischer et al. [11].

Figure 3
Defining the protein interaction site of the inactive serine protease SMIPP-S-I1

Structure of SMIPP-S-I1 (PDB code 3H7O) with regions of conservation (red) and conserved surface-exposed residues indicated. Conserved residues in the exosite targeted in mutagenesis studies are boxed and the associated regions that were focused on are tinted yellow. Modified from Fischer et al. [11].

CONLUDING REMARKS

Researchers agree that the presence of inactive enzymes is common and that an inactive enzyme tends to have evolved from the active precursor rather than vice versa [1,2]. Inactive enzymes appear to have been evolving in parallel with their active homologues within superfamilies, and are now distinguishing themselves as important players in biological systems. The discovery of so many inactive enzymes across a host of families suggests that their emergence is an evolutionary advantage rather than a misadventure. The growing importance of inactive enzymes is highlighted by the extensive range of processes with which they have been shown to be involved and is supported further by the growing interest in them as potential targets in disease therapeutics [66]. As the identification of more inactive enzymes in biological processes emerges, the need to have a thorough understanding of their structure, function and mode of action will grow. The present review has sought to highlight the diversity in structure, function and mode of action that has evolved within the inactive enzymes. Importantly, these examples demonstrate that a loss of an ancestral mechanism such as catalysis does not result in a loss of function, but rather the evolutionary incentive for the design of new mechanisms and new functions, thereby expanding the repertoire of enzyme families.

FUNDING

S.L.R. is supported by a National Health and Medical Research Council Early Career Fellowship [grant number 1054968] and K.F. is supported by an Australian Research Council Future Fellowship [grant number FT130101875].

Abbreviations

     
  • c-FLIPL

    cellular FLICE [FADD (Fas-associated death domain)-like interleukin 1β-converting enzyme]-inhibitory protein (long form)

