Proprotein convertase inhibition: Paralyzing the cell’s master switches

Abstract

Proprotein convertases are serine proteases responsible for the cleavage and subsequent activation of protein substrates, many of them relevant for the development of an ample variety of diseases. Seven of the PCs, including furin and PACE4, recognize and hydrolyze the C-terminal end of the general sequence RXRR/KXR, whereas PCSK-9 recognizes a series of non-basic amino acids. In some systems, PC-mediated substrate activation results in the development of pathological processes, such as cancer, endocrinopathies, and cardiovascular and infectious diseases.

After establishing PCs as relevant contributors to disease processes, research efforts were directed towards the development of inhibition strategies, including small and large molecules, anti-sense therapies, and antibody-based therapies. Most of these inhibitors mimic the consensus sequence of PCs, blocking the active site in a competitive manner. The most promising inhibitors were designed as bioengineered proteins; however, some non-protein and peptidomimetic agents has also proved to be effective.

These efforts led to the design of pre-clinical studies and clinical trials utilizing inhibitors to PCs. Although the initial studies were performed using non-selective PCs inhibitors, such as CMK, the search for more specific, and compartmentalized selective inhibitors resulted in specific activities ascribed to some, but not all of the PCs. For instance, PACE4 inhibitors were effective in decreasing prostate cancer cell proliferation, and neovascularization. Decreased metastatic ovarian cancer utilizing furin inhibitors represents one of the major endeavors, currently in a phase II trial stage. Antibodies targeting PCSK-9 decreased significantly the levels of HDL-cholesterol, in a phase III trial.

The study of Proprotein convertases has reached a stage of maturity. New strategies based on the alteration of their activity at the cellular and clinical level represent a promising experimental pharmacology field. The development of allosteric inhibitors, or specific agents directed against individual PCs is one of the challenges to be unraveled in the future.

Graphical abstract

1. Introduction

Proprotein convertases subtilisin and kexin type (PCSK) constitute a group of nine Calcium-dependent serine proteases; PC1/2. Furin, PC4, PC5/6, PC7, and PACE4 [1], SKI-1[2], and PCSK9 [3]. All, but the last two PCs listed, recognize and cleave at the C-terminal end of the basic sequence RXR/KR. Several inactive protein precursors gain full activity after a specific PC-mediated proteolysis. These protein substrates include growth factors and growth factor receptors, metalloproteinases, clotting factors, viral proteins, and others. Increased levels of active proteins lead to the promotion of pathophysiological processes as diverse as cell proliferation, degradation of extracellular matrix, activation of the immune response, and activation of viral proteins that allow for efficient entry into the host cells, to mention just a few [4–6]. Table 1 summarizes the main characteristics, subcellular localization, and main pathophysiological processes associated to each of the PCs. The activation of these proteins may result in the development of disease, pointing to PCs as targets for therapeutic interventions. Several strategies aiming at inhibiting the activity of PCs, including the design of competitive inhibitors and interference with mRNA transcription, have been extensively investigated to abolish or minimize the proteolytic activation of these substrates [4, 7–9].

Table 1.

Proprotein convertases. The table depicts the main characteristics of the PCs, including the recognition and cleavage sequence, localization, and cognate pathophysiological processes.

