IntroductionHeparanaseis a member of the glycoside hydrolase 79 (GH79) family of carbohydrateprocessing enzymes (1). It is further classified as an endo-?-glucuronidase andis primarily involved with the breakdown of heparan sulfate, aglycosaminoglycan polysaccharide (1). Heparan sulfate has linear disacchariderepeats with internal ?1-4 glycosidic linkages (1). Hydrolytic cleavage of these linkages iscatalyzed by heparanase (1). Heparan sulfate is a major component of theextracellular matrix, where it usually occurs in proteoglycan form (2). Theheparanase-heparan sulfate axis is involved with structural regulation of theextracellular matrix, and is heavily implicated in angiogenesis and metastaticprogression for several types of cancer (2).
The development of heparanaseinhibitors is an important part of current therapeutic strategies (3, 4). Oneof the more potent inhibitors is Roneparstat, a heparin derivative that isundergoing clinical trials for multiple myeloma (3). Recently the crystalstructures for both human bacterial heparanases have been reported andcharacterized (1, 5).Glycoside hydrolases – Overview andclassificationHydrolysisof the glycosidic bond between two carbohydrate residues is catalyzed byenzymes called glycoside hydrolases (6). These enzymes also catalyze hydrolyticcleavage of the bond between a carbohydrate residue and a non-carbohydrateaglycon unit (6). They are widespread and occur extensively in both prokaryoticand eukaryotic organisms (6).
Glycoside hydrolases are known to account forroughly one percent of the genomic makeup of most organisms and are involved inimportant biological processes (6). Carbohydrate chains or residues can beattached covalently to lipids and proteins, where they mediate cellularrecognition processes by interacting with other molecules. Glycoside hydrolasesare heavily involved in the processing of these carbohydrates through theaddition or removal of individual carbohydrate units (6). They are alsoinvolved in the metabolic processing of cellulose and starch into smaller carbohydrateunits, which is important for energy production and biopolymer construction inorganisms (6). TheInternational Union of Biochemistry and Molecular Biology uses an enzymeclassification system based on substrate specificity, but this system does notprovide much information on the structural relationships between specificenzymes (7). In general, the substrate specificity of an enzyme is less closelyrelated to its catalytic mechanism than the tertiary structure of the enzyme.
Themost commonly used classification is based on amino acid sequence comparisonsand is referred to as the carbohydrate-active enzyme (CAZy) classificationsystem (7, 8). The tertiary structure of an enzyme is determined by its primaryamino acid sequence. Therefore, glycoside hydrolases with similar primarysequences also share similar tertiary structures, active sites and catalyticmechanisms (7). The CAZy system dividesglycoside hydrolases into 145 families, with each family containing enzymes withsimilar tertiary structures and mechanisms, but different specificities (7).This classification system also provides information on the evolutionaryrelationships between glycoside hydrolase families. Families with similartertiary structures may be grouped into clans, with 15 clans known at present(7). Glycosidehydrolases may also be classified based on exo or endo activity(1).
Enzymes with exo activity cleavea substrate at the end of a chain while those with endo activity cleave a substrate within the middle of a chain (1). Manyglycoside hydrolases have long substrate binding sites that can accommodatemultiple carbohydrate residues (9). The binding location for each residue isreferred to as a subsite (9). Subsite numbering is based on the position of thetwo sugar residues that share the glycosidic oxygen atom (9). These arenumbered –1 and +1 in order to identify the non-reducing and reducing ends ofthe chain respectively. Subsite numbering increases (+1, +2, +3) towards thereducing end and decreases (–1, –2, –3) towards the non-reducing end (9). Theanomeric carbon atom involved in the hydrolysis reaction is contained by thesugar residue present at the -1 subsite position (9).
Catalytic mechanisms of glycosidehydrolasesMostglycoside hydrolases use one of two generic catalytic mechanisms, based on thestereochemistry of the hydrolysis reaction (10). Retaining mechanisms involvenet retention of the product’s anomeric configuration, while invertingmechanisms involve net inversion of the anomeric configuration (10). Both mechanismsuse two carboxylate groups which are provided by glutamate or aspartate residuesat the active site (10). One group occupies a position laterally above theplane of the sugar residue, while the other one is located below the anomericcarbon atom (11). These amino acid residues play slightly different rolesdepending on the mechanism used by the enzyme.