  •  
  • DISC

    death-inducing signalling complex

  •  
  • fXa

    Factor Xa

  •  
  • Gla

    γ-carboxyglutamic acid

  •  
  • HBP

    heparin-binding protein

  •  
  • MCA

    metacaspase

  •  
  • NPH

    non-peptidase homologue

  •  
  • Pf

    Plasmodium falciparum

  •  
  • PPAF

    prophenoloxidase-activating factor

  •  
  • PZ

    Protein Z

  •  
  • PZI

    Protein Z inhibitor

  •  
  • ROP5

    rhoptry protein 5

  •  
  • SMIPP-S

    scabies mite inactive serine protease paralogue

  •  
  • SPH3

    serine protease homologue 3

References

References
1
Pils
B.
Schultz
J.
Inactive enzyme-homologues find new function in regulatory processes
J. Mol. Biol.
2004
, vol. 
340
 (pg. 
399
-
404
)
[PubMed]
2
Todd
A.E.
Orengo
C.A.
Thornton
J.M.
Sequence and structural differences between enzyme and nonenzyme homologs
Structure
2002
, vol. 
10
 (pg. 
1435
-
1451
)
[PubMed]
3
Ordonez
G.R.
Puente
X.S.
Quesada
V.
Lopez-Otin
C.
Proteolytic systems: constructing degradomes
Methods Mol. Biol.
2009
, vol. 
539
 (pg. 
33
-
47
)
[PubMed]
4
Barrett
A.J.
Classification of peptidases
Methods Enzymol.
1994
, vol. 
244
 (pg. 
1
-
15
)
[PubMed]
5
Holliday
G.L.
Mitchell
J.B.
Thornton
J.M.
Understanding the functional roles of amino acid residues in enzyme catalysis
J. Mol. Biol.
2009
, vol. 
390
 (pg. 
560
-
577
)
[PubMed]
6
Felfoldi
G.
Eleftherianos
I.
Ffrench-Constant
R.H.
Venekei
I.
A serine proteinase homologue, SPH-3, plays a central role in insect immunity
J. Immunol.
2011
, vol. 
186
 (pg. 
4828
-
4834
)
[PubMed]
7
Kurosky
A.
Barnett
D.R.
Lee
T.H.
Touchstone
B.
Hay
R.E.
Arnott
M.S.
Bowman
B.H.
Fitch
W.M.
Covalent structure of human haptoglobin: a serine protease homolog
Proc. Natl. Acad. Sci. U.S.A.
1980
, vol. 
77
 (pg. 
3388
-
3392
)
[PubMed]
8
Greer
J.
Comparative model-building of the mammalian serine proteases
J. Mol. Biol.
1981
, vol. 
153
 (pg. 
1027
-
1042
)
[PubMed]
9
Holt
D.C.
Fischer
K.
Pizzutto
S.J.
Currie
B.J.
Walton
S.F.
Kemp
D.J.
A multigene family of inactivated cysteine proteases in Sarcoptes scabiei
J. Invest. Dermatol.
2004
, vol. 
123
 (pg. 
240
-
241
)
[PubMed]
10
Gustchina
A.
Li
M.
Wunschmann
S.
Chapman
M.D.
Pomes
A.
Wlodawer
A.
Crystal structure of cockroach allergen Bla g 2, an unusual zinc binding aspartic protease with a novel mode of self-inhibition
J. Mol. Biol.
2005
, vol. 
348
 (pg. 
433
-
444
)
[PubMed]
11
Fischer
K.
Langendorf
C.G.
Irving
J.A.
Reynolds
S.
Willis
C.
Beckham
S.
Law
R.H.
Yang
S.
Bashtannyk-Puhalovich
T.A.
McGowan
S.
, et al. 
Structural mechanisms of inactivation in scabies mite serine protease paralogues
J. Mol. Biol.
2009
, vol. 
390
 (pg. 
635
-
645
)
[PubMed]
12
Rawlings
N.D.
Waller
M.
Barrett
A.J.
Bateman
A.
MEROPS: the database of proteolytic enzymes, their substrates and inhibitors
Nucleic Acids Res.
2014
, vol. 
42
 (pg. 
D503
-
D509
)
[PubMed]
13
Holt
D.C.
Fischer
K.
Allen
G.E.
Wilson
D.
Wilson
P.
Slade
R.
Currie
B.J.
Walton
S.F.
Kemp
D.J.
Mechanisms for a novel immune evasion strategy in the scabies mite Sarcoptes scabiei: a multigene family of inactivated serine proteases
J. Invest. Dermatol.
2003
, vol. 
121
 (pg. 
1419
-
1424
)
[PubMed]
14
Wunschmann
S.
Gustchina
A.
Chapman
M.D.
Pomes
A.
Cockroach allergen Bla g 2: an unusual aspartic proteinase
J. Allergy Clin. Immunol.
2005
, vol. 
116
 (pg. 
140
-
145
)
[PubMed]
15
El Bakkouri
M.
Rathore
S.
Calmettes
C.
Wernimont
A.K.
Liu
K.
Sinha
D.
Asad
M.
Jung
P.
Hui
R.
Mohmmed
A.
Houry
W.A.
Structural insights into the inactive subunit of the apicoplast-localized caseinolytic protease complex of Plasmodium falciparum
J. Biol. Chem.
2013
, vol. 
288
 (pg. 
1022
-
1031
)
[PubMed]
16
Liu
H.P.
Chen
R.Y.
Zhang
M.
Wang
K.J.
Isolation, gene cloning and expression profile of a pathogen recognition protein: a serine proteinase homolog (Sp-SPH) involved in the antibacterial response in the crab Scylla paramamosain
Dev. Comp. Immunol.
2010
, vol. 
34
 (pg. 
741
-
748
)
[PubMed]
17
Cerenius
L.
Lee
B.L.
Söderhäll
K.
The proPO-system: pros and cons for its role in invertebrate immunity
Trends Immunol.
2008
, vol. 
29
 (pg. 
263
-
271
)
[PubMed]
18
Buda
E.S.
Shafer
T.H.
Expression of a serine proteinase homolog prophenoloxidase-activating factor from the blue crab, Callinectes sapidus
Comp. Biochem. Physiol. B Biochem. Mol. Biol.
2005
, vol. 
140
 (pg. 
521
-
531
)
[PubMed]
19
Jiang
H.
Kanost
M.R.
The clip-domain family of serine proteinases in arthropods
Insect Biochem. Mol. Biol.
2000
, vol. 
30
 (pg. 
95
-
105
)
[PubMed]
20
Willis
C.
Fischer
K.
Walton
S.F.
Currie
B.J.
Kemp
D.J.
Scabies mite inactivated serine protease paralogues are present both internally in the mite gut and externally in feces
Am. J. Trop. Med. Hyg.
2006
, vol. 
75
 (pg. 
683
-
687
)
[PubMed]
21
Reynolds
S.L.
Pike
R.N.
Mika
A.
Blom
A.M.
Hofmann
A.
Wijeyewickrema
L.C.
Kemp
D.
Fischer
K.
Scabies mite inactive serine proteases are potent inhibitors of the human complement lectin pathway
PLoS Negl. Trop. Dis.
2014
, vol. 
8
 pg. 
e2872
 
