Abstract. Proteases are often overexpressed in tumor cellsand/or the stromal compartment and can thus be exploited intumor therapy to activate cytotoxic prodrugs as, for example,in cytolytic fusion proteins, and for tumor imaging. Specifically,we discuss cathepsin B-activated prodrug conjugates, antibody-directed prodrug therapy, protease-activated peptide−thapsigargin conjugates, protease-activated cytotoxic receptorligands and other cytotoxic proteins, protease-mediatedactivation of anthrax toxin, granzyme B as a therapeuticprinciple in cytolytic fusion proteins, and tumor-imaging basedon deregulated proteases.

The human degradome consists of roughly 600 proteases,which are grouped into metallo-, serine, cysteine, threonine andaspartic proteases (1). They regulate physiological processessuch as proliferation, migration, cell survival, apoptosis andprotein turnover. In tumors, expression of proteases isfrequently dysregulated, thus, coopting these physiologicalprocesses, which are often based on complex interactionsbetween tumors and their microenvironment (2). Matrixmetalloproteinases (MMPs) are a family of proteins consistingof 26 members of secreted and transmembrane proteins (3)which degrade the extracellular matrix and regulate the activityof other proteases, growth factors and receptors, thusmodulating signaling pathways often associated with pro- orantitumoral outcome (4). The family consists of collagenases,stromelysins, gelatinases and membrane-type MMPs (5). In thecontext of tumor biology, MMP2 (gelatinase A), MMP9(gelatinase B), membrane-type1 protease MMP14, matriptase

(membrane-type serine protease 1), as well as cathepsin B, acysteine protease (6), were thoroughly investigated (5). Inaddition, the serine protease urokinase (uPA) and its specificreceptor (uPAR) control matrix degradation and many otherbiological processes by conversion of plasminogen intoplasmin and by receptor signaling based on the interaction ofuPAR with many extracellular protein ligands (7). Thederegulation of this system in several types of cancer has asignificant impact on prognosis and can serve as a possibletarget for therapeutic intervention (8-10). Other proteolyticsystems deregulated in cancer include kallikrein-relatedpeptidases (11), a disintegrin and the metalloproteinase familyof proteinases (ADAMs) (12), as well as several types ofcathepsins (6, 13). In the following, we summarize approachesfor tumor-specific activation of antitumoral agents capitalizingon dysregulated and overexpressed proteases in several typesof human cancer. We focus on MMP2, MMP9, MMP14,matriptase, uPA and cathepsin B.

Cathepsin B-activated Drugs

As outlined before, increased expression or altered localizationof proteases is typical in many types of cancer (2). Activationof prodrugs through cleavage by cathepsin B is a commonprinciple in pre-clinical and clinical development of anticancerdrugs. Cathepsin B is a lysosomal cysteine protease which isoverexperessed in numerous tumor types and can beassociated with the external cell surface (14). Procathepsin Bcan be activated by several proteases. Activated cathepsin Bcan then convert pro-uPA to uPA, which finally activatesplasminogen to plasmin, a serine protease, which can degradeseveral components of the extracellular matrix such as fibrin,fibronectin, laminin and proteoglycans and which is able toactivate MMPs (15, 16).

Several strategies for the construction of prodrugtherapeutics have been evaluated. A common motif here is thecombination of a therapeutic agent with an inactivating peptidesubstrate which is cleavable by proteases, thus leading to theliberation of the active therapeutic. In addition, more recently,

67

Correspondence to: Ulrich H. Weidle, Roche Pharma Research andEarly Development (pRED), Roche Diagnostics GmbH, ImNonnenwald 2, D 82372 Penzberg, Germany. E-mail:ulrich.weidle@roche.com

Key Words: Antibody−drug conjugates, antibody-directed enzymeprodrug therapy, Anthrax toxin, cytotoxic receptor ligand(s),granzyme B, probody, thapsigargin, review.

CANCER GENOMICS & PROTEOMICS 11: 67-80 (2014)

Review

Proteases as Activators for Cytotoxic Prodrugs in Antitumor Therapy

ULRICH H. WEIDLE, GEORG TIEFENTHALER and GUY GEORGES

Roche Pharma Research and Early Development (pRED), Roche Diagnostics GmbH, Penzberg, Germany

1109-6535/2014

such prodrug molecules were fused to a targeting moiety.Alternatively, peptide-linked therapeutic agents can also befused to a macromolecular carrier with or without a targetingmoiety (17). The targeting module can, for example, be aligand for a receptor, or an antibody-based moiety. In the caseof cathepsin B-activated prodrugs, conversion into the activedrugs should occur in the lysosome, but due to possibleassociation of cathepsin B with the cell surface in some typesof cancer, extracellular activation of the substrate is alsopossible. In addition, non-internalizable prodrug−receptorcomplexes will be activated exclusively at the cell surface.However, depending on the design of the substrate cleavagesite, prodrug activation in the serum by cathepsin B-likeactivity is a critical issue. In the present review article, wesummarize selected approaches for the design of cathepsin B-activated prodrugs, including cytotoxic antibody−drugconjugates (ADCs) which are at an advanced clinical stage.

Linking cytotoxic agents to high-molecular weight carriermolecules is a strategy to deliver drugs to tumors. The resultingmodification of the pharmacokinetic properties of the cytotoxiccompound mediates tumor accumulation in pre-clinical models,so-called passive targeting (18). For example, FCE 28068 (alsoknown as PK1) was constructed by linking doxorubicin to N-(2-hydroxypropyl) methylacrylamide copolymers via a tetrapeptidespacer which is cleavable by cathepsins, allowing release of theactive topoisomerase II inhibitor (19). This form of passivetumor targeting is also known as the enhanced permeability andretention effect, resulting from enhanced permeability of thetumor vasculature, facilitating extravasation of high-molecularweight proteins such as drug conjugates from the blood vessels(20). In a phase II study of FCE 28068 in chemotherapy-refractory patients with breast cancer, non-small cell lungcarcinoma, and colorectal cancer, 6/62 partial responses wereobserved with limited side-effects thus supporting the conceptthat polymer-based chemotherapeutics might have different andimproved anticancer activities (21). In addition, FCE 28068 waslinked to galactosamine, resulting in drug PK2, which binds tothe hepatic asialoglycoprotein receptor, thus achieving liver-specific doxorubicin delivery (22). Indeed, in a phase I study,liver-specific accumulation of PK2 after i.v. infusion was shown,as well as targeting to primary hepatocellular tumors (22).Similarly, the tubulin-binding compound paclitaxel−polyglumex(PPX) is a polymer−drug conjugate in which paclitaxel is linkedto a degradable polymer consisting of L-glutamic acid residues.The release of paclitaxel from the polymer backbone isdependent on cathepsin B and other lysosomal proteases (23).PPX was evaluated in a phase III trial versus docetaxel insecond-line treatment of patients with non-small cell lung cancer(24, 25). Although PPX and docetaxel led to similar survivalrates, different toxicity profiles were noted. Compared todocetaxel, PPX-treated patients had less frequent febrileneutropenia and alopecia, however, increased neurotoxicity wasnoted for PPX in comparison to docetaxel.

ADCs consist of a targeting moiety, a linker, and acytotoxic payload. The targeting module can be a ligand for areceptor or an antibody-based moiety directed against acancer-related antigen (26). After internalization of the ADCantigen−receptor complexes, cytotoxic payloads such ascalicheamycin, maytansine, duocarmycin, auristatin oririnotecan are released into the cytosol after cleavage of thelinker in the endosome (27). Chemically-unstable, enzyme-cleavable, but also stable, non-cleavable linkers have beenexplored in this context (26). Chemically labile linkers arehydrazine-based, whereas stable, non-cleavable linkers makeuse of thioester chemistry (26). The activity of ADCs withstable, non-cleavable linkers seems to depend on thedegradation of the thioester-linked antibody component in thelysosome, resulting in the release of the cytotoxic payload(28). Molecules with enzyme-cleavable linkers, on the otherhand, rely on the hydrolysis of a valine-citrulline (vc) linkerby cathepsin B. This results in the release of the cytotoxicpayload, such as the tubulin binders monomethyl auristatin E(MMAE) and F (MMAF) (29, 30, 31). ADCs based oncathepsin B cleavable linkers are among the most advancedADCs for the treatment of cancer. For example, brentuximabvedotin, an IgG1 antibody directed against cluster ofdifferentiation-30 (CD30) which is linked to vc MMAE isapproved for treatment of therapy-refractory Hodgkin’slymphoma and anaplastic large cell lymphoma (32, 33).Similarly, CDX-011 (glembatumumab vedotin) is an ADCcomposed of a fully human IgG2 anti-glycoprotein non-metastatic melanoma protein b/osteoactivin (gpNMB)antibody conjugated to MMAE via a vc linker (34). CDX-011is presently in phase II clinical studies in melanoma andbreast cancer (35-36). In addition, a prostate-specificmembrane antigen (PSMA)-ADC composed of a fully-humanized immunoglobulin G1 (IgG1) antibody and MMAElinked via a vc linker and is currently being evaluated inpatients with taxane-refractory metastatic prostate cancer inphase I clinical trials (37, 38). ADCs with analogous designtargeting nectin-4 and solute carrier family 4A4 (SLC4A4)are currently being evaluated in phase I clinical studies (27).Also undergoing phase I clinical studies are RG-7458,targeting CD22 in patients with non-Hodgkin’s lymphoma,and SGN75 and MDX 1203, two ADCs based on anti-CD70antibodies. These are presently being evaluated in patientswith non-Hodgkin’s lymphoma and clear cell renal cancer(27). In summary, it seems that no general design foroptimized linkers can be deduced based on previousexperience. In addition, ADCs with non-cleavable linkersmust be efficiently internalized to exert their cytotoxicfunction. On the other hand, ADCs with cleavable linkersmay mediate cytotoxicity by extracellular cleavage at thetumor cell surface in addition to cytotoxic pathways.Therefore, toxicity directed against antigen-negative tumorcells is expected only for ADCs with cleavable linkers (26).