PC	Cleavage sequence	Subcellular localization	Tissue distribution	Proteins activated after PC cleavage	Pathophysiological processes	Ref.
PCSK-1	(K/R)-(X)-(K/R)	Vesicles in neuroendocrine tissue	Neuroendocrine tissue, brain Pancreas, heart	Proenkephalin, Proopiomelanocortin	Dwarfism Obesity	[102] [103–107]
PCSK-2	(K/R)-(X)-(K/R)	Vesicles	Neuroendocrine tissue, brain	Proopiomelanocortin Prodynorphin	Susceptibility to diabetes	[108]
PCSK-3 Furin	(K/R)-(X)-(K/R)	TGN/PM	Ubiquitous	Membrane-type metalloproteinases, IGF-1R	Cancer
PCSK-4	(K/R)-(X)-(K/R)	(acrosomal) PM	Epididymis Spermatocyte	ADAM-1, IGF-2 (?)	Infertility	[109, 110]
PCSK-5	(K/R)-(X)-(K/R)	PCSK-5A: extracellular PCSK-5B: dense secretory granules	Heart, great blood vessels Kidneys, small intestine	Pro-renin Integrin α subunit MT1-MMP	Cardiovascular disease	[111]
PCSK-6	(K/R)-(X)-(K/R)	Extracellular	Ubiquitous	Similar to furin	Prostate cancer	[16, 17]
PCSK-7	(K/R)-(X)-(K/R)	TGN/PM	Ubiquitous	Pro-EGF	Learning and memory impairments	[112, 113]
PCSK-8 (SKI-1)	(R/K)X(hydrophobic)Z	TGN/PM	Ubiquitous	Pro-neurotrophic derived factor SREBP-1 and 2	Viral glycoproteins	[2, 114, 115]
PCSK-9	LVFAQSIP	ER/endosomes	Hepatocytes	PCSK-9 targets LDLR to lysosomal degradation	Atherosclerosis	[32, 116]

On the other hand, PCs seem to have minor differences in the cleavage preferences, and most of them can be equally efficient in the activations of their cognate substrates. Subtle differences among these proteases may be attributed to different tissue distribution, or subcellular locations [7, 10, 11]. In this context, PC-1 and PC-3 are restricted to neuroendocrine tissues, whereas PC4 is expressed almost exclusively in testes. However, the rest of the PCs, furin, PACE4, PC-5 and PC-7 demonstrated expression in virtually every tissue. Although PACE4 is fundamentally an extracellular proteins, furin, PC-5 and PC7 cycle between the endoplasmic reticulum and Golgi pointing to similar specificity and activity in analogous cellular and subcellular environments [9]. Although this statement may be valid in many cases, some substrates show a unique sensitivity to a specific PC. For instance, furin appears to be the only PCs that cleaves and activates TGF-β in vivo, supporting a non-redundant role for the PCs [12]. In addition, some PCs may show a differential expression in cancer or normal tissue. In fact, PACE4 is expressed predominantly in cells derived from the relatively normal ovarian surface epithelium, and silenced in ovarian neoplasia, whereas furin seems to follow the opposite pattern; highly expressed in tumors and tumor-derived cell lines, and relatively much less expressed in normal tissue and cell lines derived from them [13, 14]. Furthermore, PACE4 overexpression promotes the proliferation of prostate cancer cell lines in vitro and in vivo, whereas overexpression of other PCs do not seem to be similarly effective [15, 16].

Furin has emerged as the prototype of the family; however, several reports point to significant functions of PACE4 in prostate cancer [15, 16] and in the development of skin squamous cell carcinoma [17, 18] and musculoskeletal diseases [19] [20]. Hence, furin has been more extensively studied pointing to associations with several diseases, including cancer [6], inflammation [21, 22], cardiovascular disease [23–25] and viral infections [26], among others.

Furin is also essential for embryonic development. Knockdown of furin results in a lethal phenotype [1, 27–29]. Moreover, mouse embryos with the furin gene knockdown, specifically in endothelial cells, are born with several cardiovascular defects, including septal and valvular defects, attributable to impaired processing of TGF-β, adrenomedulin, endothelin, and bone morphogenetic protein 4; all of them furin substrates [30, 31]. On the other hand, furin knockouts in liver of adult mice are viable, pointing to certain degree of redundancy among PCs during adult life [11].