[PubMed]
22
Bergstrom
F.C.
Reynolds
S.
Johnstone
M.
Pike
R.N.
Buckle
A.M.
Kemp
D.J.
Fischer
K.
Blom
A.M.
Scabies mite inactivated serine protease paralogs inhibit the human complement system
J. Immunol.
2009
, vol. 
182
 (pg. 
7809
-
7817
)
[PubMed]
23
Beckham
S.A.
Boyd
S.E.
Reynolds
S.
Willis
C.
Johnstone
M.
Mika
A.
Simerska
P.
Wijeyewickrema
L.C.
Smith
A.I.
Kemp
D.J.
, et al. 
Characterization of a serine protease homologous to house dust mite group 3 allergens from the scabies mite Sarcoptes scabiei
J. Biol. Chem.
2009
, vol. 
284
 (pg. 
34413
-
34422
)
[PubMed]
24
Reese
M.L.
Zeiner
G.M.
Saeij
J.P.
Boothroyd
J.C.
Boyle
J.P.
Polymorphic family of injected pseudokinases is paramount in Toxoplasma virulence
Proc. Natl. Acad. Sci. U.S.A.
2011
, vol. 
108
 (pg. 
9625
-
9630
)
[PubMed]
25
Bryant
B.
Blair
C.D.
Olson
K.E.
Clem
R.J.
Annotation and expression profiling of apoptosis-related genes in the yellow fever mosquito, Aedes aegypti
Insect Biochem. Mol. Biol.
2008
, vol. 
38
 (pg. 
331
-
345
)
[PubMed]
26
Bryant
B.
Ungerer
M.C.
Liu
Q.
Waterhouse
R.M.
Clem
R.J.
A caspase-like decoy molecule enhances the activity of a paralogous caspase in the yellow fever mosquito, Aedes aegypti
Insect Biochem. Mol. Biol.
2010
, vol. 
40
 (pg. 
516
-
523
)
[PubMed]
27
Parrish
A.B.
Freel
C.D.
Kornbluth
S.
Cellular mechanisms controlling caspase activation and function
Cold Spring Harb. Perspect. Biol.
2013
, vol. 
5
 pg. 
a008672
 