CANCER GENOMICS & PROTEOMICS 11: 67-80 (2014)

68

Antibody-directed Enzyme Prodrug Therapy (ADEPT)

ADEPT is based on pre-targeting antibody−enzymecomplexes or corresponding fusion proteins to tumors,followed by administration of an inactive prodrug which isconverted to an active drug on the tumor cell surface by theseenzymes (39-41). After administration of the antibody-enzymeconjugates, the unbound conjugate is allowed to be clearedfrom the circulation. This can be accelerated by usingmannosylated conjugates (42), or a second antibody (43). Foreffective treatment, the prodrug has to be an optimal andspecific substrate for the enzyme and must not be able topenetrate cells. The inhibitory concentration 50 (IC50) of theliberated drug should be below 10 nM, and the ratio of theIC50s of the prodrug and drug should exceed 1,000 (40). Thereleased drug must be able to penetrate tumor cells and itsplasma half-life should be short enough to prevent leakage intohealthy tissues, but long enough to induce a bystander effecton neighbouring tumor cells. Activation of prodrug can bemediated by proteases, glycosidases, alkaline phosphatase,penicillin amidase, β-lactamase or cysteine deaminase (41).

In the following, we focus on prodrug activation byproteases. Most thoroughly explored, also in terms oftranslational applications in clinical studies, iscarboxypeptidase G2 (CPG2). CPG2 is a bacterial, Zn2+-dependent, homodimer forming exopeptidase, which cancleave a deactivating glutamate moiety from benzoic acidmustard (42, 43). Initial pre-clinical and clinical studies wereperformed with chemically produced antibody−enzymeconjugates. Subsequently, in order to overcome the variabilityof random chemical conjugation which results in productheterogeneity, recombinantly produced antibodymoiety−enzyme fusion proteins were explored (42, 43). Sucha fusion protein consisting of a single-chain variable domainfragment (scFv) directed against carcinoembryonic antigen(CEA), and CPG2 produced in Pichia pastoris for activationof a methotrexate-based prodrug was investigated in moredetail. This fusion protein, MFECP, bore branched mannoseresidues due to post-translational N-glycosylation (44).However, despite rapid clearance via mannose receptors,tumor-to-plasma ratios of 1,400:1 in the LS174T colonxenograft model and 339:1 in the SW122 colon xenograftmodel were recorded (45). Since the fusion protein localizedspecifically to CEA-expressing xenografts in the mouse, theprodrug was administered 6 h after administration of thefusion protein. Cycles could be repeated without obvioustoxicity (45). Consequently, this concept was explored inclinical studies in heavily pre-treated patients with CEA-expressing tumors. Treatment resulted in stable disease eightweeks after therapy in 11 out of 28 patients (46). Three cyclesof treatment could be applied after administration ofcyclosporine to suppress immune responses against the fusion

molecule (47). For second-generation approaches, humanizedantibodies and CPG2 with removed B- and T-cell epitopes areunder consideration (48).

Protease-cleavable Peptide−Thapsigargin (TG) Conjugates

The peptide−TG drug platform is based on coupling achemically modified amino acid containing TG with protease-cleavable peptides (49). TG is a sesquiterpene lactone isolatedfrom the plant Thapsia garganica (50). TG is a lipohilicmolecule and an inhibitor of the ubiquitously expressedsarcoplasmic/endoplasmic reticulum (ER) calcium ATPasepump (SERCA) (51, 52). The SERCA pump is essential forthe function of the ER. SERCA inhibition leads to a 3-5 foldincrease in intracellular free Ca2+ accompanied by thedepletion of the ER Ca2+, pool resulting in cell deathindependent of proliferation status (53, 54). Because thetoxicity of TG is not cell type-specific, systemic administrationof TG would be associated with significant toxicity. Therefore,prodrug molecules were designed with a carrier peptide thatdoes not allow TG to enter the cell until the TG analog isliberated by proteolysis. A PSMA-targeted prodrug wasevaluated (55). PSMA is a membrane-bound protease with itscatalytic domain exposed to the extracellular space. It exhibitsdual exopeptidase activity as both a pteroyl poly γ glutamylexopeptidase (folate hydrolase) and N-acetylated α-linked aciddipeptidase (56, 57). PSMA is expressed in tumor-associatedendothelium, thus potentially enabling TG prodrug activationin the tumor microenvironment (53). The TG-based prodrugG202 (Figure 1A), consisting of a PSMA-specific peptideconjugated to an analog of TG, induced tumor regression inseveral xenograft models at doses which were minimally toxicto the host (55). Critical issues are, however, the expression ofPSMA in the proximal tubules of the kidney and in the brain(58). Ongoing clinical studies will show whether a therapeuticwindow combining efficacy with an acceptable toxicity profilecan be defined.

A new emerging strategy for tumor therapy is ablation ofsupporting stromal cells in the tumor microenvironment bytargeting cancer-associated fibroblasts (59). Fibroblast activationprotein (FAP), a post-prolyl endopeptidase predominantlyexpressed in cancer-associated fibroblasts with a limited patternof expression in normal tissues has emerged as an interestingtarget (60). A FAP-activated peptidyl-TG prodrug killed FAP-positive human cancer cells at low nanomolar concentrations,efficiently reduced growth of MCF-7 and LNCaP tumorxenografts, concurrent with ablation of tumor stroma (61).However, activation of the prodrug in plasma was observedindicating a possible limitation of this approach (61). Ongoingefforts are based on activation of TG-based prodrugs byprostate-specific antigen and human kallikrein-2, proteolyticenzymes secreted by the prostate with expression orders of

Weidle et al: Proteases as Mediators of Cytotoxicity in Tumor Therapy (Review)

69

magnitude greater than in other tissues (50, 62). Inactivation ofthe enzymatic activity of prostate-specific antigen and humankallikrein-2 in the circulation by serum proteins is anencouraging aspect of this approach (63, 64).

Protease-based Activation of Cytotoxic ReceptorLigands and Other Cytotoxic Proteins

Use of tumor necrosis factor (TNF) as an antitumoral agent ishampered by its severe systemic toxicity, allowing its clinicalapplication only in the limb perfusion setting (65). Similarly,CD95−CD95L interaction can be exploited for tumor therapydue to its pro-apoptotic effect after delivery of CD95L or anagonistic antibody to tumor cells. However, this approach isalso limited because of the induction of apoptosis in normalliver cells, resulting in acute liver failure and thus the necessityfor local activiation of CD95L in tumor tissue (66). In order totackle these issues, an antibody-based targeting moiety wasfused to CD95L or TNF via a linker which can be cleaved bytumor-associated proteases such as uPA, tissue type PA, MMP2or MMP9, to effect specific tumor targeting. In the followingexamples, the antibody moiety is directed against FAP (67).

A TNF-selectokine, a homotrimeric fusion proteinconsisting of a scFv antibody fragment directed against FAPas a targeting module, a trimerization domain derived fromtenascin C, TNF, and a linker consisting of protease cleavagesites followed by the extracellular domain (ECD) of TNFreceptor-1 (TNFR1) was evaluated (68). This cell-boundselectokine mimics membrane-bound TNF. The unique featureof this TNF selectokine is the presence of the ECD of TNFR1as an intra-molecular inhibitor, thus creating a prodrug. Theprodrug exerts minimal TNF activity, but can be activatedseveral thousand-fold after cleavage by tissue-typeplasminogen activator (tPA). Activated prodrug was shown tobear TNFR-dependent cytotoxic activity in targeted andadjacent FAP-negative cells. Making use of the same speciescrossreactive anti-FAP moiety, analogous murine TNFconstructs were evaluated in which the natural processing sitefor TNFα converting enzyme was mutated in order to avoidpotential cleavage of the prodrug (69). Such cell surface-targeted fusion proteins induced strong cytotoxicity in humanHT1080 fibrosarcoma cells expressing MMPs and transfectedwith FAP. Antigen-positive and -negative cells were affectedin a juxtatropic mode. These findings were extrapolated tohuman TNF prodrugs bearing either a uPA-selective or a dualuPA−MMP-2-specific linker (70). The same basic principlewas applied to the design of CD95L-based prodrugs (71).Again, FAP-transfected HT1080 cells were used for the invitro and in vivo evaluation of the corresponding prodrugmolecules. These prodrugs are composed of an scFv-directedagainst FAP, followed by the ECD of CD95L, a protease-sensitive linker which can be cleaved by MMP-2 and relatedproteases and the ECD of CD95 as carboxyterminal inhibitory

domain (Figure 1B and C). In order to achieve tumor-specificactivation of CD95, CD95L is only unmasked after binding ofthe fusion protein to FAP-positive tumor cells and subsequentcleavage of the linker by tumor-associated proteases. Thecleaved scFv-CD95L remains membrane-bound and therebymimics membrane CD95L. It was shown that activation ofCD95L in vitro is dependent on CD95 binding and protease-dependent processing. In vivo efficacy was demonstrated byintra-tumoral injection of FAP-transfected HT1080 cellstogether with the prodrug one day after tumor inoculation with1.5×106 cells. This resulted in 80% tumor growth inhibitionwhich was not observed when non-transfected HT1080 cellswere co-injected. A critical issue of the approaches, asoutlined, is the potential activation of the fusion proteins byserum proteases. Therefore, efficient targeting of the tumorwith these effector fusion moieties is mandatory becauseexpression of tumor-associated proteases is not strictly tumor-specific.