PCSK9 represents a unique case among the PCs since it does not cleave at the typical basic motif characteristic of the PCs in atherosclerosis. This PC cleaves and inactivates the low density lipoprotein receptor, LDLR, resulting in elevated blood LDL and VLDL [32]. Elevated levels of these lipoproteins are associated with hyperlipidemia, and the devastating effects of atherosclerosis. Inhibition of this PC resulted in increased levels of LDLR, with the concomitant decrease of plasma LDL and VLDL [33] [34]. These inhibitors have been the focus of successful clinical trials. At this stage, the first two PCSK9 inhibitors, alirocumab and evolocumab, have been in clinical use for lowering LDL-particle concentrations for cases in which classical statins and other drugs were not effective or badly tolerated [35, 36]. Recently, several excellent reviews have been published regarding the spectacular effects of PCSK9 inhibition in cardiovascular therapeutics. The reader is directed to these articles [32, 37–45]

2. Paralyzing the master switches: Inhibition of PCs

Several strategies have been studied to inhibit PCs in vitro and in vivo, from small molecule inhibitors to proteins and antibodies directed against these proteases. Furthermore, subtle differences in the amino acid sequences surrounding the active site have been exploited to increase specificity of an inhibitor towards a PC.

2.1 Small molecule inhibitors

Most of the small molecule inhibitors for the PCs block the accessibility of substrates to the catalytic pocket in a competitive manner. The structure of these inhibitors contain a tetrapeptide containing the furin cleavage site, that is, RXR/KR, or modifications thereof, conserving the crucial cationic nature of the PCs recognition site [46]. These short peptides may not offer specificity towards a particular PCs; however, they proved to bind the catalytic cleft with high affinity, with K_i in the micromolar or nanomolar range. Upon binding these inhibitors, in an induced-fit manner, both PC and inhibitor structure change their conformation, favoring the formation of an antiparallel beta sheets, characteristic of furin, PC5, PC7, and PACE4, pointing some degree of redundancy among these PCs [47]. However, differences in the amino acids surrounding the pocket where the substrate binds may facilitate the design of specific PC inhibitors.

2.1.1 Chloromethylketones

Chloromethylketone derivatives of the consensus sequence RVKR provided the first and strongest evidences of the tremendous potential of inhibiting PCs in terms of the PC mechanism of action, substrate processing, and potential clinical application [48]. The compound contains the recognition sequence for the PCs, an N-terminal decanoyl extension that increases its cell permeability, and a C-terminal chloromethyketone that blocks the possible in vivo protease-mediated degradation (figure 1A). The main caveat associated to these agents is their lack of specificity for any particular PC, and its toxicity at higher doses.

Figure 1.

Structure of main furin inhibitors. (A) Decanolyl R-V-K-R chloromethyl ketone, (B) General structure of polyarginines (C) Guanidilated streptamine. (D) 1. Prodomains of furin, 2. Inhibitory peptide directed against PACE4

2.1.2 Poly-arginine derivatives

Poly-arginines represent an extension of one of the basic motif recognized by PCs; RRRR. One of the first inhibitors synthesized was the hexa-D-arginine, an effective inhibitor of inflammation in vitro and in vivo [49]. More than a decade ago, a slightly longer peptide, poly-D-nonaarginine (D9R) proved to be an efficient furin inhibitor (K_i in the order in the nanomolar range), and administration of this peptide protected cells against the anthrax toxin with an EC₅₀ of 3.6 μM [50] [51]. This highly cationic peptide not only provides the sequence that binds to the active site of furin, but to the highly acidic motifs surrounding the active site. Although the delivery of this nonapeptide presents difficulties, it has provided the bases for potential modifications that led to promising newly formulated inhibitors [52] (figure 1B). More recently, cyclic peptides constructed using arginines and variable side chains demonstrated inhibitory activity against furin, combined with an enhanced ability for cell penetration [53].

2.1.3 Streptamine derivatives

As already stated, most of the PCs inhibitors are peptides or proteins. Although they are in general excellent inhibitors of PCs, the low turnover of peptides and proteins, their potential cytotoxicity, and the considerable molecular weight, (especially proteins), may limit their clinical use. Non-peptide inhibitors may circumvent these drawbacks, while still exerting a strong inhibitory effect on PCs. The elucidation of the crystal structure of furin has facilitated the modeling of these non-peptide inhibitors [54, 55]. Guanidinilated derivatives of 2-dideoxystreptamine, modeled using docking experiments and the crystal structure of furin as template, proved to be efficient (K_i in the nanomolar range) inhibitors of furin in vitro [56]. These derivatives mimic the cationic character of the PCs recognition site, and bind the active site of furin, hence, acting in a competitive manner (figure 1). These inhibitors seem to inhibit furin (and PC6B) with a ten-fold higher efficiency than PACE4 or PC7. In fact, guanidilated streptamine derivatives bind to PACE4 and PC7 with ten-fold and 100-fold lower efficiency, respectively(figure 1B) [55].