Glycoside hydrolases with aretaining mechanism perform the reaction in two steps (10) (Figure 1). In thefirst step, one of the carboxylate groups functions as a nucleophile and formsa covalent bond with the anomeric carbon atom of the carbohydrate unit at theactive site (12). At the same time, the second carboxylate group functions asan acid catalyst and donates a proton to the glycosidic oxygen atom of theleaving group (12). This results in the formation of a glycosyl-enzymeintermediate (12). The second step involves the cleavage of the glycosyl-enzymeintermediate. The carboxylate group that acted as proton donor in the firststep now functions as a base catalyst and assists the nucleophilic attack of a watermolecule in order to form the product (12). Theinverting mechanism only requires one step (10) (Figure 2). One carboxylategroup donates a proton to the glycosidic oxygen atom of the leaving group, whilethe second carboxylate group facilitates the nucleophilic attack of a watermolecule on the anomeric carbon atom (10).
This carboxylate group collects theproton and completes the inversion reaction (10). Both retaining and inversionmechanisms involve an oxocarbenium ion-like transition state (13).Non-covalentinteractions also play an important role in the activity of glycosidehydrolases (14). The amino acid residues at the active site can form hydrogenbonds with the hydroxyl groups on carbohydrate substrates (14). Most glycosidehydrolases have a hydrogen bond acceptor that interacts with the 2-hydroxyl(C2-OH) group (15). The proximity of this hydroxyl group to the anomeric centerenables it to stabilize the transition state of the enzymatic reaction (15).Amino acids with hydrophobic side chains may also stabilize the transitionstate through stacking interactions at the –1 position (16).
Localization and biogenesis of humanheparanaseHumanheparanase is localized to lysosomes and late endosomes, where it is involvedin the breakdown of internalized heparin sulfate proteoglycans (17). It is alsotransported to the cell surface and extracellular matrix where basement membraneheparan sulfate glycosaminoglycans are degraded at sites of injury orinflammation. (17).
Heparanase is translated as a pre-proenzyme containing a 35residue signal sequence (Met1-Ala35) (18) (Figure 3). Signal peptidase cleavesthis sequence to produce an inactive 65 kDa proenzyme, which is then processedby cathepsin L to proteolytically remove a linker region (Ser110-Gln157) (18).This produces an 8 kDa N-terminal subunit (Gln36-Glu109) and a 50 kDaC-terminal subunit (Lys159-Ile543), which function as a noncovalent heterodimerin the enzyme’s active form (1, 18). The proenzyme is synthesized in the Golgicomplex and then transferred to late endosomes and lysosomes for proteolyticprocessing into its active form (18).Secondary structure and catalyticsite-human heparanaseHumanheparanase is known to belong to the GH79 family (Clan A) based on amino acid sequenceanalysis, secondary structure and mutagenic studies (1). All Clan A glycosidehydrolases have a conserved protein fold that consists of eight alternating alphahelices and parallel beta strands in a single domain (19). This is called a TIMbarrel and is named after the enzyme triosephosphate isomerase (19). Thecrystal structure of human apo-heparanase was reported in 2015 and theheterodimer with 2 subunits was observed at a resolution of 1.
75 Å (1) (Figure4). The secondary structure consists of a large (?-?)8 domain (TIM barrel) anda small ?-sandwich domain, with the small subunit contributing one ?-sheet tothe sandwich domain and the first ??? fold of the (?-?)8 domain (1). The largesubunit contributes the remaining folds (1). Thecatalytic site is located within a 10 Å cleft in the (?-?)8 domain, whichcontains the catalytic acid-base and nucleophile (Glu225 and Glu343) (1, 20)(Figure 5).