Mellitin is a 26 amino acid pore-forming protein derivedfrom honey bee venom (72). Formation of pores in the cellmembrane induces cell lysis. In order to target the cytotoxicactivity of mellitin to MMP2-overexpressing tumor cells, abiologically-inactive fusion protein consisting of mellitin andavidin connected by an MMP2 cleavage site was designed(73). This fusion protein mediated strong cytotoxic activityafter activation by MMP2 in MMP2-overexpressing tumorcells, while only little cytotoxicity was seen in cells with lowMMP2 activity. In vivo tumor shrinkage was also shown.

Fusogenic membrane glycoproteins were identified inseveral viruses inducing cell-cell fusion and syncytiaformation, which ultimately results in cytotoxicity and celldeath (74). The Gibbon Ape Leukemia Virus envelopeglycoprotein (GALV) was assessed for its potential fortargeted killing of glioma cells. Thus, GALV was linked to theextracellular domain of CD40 ligand as a binding blockingmoiety, via an MMP or factor Xa protease-cleavable linker, ora non-cleavable linker. Transfection of constructs withcleavable linkers selectively induced cytotoxicity in gliomacells in contrast to non-cleavable constructs (75). Targeting inthis case is necessary due to the ubiquitous expression of theGALV receptor, pituitary-specific positive transcription factor-1 (PIT-1). Adenoviral vectors expressing fusion proteins, asoutlined above, were shown to exhibit significant antitumoralpotential for gene therapy of glioma gene therapy (76).

Protease-mediated Activation of Anthrax Toxin for Tumor Therapy

Bacillus anthracis, a bacterium that forms highly resistantspores, can cause lethal infections after inhalation or ingestion.The causative agent for lethal infections is anthrax toxin whichconsists of three components which individually are non-toxic:protective antigen, lethal factor (LF) and edema factor (EF)

CANCER GENOMICS & PROTEOMICS 11: 67-80 (2014)

70

(77, 78). The unique processing and assembly of thecomponents of the anthrax toxin was exploited in the design ofanti-tumoral agents and for the delivery of other toxins intotumor cells. The action of anthrax toxin starts by binding ofprotective antigen to two broadly-expressed receptors: tumorendothelial marker-8 and capillary morphogenesis protein-2(79). After binding to its receptors, protective antigen is

processed by proteases such as furin, and the remainingreceptor-bound moiety of protective antigen forms a ring-shaped heptamer which transports EF, LF, or both of them intothe cell to exert their cytotoxic function. The resultingcomplexes are referred to as lethal toxin or edema toxin. Thesecomplexes are internalized by receptor-mediated endocytosisvia a clathrin-dependent process, and after formation of a

Weidle et al: Proteases as Mediators of Cytotoxicity in Tumor Therapy (Review)

71

Figure 1. Display of G202 (A) and cluster of differentiation 95 ligand (CD95L)-based (B) prodrugs. A: Schematic presentation of G202. The toxicentity thapsigargin (orange) is fused via an aliphatic linker (pink) to a cleavable prostate specific membrane antigen (PSMA)-specific peptide (grey).Protease cleavage sites are indicated by arrows. B: CD95L-based prodrug. The proteins are displayed as ribbons coloured according to domainfunction. The CD95L prodrug is composed of the CD95 receptor (magenta, pdbcode 3TJE) as the protecting moiety, followed by a trimerizationcoiled-coil domain of tenascin C (shown in green, helix modeling is based on pbdcode 2F3M), a cleavable linker (grey), a modeled scFv directedagainst fibroblast activation protein (FAP) (sc FAP) heavy and light variable domains displayed in dark and light blue and their correspondinglinker shown in pink, a polyaspartate flag (red) and finally the CD95L domain as an effector (orange, pdbcode 4MSV). The lower left shows the cyclicconfiguration of the CD95L-based prodrug, bringing the ligand and the corresponding receptor together. The protease cleavage site is indicated byan arrow. The trimeric prodrug is shown in the middle and lower right. The head-to-tail configuration is not shown. The models were generated basedon available structural data of domains or entities and were assembled and minimized using DiscoveryStudio40 (128, 129).

channel in the membrane of the endosome, LF and EF aretransported into the cytosol. LF is a zinc-dependentmetalloprotease that cleaves mitogen-activated protein kinasekinases (MAPKK). EF is a calmodulin-dependent adenylatecyclase which leads to intracellular elevation of adenosine-monophosphate or cyclic adenosine monophosphate levels,thus interfering with the balance of intracellular signaling (79,80). Anti-angiogenic effects were observed for both LF andEF due to their effects on MAPKK inhibition and cAMPlevels in endothelial cells. LF interferes with MAPK signalingin Ha-ras transformed NIH-3T3 cells, inhibits their growth insoft-agar, reverts their transformed phenotype and inhibitstheir growth in athymic mice, partly through inhibition ofangiogenesis (81). Human melanoma cells were shown to beequisitely sensitive to LF in combination with a smallmolecule MAPK inhibitor, inducing apoptosis, whereasnormal melanocytes were not killed but arrested in the G1-phase of the cell cycle (82). Due to the broad expression ofthe receptors, tumor-specific targeting of anthrax toxin isessential to generate a therapeutic window. Activation ofmodified anthrax toxins on the surface of tumor cells is astraightforward strategy. MMPs such as MMP2, MMP9,MMP14 and matriptase, as well as uPA are proteases whichare overexpressed on the surface of tumor cells or thesurrounding stromal cells. Replacement of the furin cleavagesite in protective antigen by those specific for MMPs or uPAleads to tumor-specific activation of protective antigen.Making use of MMP-activated LF, antitumor activity wasobserved in tumor xenograft models harbouring the V600Eactivating rapidly accelerated fibrosarcoma B (BRAF)mutation, but also in xenografts with non-mutated BRAF (83).In addition, anthrax toxin inhibits tumor angiogenesis byinterference with endothelial cell proliferation, migration andtube formation for which MAPK activation plays a crucialrole. Delivery of other toxin effector moieties such as theADP-ribosylation domain of pseudomonas exotoxin intotumor cells making use of the properties of the anthrax toxincomplex is another potential application of this system. Thus,fusion of passenger proteins with amino acids 1-254 of LFallows their translocation into, and eventually killing of cellsexpressing PA receptors with toxic effector molecules such asthe ADP-ribosylating moiety of pseudomonas exotoxin, the Asubunit of diphtheria toxin, and the A subunit of Shiga toxin(84, 85, 86). Similarly, when amino acids 1-254 of LF werecombined with mutated protective antigen proteins, theresulting fusion protein (FP59) mediated selective killing ofMMP-overexpressing tumor cells such as fibrosarcoma, breastand melanoma (83). Analogous observations were made withuPA activated protective antigen which resulted in selectivekilling of uPAR-expressing tumor cells (87). A critical issueis that neither the MMP nor uPA-activation systems are strictlytumor-specific, thus side-effects after administration of theseagents are an issue.

Granzyme B as Effector Moiety for Cytolytic Fusion Proteins

Receptor−ligand or antibody-based moieties directed againsttumor antigens genetically fused to granzyme B as a cytotoxicprinciple are presently pre-clinically evaluated as anti-tumoragents (88-92). Granzyme B is member of a family of serineproteases with five different human granzymes (A, B, H, K andM). They are stored in the granules of natural killer cells andcytotoxic T-lymphocytes. Granzyme B is an inducer of granule-mediated apoptosis. Granzyme B cytolytic fusion proteins asdescribed above are internalized and induce apoptosis afterrelease from the endosome. Once in the cytosol, granzyme Bcan directly activate different apoptotic pathways and atdifferent levels, ensuring induction of apoptosis even whensome apoptotic pathways are blocked. Important granzyme Bsubstrates include caspase-3 and other initiator and effectorcaspases, the inhibitor of caspase-activated DNase, and BH3-interacting domain death agonist (BID). In addition, granzymeB can cleave further death-substrates such as poly(ADPribose)polymerase, DNA-dependent protein kinase,components of the cytoskeleton and the nuclear mitoticapparatus as well as proteins involved in stress response andcellular homeostasis. Numerous variations of granzyme Bfusion proteins were studied. Transforming growth factor-α andvascular endothelial growth factor (VEGF) (Figure 2A) werethe targeting domains in the context of ligand−granzyme Bfusions, while CD64, GP240, Lewis Y antigen, humanepidermal growth factor receptor 2 (HER2) and humanluteinizing hormone receptor were evaluated as targets forantibody-based moieties of granzyme B-derived cytolyticfusion proteins (88). Despite encouraging in vitro and in vivoefficacy data in various tumor-derived cell lines and xenograftmodels, several problems preventing optimal efficacy of suchgranzyme B-based cytolytic fusion proteins have emerged: (i)granzyme B inhibitor PI-9 is expressed in several tumor typesand their corresponding cell lines, limiting efficacy; and (ii)granzyme B possesses a strong positive surface charge whichcauses non-specific binding to negatively charged cell surfaces.This leads to adsorption and off-target effects, as well astrapping of the internalized antigen−fusion proteins inendosomes, resulting in insufficient delivery of granzyme Binto the cytosol. Consequently, highly variable efficacy of thecorresponding cytolytic fusion proteins was observed. Onerecent achievement is the generation of granzyme B mutantsinsensitive to inhibition by PI-9 by mutating positionsinteracting with the variable reactive center loop of PI-9, whichis its primary site of interaction with proteases. Importantly,one of these variants, granzyme B R201K, exhibits the sameenzymatic activity in the absence or presence of PI-9 (93).Mutation of sequences responsible for interaction withheparan-sulfate-containing molecules resulted in diminishedcell binding but had a negative impact on cytotoxicity (94).