After the development of these derivatives to dideoxystreptamine, other groups developed the bisguanidinephenyl ethers derivatives of 2–5 dideoxystreptamine containing two guanidine residues [57]. These two positively charged guanidine group are attached to a phenyl group, respectively, and the guanidine phenyl moieties are linked by a three carbon bridge. This positive charge-bridge-positive charge structure resembles the minimal recognition site for the PCs-RXXR. In addition, the phenyl group increases the molecule’s hydrophobicity resulting in an enhanced penetration into the cell. The residues, bond by ether groups, confer extra chemical and biochemical stability (figure 1C).

Some of the bisguanidylated derivatives exhibit poor cell penetration, making them ideal for diseases that require a membrane-bound furin, which mainly catalyzes the cleavage of extracellular substrates, such as the anthrax toxin protective antigen. Variations in the positioning of the guanydil substituents in the aromatic group are localized to different intracellular compartments, such as endosomes and Golgi. As different substrates are putatively processed in different subcellular compartments, the selection of derivatives with a particular substitution pattern may affect the activation of different substrates, depending on the final destination of the substituted compound administered. In the future, these compounds may represent a breakthrough in PC.s, especially furin- inhibition, and may stimulate research in non-peptide PC inhibitors to increase the repertoire of drugs at our disposal.

2.2 Peptidomimetics

Small peptidomimetics combine the best of both worlds; small molecule and full-protein inhibitors. As small molecules, they exhibit better pharmacokinetic properties, better formulation, and delivery. As these compounds usually contain the PC recognition site embedded in a peptide moiety, they allow for specific interactions outside the binding pocket that are present in some, but not all of the PCs. These extra interactions strengthen the specific binding of the peptidomimetic to PC in a selective manner [58] (figure 1D).

Levesque et al (2012) [59] have synthesized a peptide containing the recognition sequence for PCs (RVKR) with a four Leucine residues extension at the N terminal end of this sequence (figure 1D). Although the binding site for furin and PACE4 are virtually identical, these investigators showed that some regions, specifically alpha helices 3 and 4, surrounding its catalytic site, contain a different array of amino acids. Furin exhibit in a highly anionic pocket between alpha helices 3 and 4 with an overall charge of −7, whereas this region is devoid of charge in PACE4. This lack of overall charge can be exploited to strengthen the inhibitor-PACE4 interaction. The four leucine residues extension provides a hydrophobic environment resulting in increased binding forces to PACE4. On the other hand, the higher hydrophobicity prohibits binding to furin. In fact, this inhibitor has proved to bind PACE4 with Ki in the nM range and about twenty times lower than furin, and is an efficient inhibitor of cancer cell proliferation [60, 61].

3. Work in progress: Peptides and proteins as models for the development of new therapeutic strategies

3.1 Pro-domain

PCs follow the secretory biosynthetic pathways, transiting from the endoplasmic reticulum to vesicles to the Golgi apparatus to the trans-Golgi compartment. PCs then cycle between the trans-Golgi network and the plasma membrane, transported back and forth through endocytic vesicles [62]. Hence, PCs exert their catalytic activity in these three compartments. PCs do not become activated until they reach the trans-Golgi network since the catalytic site is masked by the prosegment of the protein. After cleavage from the protein, the prosegment remains attached to the catalytic site, precluding substrate binding. In the trans-Golgi network, changes in pH favors the dissociation of the prosegment to the PC core, enabling the enzyme to exert its catalytic activities [62] (figure 1D). This approach has been used in several research studies [63] Furin prosegment, for instance decreased the activity of MMP-9 in breast cancer cells [64], inhibition of growth and invasiveness in head and neck cancer cell lines [65], prohepcidin maturation [66], and inhibition of the maturation of brain-derived neurotrophic factor [67]. Despite these promising studies, prosegments have not been evaluated in vivo, probably because full proteins display a similar or better inhibitory activity.