The catalytic cleft is lined with many basic residues (Arg35,Lys158/159/161/231, Arg272/273/303) that interact with the negative charge onheparan sulfate (1). The position of the 8-kDa and 50-kDa subunits in thecrystal structure indicates that the excised linker region of the proenzymelies within the active site cleft in the (?-?)8 domain (1). The linker regionthus blocks substrate binding prior to enzyme activation (1). Recently, 1H NMRspectroscopic studies have confirmed that human heparanase has a retainingcatalytic mechanism and hydrolyzes its substrate with a net retention ofconfiguration (21).
Substrate-human heparanaseHeparansulfate is the primary substrate for human heparanase (1). The disacchariderepeats are composed of glucuronic or iduronic acid with a ?1-4 glycosidiclinkage to either N-acetylglucosamine or N-sulfoglucosamine (1). Heparansulfate chains are generally modified by O-sulfation at O2 ofglucuronic/iduronic acid and O6 of N-acetylglucosamine /N-sulfoglucosamine (1).
This results in the presence of distinct domains with different degrees ofO-sulfation, which allows different binding partners to interact with heparansulfate (1). Heparanase targets heparan sulfate chains with specific sulfationpatterns (1). Experimental studies have shown that it preferentially hydrolyzesa trisaccharide unit with sulfated N-acetylglucosamine/N-sulfoglucosamineresidues at the –2 and +1 positions (22). The -2 N-sulfate and +1 O-sulfate are connected to the enzyme throughhydrogen bonding, which allows it to open up the substrate and access the -1sugar residue (1). The –2 6O-sulfate and +1 N-sulfate are also known to stabilizethe bound trisaccharide substrate through electrostatic interactions to basic aminoacid residues in the active site (1).Twomodes of substrate cleavage have been identified for heparanase – consecutive cleavageand gapped cleavage (23) (Figure 6). Consecutive cleavage occurs when anN-acetylglucosamine residue is present in the trisaccharide unit at thenon-reducing end of the sugar (23).
The enzyme removes the trisaccharide unitand then immediately removes 2 more residues from the non-reducing end (23).Gapped cleavage occurs when an N-sulfoglucosamine is present in thetrisaccharide unit, causing the enzyme to skip the next cleavage site afterremoving the trisaccharide (23).Bacterial heparanaseThecrystal structures of a bacterial heparanse (BpHPSE) from Burkholderiapseudomallei and a bacterial exo-glucuronidase (AcaGH79) from Acidobacterium capsulatum have alsobeen reported recently (5, 24). The bacterial heparanase is knownto have similar activity to human heparanase but shows different substratespecificity (5). Human heparanase prefers heparan sulfate withN-sulfoglucosamine residues while the bacterial enzyme prefers N-acetylglucosamineresidues (5). The bacterial enzyme is also known to act on chondroitin sulfate,which is not known to be a substrate for human heparanase (5). Sequencealignment studies of eukaryotic heparanases with BpHPSE and AcaGH79have revealed that amino acid residues at the –1 subsite of human heparanase arewell conserved across all species, while residues at the –2 and +1 subsites arepoorly conserved in the bacterial enzymes (1, 5).
This accounts for thedifferences in substrate specificity between human and bacterial heparanases(1, 5). Theevolutionary relationship between GH79 enzymes with exo and endo activityhas also been studied with reference to the loop that forms part of the exosubstrate binding pocket in AcaGH79 (1). This loop is much smaller in endo-actingBpHPSE, which converts the binding pocket into a cleft capable ofaccommodating long substrate chains (1). Human heparanase seems to havedeveloped endo specificity via expansion of the loop into a linkerregion that is excised to produce the active enzyme (1) (Figure 7).Heparanase and cancerHeparanaseis heavily involved in the metastatic progression of malignant tumors (25). Itmediates the breakdown of heparan sulfate in the extracellular matrix (25).Studies have shown that a direct correlation exists between the levels ofheparanase expression and the rates of invasion and metastatic progression (26).