CANCER GENOMICS & PROTEOMICS 11: 67-80 (2014)

72

Residues 96-103 and 221-226 confer a positive surface chargeon mature granzyme B, causing interactions withglycosaminoglycans. Substitution of such cationic residues wasshown to have no impact on the enzymatic and pro-apoptoticactivity of granzyme B but reduced binding to antigen-negativecells (95). Improved efficacy of granzyme B-based cytolyticfusion proteins was achieved by co-administration withchloroquine, which accumulates in acidic compartments andlysosomes and causes osmotic rupture of these vesicles (95-97). Further approaches for improved endosomal release arethe incorporation of (i) protein transduction domains derivedfrom pseudomonas exotoxin or diphtheria toxin; (ii) cell-

penetrating peptides or membrane transfer sequences into thecorresponding cytolytic fusion proteins; and (iii) insertion ofsynthetic, multifunctional linkers, as outlined below (98). Theselinkers are composed of a membrane transfer peptide flankedby an endosomal cleavable peptide and a cytosolic cleavablepeptide (98). Potential immunogenicity after introduction ofamino acid substitutions has not yet been properly investigated.An interesting example for the improvement of efficacy is afusion protein between a scFv directed against HER2, afusogenic pH-sensitive peptide, and granzyme B (99). Tumorcells resistant to lapatinib or Herceptin and expressing MDR1showed no crossresistance to the granzyme B-based fusion

Weidle et al: Proteases as Mediators of Cytotoxicity in Tumor Therapy (Review)

73

Figure 2. Display of a targeted granzyme B-based fusion protein and two different versions of a probody. A: Granzyme B-vascular endothelial growthfactor 121 (VEGF121) fusion protein. The model is based on the structures of granzyme B (orange, pdbcode 3PW9) and VEGF121 (blue, pbdcode1FQ4). The models were generated based on available structural data of domains or entities and were assembled and minimized usingDiscoveryStudio40 (128, 129). B: Schematic presentation of a probody. The principle design is derived from a probody directed against vascularcellular adhesion molecule 1 (VCAM1) (117). The peptide with lid function (magenta) has affinity to the antibody paratope. Dark and light orangevariable domains were taken from a non-related Fab complex (pdbcode 4RNRX), the peptide is linked via a flexible glycine-serine linker (pink) anda cleavable sequence (grey) to the heavy chain variable domain (VH domain). The constant part of the antibody is shown in grey. Protease cleavagesites are shown in grey. C: Schematic presentation of a bi-specific probody. Targeting to antigen 1 is mediated by the V domains shown in dark andlight blue. The molecule is based on the knob-into-hole differentiation of the two fragment crystallization region (Fcs), highlighted in dark and lightgrey. V-domains of an antibody directed against antigen 2 are fused to the C-termini of the Fc portions by connectors with and without a proteasecleavage site. VH and VL domains are disulfide bridged. Binding functionality is reconstituted after cleavage of one of the two linkers. The cleavableelement is shown in grey and G4S in pink. The schematic presentation is based on the description of a bi-specific prodbody directed against humanepidermal growth factor receptor 3 (HER3) and cellular proto-oncogene of mesenchymal epithelial transition factor (c MET) (120). The proteasecleavage site is indicated by an arrow and the disulfide bridge connecting VL and VH domains is shown in green.

protein, and IC50s between 20 and 90 nM were noted forvarious HER2-positive breast cancer cell lines. Significanttumor growth suppression was noted in vivo in mice bearingBT474 breast tumors after i.v. administration (99).

Protease-cleavable Adapters for Improved Cytolytic Fusion Proteins

Cytolytic fusion proteins are composed of a targeting moduleand a protein-based cytotoxic moiety. The targeting modulecan be a ligand for a cell surface receptor or an antibody-basedmoiety, the therapeutic principle as a rule is an enzyme withcytotoxic properties. The concept of cleavable adapters wasraised (98, 100) to improve the characteristics of cytolyticfusion proteins. The prototype adapter is composed of acytosolic cleavable peptide, a cell-penetrating peptide (CPP)which is a membrane transfer peptide, and an endosomalcleavable peptide (98). For example, a cytosolic cleavablepeptide was designed by incorporating cleavage sites forcaspases 1, 3 and 7 and a cytosolic protease from yeast into acorresponding peptide sequence (101). After cleavage, thecompound is trapped within the cell, converting a formerlycell-permeable compound into a cell-impermeable drug. CPPscan penetrate membranes by an energy-independent processor by an endocytosis-related process (102, 103). CPPs are ofdiverse origin, 10-16 amino acids long and can be groupedinto peptides composed of basic amino acids such as poly-lysine or poly-arginine, α- helical, and β-sheet peptides (102,103). Examples are TAT from HIV, penetratin fromDrosophila homeotic transcription factor antennapedia, or thepreS2-domain of hepatitis B virus surface protein (104). Thefunction of membrane transfer peptides and CPPs is toimprove cellular uptake and deliver the cytotoxic moleculedirectly into the cytosol. Endosomal cleavable peptides on theother hand are based on dedicated translocation sequenceswhich are activated after protease cleavage and usually arederived from diphtheria toxin or pseudomonas exotoxin (105).One exemplary adapter was designed to improve the uptakeand transport of the catalytic cytotoxic moiety into the cytosoland subsequently trap it in the cell. This trapping might beresponsible for the reduced cytotoxicity which was associatedwith long circulation times and damage of target-negative cells(98). Saporin linked to EGF with or without adapter wasevaluated, and reduced non-specific cytotoxicity was noted byinserting a cleavable adapter, as outlined (101). Similarly,treatment of human EGFR-transfected murine tumor cells withadapter-containing and adapter-free fusion protein showedenhanced efficacy for the adapter-containing fusion protein(106). Surprisingly, however, in vitro cytotoxicity was notaffected by the insertion of an adapter (101). Another exampleis a fusion protein between a humanized scFv directed againstCD64 and human ribonuclease angiogenin (107). Insertion ofan adapter improved cytotoxicity 20-fold, however, serum

stability was reduced, resulting in complete cleavage of theadapter after 1 h, whereas the adapter-free fusion protein wasstable with no evidence for cleavage after 24 h. Consequently,a fusion protein bearing a modified adapter with deletedendosomal cleavable peptide was evaluated. Thus, whilecytotoxicity of this molecule was still 10-fold better than theadapter-free fusion protein, almost no cleavage was observedafter 24 h in the presence of serum.

Finally, the findings as outlined above indicate that anadapter cannot be transferred between fusion proteins bydefault to improve cytotoxicity and stability but has to beoptimized specifically for each fusion protein underconsideration.