[PubMed]
28
Boatright
K.M.
Renatus
M.
Scott
F.L.
Sperandio
S.
Shin
H.
Pedersen
I.M.
Ricci
J.E.
Edris
W.A.
Sutherlin
D.P.
Green
D.R.
Salvesen
G.S.
A unified model for apical caspase activation
Mol. Cell
2003
, vol. 
11
 (pg. 
529
-
541
)
[PubMed]
29
Boatright
K.M.
Deis
C.
Denault
J.B.
Sutherlin
D.P.
Salvesen
G.S.
Activation of caspases-8 and -10 by FLIPL
Biochem J.
2004
, vol. 
382
 (pg. 
651
-
657
)
[PubMed]
30
Yu
J.W.
Jeffrey
P.D.
Shi
Y.
Mechanism of procaspase-8 activation by c-FLIPL
Proc. Natl. Acad. Sci. U.S.A.
2009
, vol. 
106
 (pg. 
8169
-
8174
)
[PubMed]
31
van Raam
B.J.
Salvesen
G.S.
Proliferative versus apoptotic functions of caspase-8 hetero or homo: the caspase-8 dimer controls cell fate
Biochim. Biophys. Acta
2012
, vol. 
1824
 (pg. 
113
-
122
)
[PubMed]
32
Pop
C.
Oberst
A.
Drag
M.
Van Raam
B.J.
Riedl
S.J.
Green
D.R.
Salvesen
G.S.
FLIPL induces caspase 8 activity in the absence of interdomain caspase 8 cleavage and alters substrate specificity
Biochem. J.
2011
, vol. 
433
 (pg. 
447
-
457
)
[PubMed]
33
Oberst
A.
Dillon
C.P.
Weinlich
R.
McCormick
L.L.
Fitzgerald
P.
Pop
C.
Hakem
R.
Salvesen
G.S.
Green
D.R.
Catalytic activity of the caspase-8–FLIPL complex inhibits RIPK3-dependent necrosis
Nature
2011
, vol. 
471
 (pg. 
363
-
367
)
[PubMed]
34
Scaffidi
C.
Schmitz
I.
Krammer
P.H.
Peter
M.E.
The role of c-FLIP in modulation of CD95-induced apoptosis
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
1541
-
1548
)
[PubMed]
35
Pohl
J.
Pereira
H.A.
Martin
N.M.
Spitznagel
J.K.
Amino acid sequence of CAP37, a human neutrophil granule-derived antibacterial and monocyte-specific chemotactic glycoprotein structurally similar to neutrophil elastase
FEBS Lett.
1990
, vol. 
272
 (pg. 
200
-
204
)
[PubMed]
36
Ostergaard
E.
Flodgaard
H.
A neutrophil-derived proteolytic inactive elastase homologue (hHBP) mediates reversible contraction of fibroblasts and endothelial cell monolayers and stimulates monocyte survival and thrombospondin secretion
J. Leukoc. Biol.
1992
, vol. 
51
 (pg. 
316
-
323
)
[PubMed]
37
Flodgaard
H.
Ostergaard
E.
Bayne
S.
Svendsen
A.
Thomsen
J.
Engels
M.
Wollmer
A.
Covalent structure of two novel neutrophile leucocyte-derived proteins of porcine and human origin: neutrophile elastase homologues with strong monocyte and fibroblast chemotactic activities
Eur. J. Biochem.
1991
, vol. 
197
 (pg. 
535
-
547
)
[PubMed]
38
Rasmussen
P.B.
Bjorn
S.
Hastrup
S.
Nielsen
P.F.
Norris
K.
Thim
L.
Wiberg
F.C.
Flodgaard
H.
Characterization of recombinant human HBP/CAP37/azurocidin, a pleiotropic mediator of inflammation-enhancing LPS-induced cytokine release from monocytes
FEBS Lett.
1996
, vol. 
390
 (pg. 
109
-
112
)
[PubMed]
39
Iversen
L.F.
Kastrup
J.S.
Bjorn
S.E.
Rasmussen
P.B.
Wiberg
F.C.
Flodgaard
H.J.
Larsen
I.K.
Structure of HBP, a multifunctional protein with a serine proteinase fold