This is also true for tumor vascularity levels which suggests that heparanaseplays an important role in tumor angiogenesis (26). Postoperative survival time is known to bemuch shorter for cancer patients with high levels of heparanase expression whencompared to patients with low levels of heparanase expression (27). Heparanaseis known to be involved with the mobilization of tumor-supporting immunocytepopulations, such as tumor-associated macrophages and neutrophils (28). Theenzyme is thought to create a tumor promoting microenvironment characterized byincreased NF-kB and STAT3 signaling, high cyclooxygenase-2 levels and activationof macrophages that provide cytokines such as TNF-a and interleukin-1 (28). Anovel mechanism linking heparanase to angiogenesis and metastasis has recentlybeen described (2) (Figure 8).
Upregulation of heparanase causes the cleavageof heparan sulfate chains linked to syndecan-1, a cell surface proteinexpressed by myeloma and endothelial cells (2). The core syndecan-1 protein isthen cleaved by proteases like MMP-9, which results in shedding of syndecan-1from the cell surface (2). A juxtamembrane domain of the core syndecan-1 isthen exposed, which facilitates the coupling of VEGFR-2 (vascular endothelialgrowth factor receptor– 2) and VLA-4 (Very late antigen-4) at the cell surface(2). This results in a signaling cascade that promotes cell invasion andendothelial tube formation (2).
Roneparstatis a heparin derivative that decreases syndecan-1 shedding, while synstatinpeptides mimic the syndecan-1 core protein and inhibit the binding of VLA-4 andVEGFR-2 to shed syndecan-1 (2). Roneparstat has proven to be effective againstmyelomas and brain metastases of breast cancer, in combination withchemotherapeutic agents like bortezomib and lapatinib (3, 29, 30). It iscurrently in Phase 2 clinical trials in myeloma patients (31). In vivo studies have shown that Roneparstatcauses a decrease in the release of ECM bound FGF-2 (fibroblast growthfactor-2) and a lack of anticoagulant activity (31).
Otheroligosaccharide mimetics have also progressed to clinical trials (32). PI-88 isa highly sulfated phosphosulfomannan which is currently in Phase 3 trials forhepatocellular carcinoma (32). It is a complex mixture of sulfatedoligosaccharides, with the major components being pentasaccharides andtetrasaccharides (32).
In addition to its function as a heparanase inhibitor, PI-88also inhibits angiogenesis directly by preventing the interactions of VEGF andits receptors with heparan sulfate (32). PI-88 is linked to decreased tumor cellproliferation, upregulation of apoptosis and a notable reduction in the numberof invasive carcinomas (32). Ongoing clinical trials have also demonstratedthat PI-88 has good toxicity and tolerability profiles, thus increasing thelikelihood that it will be approved by the FDA (32).Avery recent study has also shown that heparanase is a target of the drugaspirin (33). Aspirin binds directly to the Glu225 residue in the heparanaseactive site, thus inhibiting enzymatic activity (33). These studies have shownthat aspirin reduces tumor metastasis and angiogenesis both in vitro and invivo (33). Heparanase activity assays Severalstudies have highlighted the role of heparanase as a major player in tumorgrowth, angiogenesis and metastatic progression. Based on the increasingimportance of heparanase as a drug target, two commercial heparanase activityassays have been developed recently (34).
The first method is based on thereaction between chimeric recombinant heparanase and the commercial substratefondaparinux (34). The enzyme and substrate are incubated overnight, afterwhich a fluorescent redox marker (resazurin) is added (34). Reduction of thismarker is dependent on the amount of glucuronic acid produced through heparanaseactivity (34). Excitation and emission wavelengths of 560 nm and 590 nm can beused for the fluorescence measurements (34).Thesecond method involves the incubation of a 96-well plate with protaminesulfate, which is followed by the immobilization of biotinylated heparansulfate (34). Different concentrations of chimeric recombinant heparanase areused for this method (34). The immobilized substrate not digested by heparanaseis then bound to streptavidin conjugated with europium and fluorescence can bemeasured with a time-resolved fluorometer (34).
The fondaparinux assay can beused to screen heparanase inhibitors while the biotinylated heparan sulfateassay can be used for quantitative analysis of heparanase in biological samples(34).Areas of future researchRecentstudies have shown that heparanase trafficking is implicated in the enhancedautophagy exhibited by several tumor cell types (2). When heparanase issecreted, it interacts very quickly with syndecans and other heparan sulfateproteoglycans associated with the cell membrane (2). This is followed by rapidendocytosis of the enzyme-substrate complex (2). These endosomes are then converted tolysosomes, which results in the processing and activation of heparanase (2).