Activation of Antibodies with ImpairedFunctionality by Proteases

Antibody-based therapies often exploit targets which play acrucial role in physiological processes and therefore can beassociated with serious toxicities (108). For example, blockageof delta-like ligand 4, a notch ligand, leads to inhibition ofangiogenesis in pre-clinical models. However, liver pathology,gut and T-cell toxicity, and induction of vascular neoplasmswere observed in mice and rats (109, 110, 111). Similarly, theanti-angiogenic agent bevacizumab, a VEGF-neutralizingmonoclonal antibody, can cause side-effects such ashypertension, proteinuria and gastrointestinal symptoms, andin rare cases, gastrointestinal perforation and arterialthromboembolic complications (112). Skin toxicity as dose-limiting toxicity was observed after treatment of cancer patientswith antibodies or antibody conjugates directed against EGFR(113, 114). An increased risk of serious infection and a dose-dependent risk for malignancies was observed in patients withrheumatoid arthritis after anti-TNF antibody therapy (115).HER2-directed therapy with trastuzumab caused an incidenceof 0.2-3.8% of class III congestive heart failure in the adjuvanttrastuzumab trials of patients with HER2-positive breast cancer.This was probably due to the expression of HER2 oncardiomyocytes in addition to tumor tissues (116). This list ofserious target-related side-effects could be extended tonumerous other examples. Therefore, the probody concept,activation of an inactive preform of an antibody by tumor-associated proteases was raised (117) (Figure 2B). Accordingly,a probody with masked antibody binding site was designed bytethering a peptide to the N-terminus of an antibody-derivedbinding domain via a flexible linker containing a protease-cleavage site in the context of a scFv-Fc format (117). In aproof-of-concept experiment, such a probody directed againstvascular cell adhesion molecule-1, a marker of atheroscleroticplaques, was evaluated (117). In vitro activation of the probodywith MMP1 resulted in a 200-fold increase of binding affinityand restored binding to tissue sections from atheroscleroticmice ex vivo demonstrating MMP-dependent probody

CANCER GENOMICS & PROTEOMICS 11: 67-80 (2014)

74

accumulation in aortic plaques. Analysis of frozen aorta tissuesections from mice injected with such probodies revealed largeplaques that stained strongly for the presence of the antibody tovascular cell adhesion molecule-1 and probody. Regardingpossible oncological applications of the probody technology,antibodies against matriptase were successfully used forimaging of colon cancer xenografts, pointing to matriptase as apossible activator of probodies (118). Tumor-specific activationof therapeutic antibodies potentially results in an improvedtherapeutic index (119). For example, the EGFR-targetingantibody cetuximab was converted into a probody by fusing amasking peptide which blocks binding to EGFR to the N-terminus of the L-chain. The masking peptide (21 amino acids)was fused to a cleavable linker (26 amino acids), sensitive touPA, matriptase, and legumain. This probody possessed in vitroactivity equivalent to cetuximab, was activated ex vivo byxenograft tumors, and suppressed tumor growth in vivo inmouse xenograft models with efficacy equivalent to cetuximab.However, due to the lack of crossreactivity of cetuximab withrodent EGFR, final conclusions are limited. This probody wasshown to be activated by human tumor samples in vitro. Theprobody was nearly inert in the circulation and possessedmarkedly improved safety with respect to cutaneous toxicityand increased plasma half-life in non-human primates.Therefore, this probdody could be dosed safely at much higherlevels than cetuximab (119). Finally, proof-of-conceptexperiments have shown that bi-specific antibodies withrestricted binding functionality can be activated by proteolyticprocessing (120). The principal design of such a bi-specificprobody is shown in Figure 2C. A bispecific antibody whichbinds HER3 in a conventional, bivalent IgG-like manner andwhich carries in addition a cellular proto-oncogene ofmesenchymal epithelial transition factor (C-MET) bindingmoiety composed of a variable heavy chain (VH)- and avariable light chain (VL)-domain which can be connected bya disulfide bond was designed (120). The H- and L- chaindomains are linked to the C-termini of the constant heavy chain3 domains, respectively. One of the connectors contains aprotease cleavage site for MMP2, MMP9 or uPA. Aftercleavage of the protease-sensitive connector, steric hindranceis resolved and antigen access and affinity in vitro is fullyrestored.

Proteases and Imaging

Tumor-associated expression of enzymatically activematriptase has been exploited for tumor imaging (122).Matriptase is a type II transmembrane protease of the serineprotease family. It is found on the surface of epithelial cellsand its activity is tightly regulated by its cognate inhibitor,hepatocyte growth factor activator inhibitor-1. In normaltissues, the ratio of matriptase to its inhibitor is low. This ratioincreases during progression of some types of cancer, resulting

in active matriptase on the cell surface (121, 122). Using arecombinant human antibody against an epitope covering theactive site of matriptase, it was shown that tumor epitheliumcan be visualized selectively (123, 124). Live cell fluorescenceimaging showed that the antibody localized only to the surfaceof matriptase-positive tumor cells, and immunofluorescenceexperiments with tissue microarray sections from primary andmetastatic colon cancer indicated active matriptase in 68% ofprimary and metastatic colon cancer specimens (122). Nearinfrared and single-photon emission computed tomographicimaging visualized matriptase in human colon cancer tumorsin a patient-derived xenograft model (122). The antibody usedin these studies was crossreactive with the murine matriptaseortholog epithin, and radiolabeled antibody was taken-up bytumor xenografts. Specificity of the interaction wasdemonstrated by inhibition of uptake of the antibody afterinjection of ecotin, a macromolecular inhibitor of matriptase,before injection of the radiolabeled antibody. Accordingly,active site probes may become important tools for theelucidation of the role of proteases in the pathogenesis ofdifferent types of cancer.

Concluding Remarks

Personalized medicine is one of the key issues of cancertherapy (125-127) meaning that patients are treated based onmolecular characteristics as revealed by diagnostic testsand/or high throughput sequencing. Overexpressed plasmamembrane-associated receptors, oncogenic driver mutationsand other deregulated cancer promoting pathways are targetsfor small molecule- or antibody-derived therapeutics,resulting in inhibition of migration, induction of apoptosis,inhibition of angiogenesis, stimulation of an antitumoralimmune response, or synthetic lethality. Knowledge about,and specification of, the expression pattern of proteases intumors, and quantification of the ratio of inactiveproenzymes versus enzymatically active ones in individualtumors could advise treatment with prodrugs activated by theproteases characteristic of the individual tumors. Finally,these findings will also have an impact on optimal tumorimaging. A further future application is the administration ofcytolytic fusion proteins with proteases such as granzyme Bas an effector moiety.

References1 López-Otín C and Matrisian LM: Emerging roles of proteases in

tumour suppression. Nat Rev Cancer 7: 800-808, 2007.2 Coussens LM, Fingleton B and Matrisian LM: Matrix

metalloproteinase inhibitors and cancer: Trials and tribulations.Science 295: 2387-2392, 2002.

3 Overall CM and López-Otín C: Strategies for MMP inhibition incancer: innovations for the post-trial era. Nat Rev Cancer 2: 657-672, 2002.

Weidle et al: Proteases as Mediators of Cytotoxicity in Tumor Therapy (Review)

75

4 Hadler-Olsen E, Winberg JO and Uhlin-Hansen L: Matrixmetalloproteinases in cancer: Their value as diagnostic andprognostic markers and therapeutic targets. Tumour Biol 34:2041-2051, 2013.

5 Vartak DG and Gemeinhart RA: Matrix metalloproteases:Underutilized targets for drug delivery. J Drug Target 15: 1-20,2007.

6 Mohamed MM and Sloane BF: Cysteine cathepsins:Multifunctional enzymes in cancer. Nat Rev Cancer 6: 764-775, 2006.

7 D’Alessio S and Blasi F: The urokinase receptor as an entertainerof signal transduction. Front Biosci (Landmark Ed) 14: 4575-4587, 2009.

8 Schmitt M, Harbeck N, Thomssen C, Wilhelm O, Magdolen V,Reuning U, Ulm K, Höfler H, Jänicke F and Graeff H: Clinicalimpact of the plasminogen activation system in tumor invasionand metastasis: Prognostic relevance and target for therapy.Thromb Haemost 78: 285-296, 1997.

9 Min HY, Doyle LV, Vitt CR, Zandonella CL, Stratton-ThomasJR, Shuman MA and Rosenberg S: Urokinase receptorantagonists inhibit angiogenesis and primary tumor growth insyngeneic mice. Cancer Res 56: 2428-2433, 1996.

10 Blasi F and Carmeliet P: uPAR: A versatile signallingorchestrator. Nat Rev Mol Cell Biol 3: 932-943, 2002.

11 Oikonomopoulou K, Diamandis EP and Hollenberg MD:Kallikrein-related peptidases: proteolysis and signaling in cancer,the new frontier. Biol Chem 391: 299-310, 2010.

12 Murphy G: The ADAMs: Signalling scissors in the tumourmicroenvironment. Nat Rev Cancer 8: 929-941, 2008.

13 Benes P, Vetvicka V and Fusek M: Cathepsin D−many functions ofone aspartic protease. Crit Rev Oncol Hematol 68: 12-28, 2008.

14 Koblinski JE, Ahram M and Sloane BF: Unraveling the role ofproteases in cancer. Clin Chim Acta 291: 113-135, 2000.

15 Danø K, Andreasen PA, Grøndahl-Hansen J, Kristensen P,Nielsen LS and Skriver L: Plasminogen activators, tissuedegradation, and cancer. Adv Cancer Res 44: 139-266, 1985.

16 Baramova EN, Bajou K, Remacle A, L’Hoir C, Krell HW, WeidleUH, Noel A and Foidart JM: Involvement of PA/plasmin systemin the processing of pro-MMP-9 and in the second step of pro-MMP-2 activation. FEBS Lett 405: 157-162, 1997.

17 Choi KY, Swierczewska M, Lee S and Chen X: Protease-activated drug development. Theranostics 2: 156-178, 2012.

18 Seymour LW: Passive tumor targeting of soluble macromoleculesand drug conjugates. Crit Rev Ther Drug Carrier Syst 9: 135-187,1992.

19 Duncan R, Kopecek J, Rejmanová P and Lloyd JB: Targeting ofN-(2-hydroxypropyl)methacrylamide copolymers to liver byincorporation of galactose residues. Biochim Biophys Acta 755:518-521, 1983.

20 Seymour LW, Ulbrich K, Steyger PS, Brereton M, Subr V,Strohalm J and Duncan R: Tumour tropism and anti-cancerefficacy of polymer-based doxorubicin prodrugs in the treatmentof subcutaneous murine B16F10 melanoma. Br J Cancer 70: 636-641, 1994.