Nat. Struct. Biol.
1997
, vol. 
4
 (pg. 
265
-
268
)
[PubMed]
40
Andersen
C.B.
Torvund-Jensen
M.
Nielsen
M.J.
de Oliveira
C.L.
Hersleth
H.P.
Andersen
N.H.
Pedersen
J.S.
Andersen
G.R.
Moestrup
S.K.
Structure of the haptoglobin–haemoglobin complex
Nature
2012
, vol. 
489
 (pg. 
456
-
459
)
[PubMed]
41
Oliviero
S.
Cortese
R.
The human haptoglobin gene promoter: interleukin-6-responsive elements interact with a DNA-binding protein induced by interleukin-6
EMBO J.
1989
, vol. 
8
 (pg. 
1145
-
1151
)
[PubMed]
42
Kristiansen
M.
Graversen
J.H.
Jacobsen
C.
Sonne
O.
Hoffman
H.J.
Law
S.K.
Moestrup
S.K.
Identification of the haemoglobin scavenger receptor
Nature
2001
, vol. 
409
 (pg. 
198
-
201
)
[PubMed]
43
Nielsen
M.J.
Petersen
S.V.
Jacobsen
C.
Thirup
S.
Enghild
J.J.
Graversen
J.H.
Moestrup
S.K.
A unique loop extension in the serine protease domain of haptoglobin is essential for CD163 recognition of the haptoglobin–hemoglobin complex
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
1072
-
1079
)
[PubMed]
44
Perona
J.J.
Craik
C.S.
Evolutionary divergence of substrate specificity within the chymotrypsin-like serine protease fold
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
29987
-
29990
)
[PubMed]
45
Xie
Y.
Li
Y.
Zhang
Q.
Stiller
M.J.
Wang
C.L.
Streilein
J.W.
Haptoglobin is a natural regulator of Langerhans cell function in the skin
J. Dermatol. Sci.
2000
, vol. 
24
 (pg. 
25
-
37
)
[PubMed]
46
Proto
W.R.
Castanys-Munoz
E.
Black
A.
Tetley
L.
Moss
C.X.
Juliano
L.
Coombs
G.H.
Mottram
J.C.
Trypanosoma brucei metacaspase 4 is a pseudopeptidase and a virulence factor
J. Biol. Chem.
2011
, vol. 
286
 (pg. 
39914
-
39925
)
[PubMed]
47
Khersonsky
O.
Tawfik
D.S.
Enzyme promiscuity: a mechanistic and evolutionary perspective
Annu. Rev. Biochem.
2010
, vol. 
79
 (pg. 
471
-
505
)
[PubMed]
48
Copley
S.D.
An evolutionary biochemist's perspective on promiscuity
Trends Biochem. Sci.
2015
, vol. 
40
 (pg. 
72
-
78
)
[PubMed]
49
Nobeli
I.
Favia
A.D.
Thornton
J.M.
Protein promiscuity and its implications for biotechnology
Nat. Biotechnol.
2009
, vol. 
27
 (pg. 
157
-
167
)
[PubMed]
50
Brown
S.D.
Babbitt
P.C.
New insights about enzyme evolution from large scale studies of sequence and structure relationships
J. Biol. Chem.
2014
, vol. 
289
 (pg. 
30221
-
30228
)
[PubMed]
51
Pandya
C.
Farelli
J.D.
Dunaway-Mariano
D.
Allen
K.N.
Enzyme promiscuity: engine of evolutionary innovation
J. Biol. Chem.
2014
, vol. 
289
 (pg. 
30229
-
30236
)
[PubMed]
52
Tokuriki
N.
Tawfik
D.S.
Protein dynamism and evolvability
Science
2009
, vol. 
324
 (pg. 
203
-
207
)
[PubMed]
53
Hiblot
J.
Gotthard
G.
Elias
M.
Chabriere
E.
Differential active site loop conformations mediate promiscuous activities in the lactonase SsoPox
PLoS ONE
2013
, vol. 
8
 pg. 
e75272
 