Lysosomal heparanase regulates the basal level of autophagy and is storedwithin autophagosomes (2). Tumor cells with higher heparanase content exhibithigher levels of autophagy, which promotes chemoresistance and tumor cellproliferation (2). Thesecells are characterized by a decrease in p70 S6-kinase phosphorylation levelsand accumulation of mTOR1 near the nucleus (2). At present, there are noinhibitors that specifically target this process. Therefore, there is greatpotential for the development of a novel class of inhibitors that targetheparanase trafficking and the lysosomal accumulation of heparanase. Inaddition to the role that heparanase plays in cancer, recent studies have alsoindicated that the enzyme may be involved in the development of multiplesclerosis (35). Disruption of the extracellular matrix, inflammation of thecentral nervous system and T-cell activation are key features of multiplesclerosis (35). As heparanase influences and mediates similar processes incancer, it has been hypothesized that it may function similarly in multiplesclerosis (35).
Recent experimental results from animal studies suggest thatthat there may be a direct correlation between increased heparanase expressionand severe disease symptoms (35). Polymorphisms in glypican 5 (a cell surfaceproteoglycan) are also known to be a risk factor for the disease, thus supportinga potential role for heparanase (35). Therefore, there is a need to investigateand develop heparanase inhibitors that specifically target multiple sclerosis.Heparanaseis also known to mediate the onset of renal damage and proteinuria during diabeticnephropathy (36). Glomerular heparanase expression is known to beelevated in other similar diseases (36). Recent studies have shown that heparanaseplays a key role in proteinuria and renal damage by causing a reduction inglomerular heparan sulfate expression and enhancing renal macrophage and leukocyteinflux (36).
Heparanase inhibitors could potentially reduce the levels of renaldamage and proteinuria in patients with diabetic nephropathy.Therole of heparanase in the pathogenesis of HSV-1 (Herpes simplex virus-1) hasalso been investigated recently (37). Viral upregulation of heparanase at thenucleus causes a reduction in interferon signaling and an increase in theactivation of NF-kB (37). Inhibition of type I interferon signaling increasesthe susceptibility of neighboring cells to HSV-1 infection, while an increasein NF-kB activation results in increased recruitment of endothelial and immunecells at the site of infection (37). These conditions favor inflammation andincreased cytokine levels which leads to defects in wound-healing (37). Inhibition of heparanase upregulation maypotentially reduce virulence and symptoms of the disease (37). Heparanase hasthus been characterized as host-encoded pathogenic factor (37).
Several heparanaseinhibitors are currently in clinical trials for multiple cancer types. As aresult, a lot of pre-clinical and clinical information is readily available forseveral classes of heparanase inhibitors. These inhibitors could potentially beadapted to target heparanase in HSV-1 and other similar infectious diseases. Computationaltechniques are being used to develop heparanase inhibitors, by extractingnatural heparanase-heparan sulfate interactions as a design template forheparan sulfate mimicking glycopolymers (38). Multiple compounds have beenevaluated in this manner.
A glycopolymer with 12 repeating units has shown themost potent inhibitory activity to date and is known to bind strongly toheparanase (38). Experiments have also demonstrated that this compound lacksanticoagulant activity (38). This compound is just one of many promisingheparanase inhibitors that are currently being developed.ConclusionInconclusion, heparanase plays a very important role in the development ofseveral conditions including various types of cancer, multiple sclerosis, viralinfections and renal diseases. The role that heparanase plays in cancer progressionis currently in the spotlight and the vast majority of drug developmentresearch is focused on cancer. However, there is a lot of scope and potentialfor the development of effective heparanase inhibitors that specifically targetdiseases like multiple sclerosis or diabetic nephropathy.
Until recently, drug development was hinderedby the lack of crystal structures for human and bacterial heparanases. Now thatcrystal structures are available, effective heparanase inhibitors can be designedmore efficiently and with greater ease.