21 Seymour LW, Ferry DR, Kerr DJ, Rea D, Whitlock M, PoynerR, Boivin C, Hesslewood S, Twelves C, Blackie R, Schatzlein A,Jodrell D, Bissett D, Calvert H, Lind M, Robbins A, Burtles S,Duncan R and Cassidy J: Phase II studies of polymer-doxorubicin (PK1, FCE28068) in the treatment of breast, lungand colorectal cancer. Int J Oncol 34: 1629-1636, 2009.

22 Seymour LW, Ferry DR, Anderson D, Hesslewood S, Julyan PJ,Poyner R, Doran J, Young AM, Burtles S and Kerr DJ: Hepaticdrug targeting: Phase I evaluation of polymer-bound doxorubicin.J Clin Oncol 20: 1668-1676, 2002.

23 Shaffer SA, Baker-Lee C, Kennedy J, Lai MS, de Vries P, BuhlerK and Singer JW: In vitro and in vivo metabolism of paclitaxelpoliglumex: I dentification of metabolites and active proteases.Cancer Chemother Pharmacol 59: 537-548, 2007.

24 Chipman SD, Oldham FB, Pezzoni G and Singer JW: Biologicaland clinical characterization of paclitaxel poliglumex (PPX, CT-2103), a macromolecular polymer-drug conjugate. Int JNanomedicine 1: 375-383, 2006.

25 Paz-Ares L, Ross H, O’Brien M, Riviere A, Gatzemeier U, VonPawel J, Kaukel E, Freitag L, Digel W, Bischoff H, García-Campelo R, Iannotti N, Reiterer P, Bover I, Prendiville J,Eisenfeld AJ, Oldham FB, Bandstra B, Singer JW and BonomiP: Phase III trial comparing paclitaxel poliglumex vs. docetaxelin the second-line treatment of non-small-cell lung cancer. Br JCancer 98: 1608-1613, 2008.

26 Ducry L and Stump B: Antibody-drug conjugates: Linkingcytotoxic payloads to monoclonal antibodies. Bioconjug Chem21: 5-13, 2010.

27 Sassoon I and Blanc V: Antibody-drug conjugate (ADC) clinicalpipeline: A review. Methods Mol Biol 1045: 1-27, 2013.

28 Erickson HK, Park PU, Widdison WC, Kovtun YV, Garrett LM,Hoffman K, Lutz RJ, Goldmacher VS and Blättler WA:Antibody-maytansinoid conjugates are activated in targetedcancer cells by lysosomal degradation and linker-dependentintracellular processing. Cancer Res 66: 4426-4433, 2006.

29 Widdison WC, Wilhelm SD, Cavanagh EE, Whiteman KR, LeeceBA, Kovtun Y, Goldmacher VS, Xie H, Steeves RM, Lutz RJ,Zhao R, Wang L, Blättler WA and Chari RV: Semisyntheticmaytansine analogues for the targeted treatment of cancer. J MedChem 49: 4392-4408, 2006.

30 Carter PJ and Senter PD: Antibody−drug conjugates for cancertherapy. Cancer J 14: 154-169, 2008.

31 Beck A, Haeuw JF, Wurch T, Goetsch L, Bailly C and CorvaïaN: The next generation of antibody−drug conjugates comes ofage. Discov Med 10: 329-339, 2010.

32 Senter PD and Sievers EL: The discovery and development ofbrentuximab vedotin for use in relapsed Hodgkin lymphoma andsystemic anaplastic large cell lymphoma. Nat Biotechnol 30:631-637, 2012.

33 Minich SS: Brentuximab vedotin: A new age in the treatment ofHodgkin lymphoma and anaplastic large cell lymphoma. AnnPharmacother 46: 377-383, 2012.

34 Naumovski L and Junutula JR: Glembatumumab vedotin, aconjugate of an anti-glycoprotein non-metastatic melanomaprotein B mAb and monomethyl auristatin E for the treatment ofmelanoma and breast cancer. Curr Opin Mol Ther 12: 248-257,2010.

35 Keir CH and Vahdat LT: The use of an antibody−drug conjugate,glembatumumab vedotin (CDX-011), for the treatment of breastcancer. Expert Opin Biol Ther 12: 259-263, 2012.

36 Hoashi T, Sato S, Yamaguchi Y, Passeron T, Tamaki K andHearing VJ: Glycoprotein nonmetastatic melanoma protein b, amelanocytic cell marker, is a melanosome-specific andproteolytically released protein. FASEB J 24: 1616-1629, 2010.

37 Wang X, Ma D, Olson WC and Heston WD: In vitro and in vivoresponses of advanced prostate tumors to PSMA ADC, an

CANCER GENOMICS & PROTEOMICS 11: 67-80 (2014)

76

auristatin-conjugated antibody to prostate-specific membraneantigen. Mol Cancer Ther 10: 1728-1739, 2011.

38 Akhtar NH, Pail O, Saran A, Tyrell L and Tagawa ST: Prostate-specific membrane antigen-based therapeutics. Adv Urol 973820,2012.

39 Andrady C, Sharma SK and Chester KA: Antibody-enzyme fusionproteins for cancer therapy. Immunotherapy 3: 193-211, 2011.

40 Tietze LF and Schmuck K: Prodrugs for targeted tumor therapies:Recent developments in ADEPT, GDEPT and PMT. Curr PharmDes 17: 3527-3547, 2011.

41 Melton RG and Sherwood RF: Antibody−enzyme conjugates forcancer therapy. J Natl Cancer Inst 88: 153-165, 1996.

42 Bagshawe KD: Antibody-directed enzyme prodrug therapy(ADEPT) for cancer. Expert Rev Anticancer Ther 6: 1421-1431,2006.

43 Springer CJ, Antoniw P, Bagshawe KD, Searle F, Bisset GM andJarman M: Novel prodrugs which are activated to cytotoxicalkylating agents by carboxypeptidase G2. J Med Chem 33: 677-681, 1990.

44 Medzihradszky KF, Spencer DI, Sharma SK, Bhatia J, PedleyRB, Read DA, Begent RH and Chester KA: Glycoforms obtainedby expression in Pichia pastoris improve cancer targetingpotential of a recombinant antibody−enzyme fusion protein.Glycobiology 14: 27-37, 2004.

45 Sharma SK, Pedley RB, Bhatia J, Boxer GM, El-Emir E, QureshiU, Tolner B, Lowe H, Michael NP, Minton N, Begent RH andChester KA: Sustained tumor regression of human colorectalcancer xenografts using a multifunctional mannosylated fusionprotein in antibody-directed enzyme prodrug therapy. ClinCancer Res 11: 814-825, 2005.

46 Mayer A, Francis RJ, Sharma SK, Tolner B, Springer CJ, MartinJ, Boxer GM, Bell J, Green AJ, Hartley JA, Cruickshank C, WrenJ, Chester KA and Begent RH: A phase I study of singleadministration of antibody-directed enzyme prodrug therapy withthe recombinant anti-carcinoembryonic antigen antibody-enzymefusion protein MFECP1 and a bis-iodo phenol mustard prodrug.Clin Cancer Res 12: 6509-6516, 2006.

47 Bagshawe KD and Sharma SK: Cyclosporine delays hostimmune response to antibody−enzyme conjugate in ADEPT.Transplant Proc 28: 3156-3158, 1996.

48 Chester KA, Baker M and Mayer A: Overcoming theimmunologic response to foreign enzymes in cancer therapy.Expert Rev Clin Immunol 1: 549-559, 2005.

49 Denmeade SR and Isaacs JT: Engineering enzymatically activated‘molecular grenades’ for cancer. Oncotarget 3: 666-667, 2012.

50 Christensen SB, Skytte DM, Denmeade SR, Dionne C, MøllerJV, Nissen P and Isaacs JT: A Trojan horse in drug development:targeting of thapsigargins towards prostate cancer cells.Anticancer Agents Med Chem 9: 276-294, 2009.

51 Thastrup O, Cullen PJ, Drøbak BK, Hanley MR and Dawson AP:Thapsigargin, a tumor promoter, discharges intracellular Ca2+stores by specific inhibition of the endoplasmic reticulum Ca2+-ATPase. Proc Natl Acad Sci USA 87: 2466-2470, 1990.

52 Breckenridge DG, Germain M, Mathai JP, Nguyen M and ShoreGC: Regulation of apoptosis by endoplasmic reticulum pathways.Oncogene 22: 8608-8618, 2003.

53 Denmeade SR, Jakobsen CM, Janssen S, Khan SR, Garrett ES,Lilja H, Christensen SB and Isaacs JT: Prostate-specific antigen-activated thapsigargin prodrug as targeted therapy for prostatecancer. J Natl Cancer Inst 95: 990-1000, 2003.

54 Christensen SB, Andersen A, Kromann H, Treiman M, TombalB, Denmeade S and Isaacs JT: Thapsigargin analogues fortargeting programmed death of androgen-independent prostatecancer cells. Bioorg Med Chem 7: 1273-1280, 1999.