[PubMed]
54
Babtie
A.
Tokuriki
N.
Hollfelder
F.
What makes an enzyme promiscuous?
Curr. Opin. Chem. Biol.
2010
, vol. 
14
 (pg. 
200
-
207
)
[PubMed]
55
Fuchs
J.E.
von Grafenstein
S.
Huber
R.G.
Margreiter
M.A.
Spitzer
G.M.
Wallnoefer
H.G.
Liedl
K.R.
Cleavage entropy as quantitative measure of protease specificity
PLoS Comput. Biol.
2013
, vol. 
9
 pg. 
e1003007
 
[PubMed]
56
Bishop
B.
Aricescu
A.R.
Harlos
K.
O'Callaghan
C.A.
Jones
E.Y.
Siebold
C.
Structural insights into hedgehog ligand sequestration by the human hedgehog-interacting protein HHIP
Nat. Struct. Mol. Biol.
2009
, vol. 
16
 (pg. 
698
-
703
)
[PubMed]
57
Bosanac
I.
Maun
H.R.
Scales
S.J.
Wen
X.
Lingel
A.
Bazan
J.F.
de Sauvage
F.J.
Hymowitz
S.G.
Lazarus
R.A.
The structure of SHH in complex with HHIP reveals a recognition role for the Shh pseudo active site in signaling
Nat. Struct. Mol. Biol.
2009
, vol. 
16
 (pg. 
691
-
697
)
[PubMed]
58
Maun
H.R.
Wen
X.
Lingel
A.
de Sauvage
F.J.
Lazarus
R.A.
Scales
S.J.
Hymowitz
S.G.
The hedgehog pathway antagonist 5E1 binds hedgehog at the pseudo-active site
J. Biol. Chem.
2010
, vol. 
285
 (pg. 
26570
-
26580
)
[PubMed]
59
Pils
B.
Schultz
J.
Evolution of the multifunctional protein tyrosine phosphatase family
Mol. Biol. Evol.
2004
, vol. 
21
 (pg. 
625
-
631
)
[PubMed]
60
Han
X.
Fiehler
R.
Broze
G.J.
Jr
Characterization of the protein Z-dependent protease inhibitor
Blood
2000
, vol. 
96
 (pg. 
3049
-
3055
)
[PubMed]
61
Rezaie
A.R.
Bae
J.S.
Manithody
C.
Qureshi
S.H.
Yang
L.
Protein Z-dependent protease inhibitor binds to the C-terminal domain of protein Z
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
19922
-
19926
)
[PubMed]
62
Broze
G.J.
Jr
Miletich
J.P.
Human Protein Z
J. Clin. Invest.
1984
, vol. 
73
 (pg. 
933
-
938
)
[PubMed]
63
Ichinose
A.
Takeya
H.
Espling
E.
Iwanaga
S.
Kisiel
W.
Davie
E.W.
Amino acid sequence of human protein Z, a vitamin K-dependent plasma glycoprotein
Biochem. Biophys. Res. Commun.
1990
, vol. 
172
 (pg. 
1139
-
1144
)
[PubMed]
64
Chandrasekaran
V.
Lee
C.J.
Duke
R.E.
Perera
L.
Pedersen
L.G.
Computational study of the putative active form of protein Z (PZa): sequence design and structural modeling
Protein Sci.
2008
, vol. 
17
 (pg. 
1354
-
1361
)
[PubMed]
65
Huang
X.
Dementiev
A.
Olson
S.T.
Gettins
P.G.
Basis for the specificity and activation of the serpin protein Z-dependent proteinase inhibitor (ZPI) as an inhibitor of membrane-associated factor Xa
J. Biol. Chem.
, vol. 
285
 (pg. 
20399
-
20409
)
[PubMed]
66
Dunn
B.M.
Proteinases as Drug Targets
2012
Cambridge
Royal Society of Chemistry

Supplementary data