55 Denmeade SR, Mhaka AM, Rosen DM, Brennen WN, DalrympleS, Dach I, Olesen C, Gurel B, Demarzo AM, Wilding G,Carducci MA, Dionne CA, Møller JV, Nissen P, Christensen SBand Isaacs JT: Engineering a prostate-specific membrane antigen-activated tumor endothelial cell prodrug for cancer therapy. SciTransl Med 4: 140ra86, 2012.

56 Pinto JT, Suffoletto BP, Berzin TM, Qiao CH, Lin S, Tong WP,May F, Mukherjee B and Heston WD: Prostate-specificmembrane antigen: a novel folate hydrolase in human prostaticcarcinoma cells. Clin Cancer Res 2: 1445-1451, 1996.

57 Carter RE, Feldman AR and Coyle JT: Prostate-specificmembrane antigen is a hydrolase with substrate andpharmacologic characteristics of a neuropeptidase. Proc NatlAcad Sci USA 93: 749-753, 1996.

58 Chang SS, O’Keefe DS, Bacich DJ, Reuter VE, Heston WD andGaudin PB: Prostate-specific membrane antigen is produced intumor-associated neovasculature. Clin Cancer Res 5: 2674-2681,1999.

59 Brennen WN, Isaacs JT and Denmeade SR: Rationale behindtargeting fibroblast activation protein-expressing carcinoma-associated fibroblasts as a novel chemotherapeutic strategy. MolCancer Ther 11: 257-266, 2012.

60 Wolf BB, Quan C, Tran T, Wiesmann C and Sutherlin D: On theedge of validation-cancer protease fibroblast activation protein.Mini Rev Med Chem 8: 719-727, 2008.

61 Brennen WN, Rosen DM, Wang H, Isaacs JT and Denmeade SR:Targeting carcinoma-associated fibroblasts within the tumorstroma with a fibroblast activation protein-activated prodrug. JNatl Cancer Inst 104: 1320-1334, 2012.

62 Olsson AY, Bjartell A, Lilja H and Lundwall A: Expression ofprostate-specific antigen (PSA) and human glandular kallikrein2 (hK2) in ileum and other extraprostatic tissues. Int J Cancer113: 290-297, 2005.

63 Otto A, Bär J and Birkenmeier G: Prostate-specific antigen formscomplexes with human alpha 2-macroglobulin and binds to thealpha 2-macroglobulin receptor/LDL receptor-related protein. JUrol 159: 297-303, 1998.

64 Mikolajczyk SD, Millar LS, Kumar A and Saedi MS: Humanglandular kallikrein, hK2, shows arginine-restricted specificityand forms complexes with plasma protease inhibitors. Prostate34: 44-50, 1998.

65 Eggermont AM and ten Hagen TL: Isolated limb perfusion forextremity soft-tissue sarcomas, in-transit metastases, and otherunresectable tumors: credits, debits, and future perspectives. CurrOncol Rep 3: 359-367, 2001.

66 Wajant H, Gerspach J and Pfizenmaier K: Tumor therapeutics bydesign: Targeting and activation of death receptors. CytokineGrowth Factor Rev 16: 55-76, 2005.

67 Kelly T: Fibroblast activation protein-alpha and dipeptidylpeptidase IV (CD26): cell-surface proteases that activate cellsignaling and are potential targets for cancer therapy. Drug ResistUpdat 8: 51-58, 2005.

68 Wüest T, Gerlach E, Banerjee D, Gerspach J, Moosmayer D andPfizenmaier K: TNF-Selectokine: A novel prodrug generated fortumor targeting and site-specific activation of tumor necrosisfactor. Oncogene 21: 4257-4265, 2002.

Weidle et al: Proteases as Mediators of Cytotoxicity in Tumor Therapy (Review)

77

69 Gerspach J, Müller D, Münkel S, Selchow O, Nemeth J, NoackM, Petrul H, Menrad A, Wajant H and Pfizenmaier K:Restoration of membrane TNF-like activity by cell surfacetargeting and matrix metalloproteinase-mediated processing of aTNF prodrug. Cell Death Differ 13: 273-284, 2006.

70 Gerspach J, Németh J, Münkel S, Wajant H and Pfizenmaier K:Target-selective activation of a TNF prodrug by urokinase-typeplasminogen activator (uPA) mediated proteolytic processing atthe cell surface. Cancer Immunol Immunother 55: 1590-1600,2006.

71 Watermann I, Gerspach J, Lehne M, Seufert J, Schneider B,Pfizenmaier K and Wajant H: Activation of CD95L fusion proteinprodrugs by tumor-associated proteases. Cell Death Differ 14:765-774, 2007.

72 Hristova K, Dempsey CE and White SH: Structure, location, andlipid perturbations of melittin at the membrane interface. BiophysJ 80: 801-811, 2001.

73 Holle L, Song W, Holle E, Wei Y, Wagner T and Yu X: A matrixmetalloproteinase 2 cleavable melittin/avidin conjugatespecifically targets tumor cells in vitro and in vivo. Int J Oncol22: 93-98, 2003.

74 Bateman A, Bullough F, Murphy S, Emiliusen L, Lavillette D,Cosset FL, Cattaneo R, Russell SJ and Vile RG: Fusogenicmembrane glycoproteins as a novel class of genes for the localand immune-mediated control of tumor growth. Cancer Res 60:1492-1497, 2000.

75 Johnson KJ, Peng KW, Allen C, Russell SJ and Galanis E:Targeting the cytotoxicity of fusogenic membrane glycoproteinsin gliomas through protease-substrate interaction. Gene Ther 10:725-732, 2003.

76 Allen C, McDonald C, Giannini C, Peng KW, Rosales G, RussellSJ and Galanis E: Adenoviral vectors expressing fusogenicmembrane glycoproteins activated via matrix metalloproteinasecleavable linkers have significant antitumor potential in the genetherapy of gliomas. J Gene Med 6: 1216-1227, 2004.

77 Liu S, Schubert RL, Bugge TH and Leppla SH: Anthrax toxin:structures, functions and tumour targeting. Expert Opin Biol Ther3: 843-853, 2003.

78 Abrami L, Liu S, Cosson P, Leppla SH and van der Goot FG:Anthrax toxin triggers endocytosis of its receptor via a lipid raft-mediated clathrin-dependent process. J Cell Biol 160: 321-328,2003.

79 Cryan LM and Rogers MS. Targeting the anthrax receptors,TEM-8 and CMG-2, for anti-angiogenic therapy. Front Biosci(Landmark Ed) 16: 1574-1588, 2011.

80 Shapira A and Benhar I: Toxin-based therapeutic approaches.Toxins 2: 2519-2583, 2010.

81 Duesbery NS, Webb CP, Leppla SH, Gordon VM, Klimpel KR,Copeland TD, Ahn NG, Oskarsson MK, Fukasawa K, PaullKD and Vande Woude GF. Proteolytic inactivation of MAP-kinase-kinase by anthrax lethal factor. Science 280: 734-737,1998.

82 Koo HM, VanBrocklin M, McWilliams MJ, Leppla SH,Duesbery NS and Vande Woude GF. Apoptosis andmelanogenesis in human melanoma cells induced by anthraxlethal factor inactivation of mitogen-activated protein kinasekinase. Proc Natl Acad Sci USA 99: 3052-3057, 2002.

83 Liu S, Netzel-Arnett S, Birkedal-Hansen H and Leppla SH:Tumor cell-selective cytotoxicity of matrix metalloproteinase-activated anthrax toxin. Cancer Res 60: 6061-6067, 2000.

84 Arora N, Klimpel KR, Singh Y and Leppla SH: Fusions ofanthrax toxin lethal factor to the ADP-ribosylation domain ofPseudomonas exotoxin A are potent cytotoxins which aretranslocated to the cytosol of mammalian cells. J Biol Chem267: 15542-15548, 1992.

85 Milne JC, Blanke SR, Hanna PC and Collier RJ: Protectiveantigen-binding domain of anthrax lethal factor mediatestranslocation of a heterologous protein fused to its amino- orcarboxy-terminus. Mol Microbiol 15: 661-666, 1995.

86 Arora N and Leppla SH: Fusions of anthrax toxin lethalfactor with shiga toxin and diphtheria toxin enzymaticdomains are toxic to mammalian cells. Infect Immun 62:4955-4961, 1994.

87 Liu S, Bugge TH and Leppla SH: Targeting of tumor cells bycell surface urokinase plasminogen activator-dependent anthraxtoxin. J Biol Chem 276: 17976-17984, 2001.

88 Weidle UH, Georges G and Brinkmann U: Fully humantargeted cytotoxic fusion proteins: New anticancer agents on thehorizon. Cancer Genomics Proteomics 9: 119-133, 2012.

89 Hehmann-Titt G, Schiffer S, Berges N, Melmer G and Barth S:Improving the therapeutic potential of human granzyme B fortargeted cancer therapy. Antibodies 2: 19-49, 2013.

90 Oberoi P, Jabulowsky RA and Wels WS: Selective induction ofcancer cell death by targeted granzyme B. Antibodies 2: 130-151, 2013.

91 Kurschus FC and Jenne DE: Delivery and therapeutic potentialof human granzyme B. Immunol Rev 235: 159-171, 2010.

92 Rosenblum MG and Barth S: Development of novel, highlycytotoxic fusion constructs containing granzyme B: Uniquemechanisms and functions. Curr Pharm Des 15: 2676-2692,2009.

93 Losasso V, Schiffer S, Barth S and Carloni P: Design of humangranzyme B variants resistant to serpin B9. Proteins 80: 2514-2522, 2012.

94 Bird CH, Sun J, Ung K, Karambalis D, Whisstock JC, TrapaniJA and Bird PI: Cationic sites on granzyme B contribute tocytotoxicity by promoting its uptake into target cells. Mol CellBiol 25: 7854-7867, 2005.

95 Jabulowsky RA, Oberoi P, Bähr-Mahmud H, Dälken B andWels WS: Surface charge-modification prevents sequestrationand enhances tumor-cell specificity of a recombinant granzymeB-TGFα fusion protein. Bioconjug Chem 23: 1567-1576, 2012.

96 Kurschus FC, Fellows E, Stegmann E and Jenne DE: GranzymeB delivery via perforin is restricted by size, but not by heparansulfate-dependent endocytosis. Proc Ntl Acad Sci USA 105:13799-13804, 2008

97 Kurschus FC, Kleinschmidt M, Fellows E, Dornmair K,Rudolph R, Lilie H and Jenne DE: Killing of target cells byredirected granzyme B in the absence of perforin. FEBS Lett562: 87-92, 2004.

98 Keller J, Heisler I, Tauber R and Fuchs H: Development of anovel molecular adapter for the optimization of immunotoxins.J Control Release 74: 259-261, 2001.

99 Cao Y, Mohamedali KA, Marks JW, Cheung LH, HittelmanWN and Rosenblum MG: Construction and characterization ofnovel, completely human serine protease therapeutics targetingHER2/NEU. Mol Cancer Ther 12: 979-991, 2013.

100 Hetzel C, Bachran C, Tur MK, Fuchs H and Stöcker M:Improved immunotoxins with novel functional elements. CurrPharm Des 15: 2700-2711, 2009.

CANCER GENOMICS & PROTEOMICS 11: 67-80 (2014)

78

101 Heisler I, Keller J, Tauber R, Sutherland M and Fuchs H: Acleavable adapter to reduce nonspecific cytotoxicity ofrecombinant immunotoxins. Int J Cancer 103: 277-282, 2003.

102 Säälik P, Elmquist A, Hansen M, Padari K, Saar K, Viht K,Langel U and Pooga M: Protein cargo delivery properties of cell-penetrating peptides. A comparative study. Bioconjug Chem 15:1246-1253, 2004.

103 Jones AT: Gateways and tools for drug delivery: endocyticpathways and the cellular dynamics of cell penetrating peptides.Int J Pharm 354: 34-38, 2008.

104 Fuchs H, Bachran C, Heisler I and Sutherland M: A closer lookat protein transduction domains as a tool in drug delivery. CurrNanosci 1: 117-124, 2005.

105 Hudson TH and Neville DM Jr: Enhancement of immunotoxinaction: manipulation of the cellular routing of proteins. CancerTreat Res 37: 371-389, 1988.

106 Fuchs H, Bachran C, Li T, Heisler I, Dürkop H and SutherlandM: A cleavable molecular adapter reduces side effects andconcomitantly enhances efficacy in tumor treatment by targetedtoxins in mice. J Control Release 117: 342-350, 2007.

107 Hetzel C, Bachran C, Fischer R, Fuchs H, Barth S and StöckerM: Small cleavable adapters enhance the specific cytotoxicity ofa humanized immunotoxin directed against CD64-positive cells.J Immunother 31: 370-376, 2008.

108 Hansel TT, Kropshofer H, Singer T, Mitchell JA and George AJ:The safety and side effects of monoclonal antibodies. Nat RevDrug Discov 9: 325-338, 2010.

109 Yan M, Callahan CA, Beyer JC, Allamneni KP, Zhang G,Ridgway JB, Niessen K and Plowman GD: Chronic DLL4blockade induces vascular neoplasms. Nature 463: E6-7, 2010.

110 Wu Y, Cain-Hom C, Choy L, Hagenbeek TJ, de Leon GP, ChenY, Finkle D, Venook R, Wu X, Ridgway J, Schahin-Reed D, DowGJ, Shelton A, Stawicki S, Watts RJ, Zhang J, Choy R, HowardP, Kadyk L, Yan M, Zha J, Callahan CA, Hymowitz SG andSiebel CW: Therapeutic antibody targeting of individual Notchreceptors. Nature 464: 1052-1057, 2010.

111 Li JL, Jubb AM and Harris AL: Targeting DLL4 in tumors showspreclinical activity but potentially significant toxicity. FutureOncol 6: 1099-1103, 2010.

112 Kamba T and McDonald DM: Mechanisms of adverse effects ofanti-VEGF therapy for cancer. Br J Cancer 96: 1788-1795, 2007.

113 Segaert S and Van Cutsem E: Clinical signs, pathophysiology andmanagement of skin toxicity during therapy with epidermal growthfactor receptor inhibitors. Ann Oncol 16: 1425-1433, 2005.

114 Tijink BM, Buter J, de Bree R, Giaccone G, Lang MS, Staab A,Leemans CR and van Dongen GA: A phase I dose escalationstudy with anti-CD44v6 bivatuzumab mertansine in patients withincurable squamous cell carcinoma of the head and neck oresophagus. Clin Cancer Res 12: 6064-6072, 2006.

115 Bongartz T, Sutton AJ, Sweeting MJ, Buchan I, Matteson EL andMontori V: Anti-TNF antibody therapy in rheumatoid arthritisand the risk of serious infections and malignancies: systematicreview and meta-analysis of rare harmful effects in randomizedcontrolled trials. JAMA 295: 2275-2285, 2006.

116 Perez EA: Cardiac toxicity of ErbB2-targeted therapies: What dowe know? Clin Breast Cancer 8(Suppl 3): S114-120, 2008.

117 Erster O, Thomas JM, Hamzah J, Jabaiah AM, Getz JA, SchoepTD, Hall SS, Ruoslahti E and Daugherty PS: Site-specifictargeting of antibody activity in vivo mediated by disease-associated proteases. J Control Release 161: 804-812, 2012.

118 LeBeau AM, Lee M, Murphy ST, Hann BC, Warren RS, DelosSantos R, Kurhanewicz J, Hanash SM, VanBrocklin HF andCraik CS: Imaging a functional tumorigenic biomarker in thetransformed epithelium. Proc Natl Acad Sci USA 110: 93-98,2013.

119 Desnoyers LR, Vasiljeva O, Richardson JH, Yang A, MenendezEEM, Liang TW, Wong C, Bessette PH, Kamath K, Moore SJ,Sagert JG, Hofstetter DR, Han F, Gee J, Flandez J, MarkhamK, Nguyen M, Krimm M, Wong KR, Liu S, Daugherty, WestJW and Lowman HB: Tumor-specific activation of EGFR-targeting probody enhances therapeutic index. Sci Transl Med5: 207ra144, 2013.

120 Metz S, Panke C, Haas AK, Schanzer J, Lau W, Croasdale R,Hoffmann E, Schneider B, Auer J, Gassner C, Bossenmaier B,Umana P, Sustmann C and Brinkmann U: Bispecific antibodyderivatives with restricted binding functionalities that areactivated by proteolytic processing. Protein Eng Des Sel 25: 571-580, 2012.

121 Parr C, Sanders AJ and Jiang WG: Hepatocyte growth factoractivation inhibitors − therapeutic potential in cancer. AnticancerAgents Med Chem 10: 47-57, 2010.

122 Saleem M, Adhami VM, Zhong W, Longley BJ, Lin CY, DicksonRB, Reagan-Shaw S, Jarrard DF and Mukhtar H: A novelbiomarker for staging human prostate adenocarcinoma:overexpression of matriptase with concomitant loss of itsinhibitor, hepatocyte growth factor activator inhibitor-1. CancerEpidemiol Biomarkers Prev 15: 217-227, 2006.

123 Sun J, Pons J and Craik CS: Potent and selective inhibition ofmembrane-type serine protease 1 by human single-chainantibodies. Biochemistry 42: 892-900, 2003.

124 Farady CJ, Egea PF, Schneider EL, Darragh MR and Craik CS:Structure of a Fab-protease complex reveals a highly specificnon-canonical mechanism of inhibition. J Mol Biol 380: 351-360, 2008.

125 Hayden, C.E: News. Personalized cancer therapy gets closer.Genetic testing allows doctors to select best treatment. Nature458: 131-132, 2009.

126 Simon R and Roychowdhury S: Implementing personalizedcancer genomics in clinical trials. Nat Rev Drug Discov 12: 358-369, 2013.

127 Martini M, Vecchione L, Siena S, Tejpar S and Bardelli A:Targeted therapies: how personal should we go? Nat Rev ClinOncol 9: 87-97, 2011.

128 Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN,Weissig H, Shindyalov IN and Bourne PE: The Protein DataBank (www.pdb.org). Nucleic Acids Res 28: 235-242, 2000.

129 Accelrys Software Incorp., Discovery Studio ModelingEnvironment, Release 3.1, San Diego: Accelrys Software Inc.,2007.

Received February 21, 2014Revised March 6, 2014Accepted March 7, 2014

Weidle et al: Proteases as Mediators of Cytotoxicity in Tumor Therapy (Review)

79