3.2 Alpha-1-antitrypsin and derivatives

Proteins provide not only the tetrapeptide domain that specifically binds PCs, but a scaffold that may provide specificity, increased affinity, and improved kinetic properties. Alpha1- antitrypsin, var Portland (PDX), is a bioengineered variant of antitrypsin containing the PCs recognition sequence. The original sequence of α₁-antitrypsin contains the sequence AIPM r 358 and selectively inhibits the activity of elastase. The change AIPM to AIPR, named α₁-antitrypsin Pittsburgh, not only lacks affinity to elastase, but originates a motif that serves as thrombin inhibitor. In fact, this variant has been discovered in a patient suffering from uncontrolled hemorrhages [68]. A bioengineered change in the sequence resulted in the motif RIPR 358 creating the minimal consensus sequence for PC binding [69]. This consensus sequence has minimal effect on the overall structure of protein, since it is in a loop that projects away from the core structure of the protein, a fact that has been exploited to inhibit extracellular substrates. This antitrypsin variant forms a stable, SDS and heat-resistant complex with PCs, especially furin, and PC6B, and to a lesser extent with PACE4, [70, 71]. Several in vitro and in vivo systems have employed PDX in order to decrease the activation of substrates relevant to cancer progression [72–74] viral infection and inflammation [75, 76].

PDX, a 50kDa protein, may establish multiple interactions with PCs besides those in the active site. In this context, some variants contain a recognition sequence that depart from the minimal consensus for PCs, and still binds efficiently to PCs. In this context, Hada et al, [77] demonstrated that the variant AVNR (instead of RIPR) binds furin with great selectivity, albeit with less efficiency. The selectivity displayed by these variants, that do not contain the minimal consensus for PCs, suggests that other important domains in the protein may contribute to the efficiency and selectivity of protein-based inhibitors. PDX has been a source of inspiration for many protein-based approaches.

4. Nanobodies: beyond competitive inhibition

PCs efficient cleavage relies not only on the integrity of the active site, but on the P-domain, located towards the C-terminal end of the protein, as shown in mutagenesis experiments [78]. Nanobodies, the variable region of the heavy chain of antibodies, VHH fragments, have been produced to target the P-domain of furin. The resulting antibody, Nb14, binds to the P-domain, without interfering with the activity of the catalytic domain, measured as binding and processing of small substrates [79]. Interestingly, binding of the nanobody Nb14 to furin precludes activation of full-length proteins, such as factor X. These somewhat puzzling results suggest that Nb14 may cover partially the catalytic site, preventing molecules with large molecular weights such as proteins to reach the binding site and be cleaved. In contrast, Nb14 binding to the P domain does not seem to disrupt catalytic activity, since small molecules, such as the fluorescent substrate 7-amino-4 methyl-coumarin, can be cleaved with both, furin or NB14-bound furin. Furthermore, Dec-RVKR-CMK binding to the catalytic site occurs regardless of the presence or not of Nb14.

The design and usage of a nanobody represents a novel finding in the field of PC inhibition. Nb14 binds to a regulatory region (P-region) within the molecule of furin rather than the catalytic site as most of the inhibitors developed so far. This non-competitive mode of inhibition represents a promising approach, and may stimulate the design of similar allosteric inhibitors with potential applications in the clinic. Furthermore, the mode of non-competitive action of Nb14 differs markedly from the mechanism of actions of other antibodies used to block the activity of other proteases, where these nanobodies bock the catalytic site, or induce a conformational change in the active site, preventing the binding of any substrate-large or small.

Nb14 is specifically recognizes furin P-domain; other PCs, such as PACE-4, PC1 and PC2 do not bind this nanobody. Cleavage of substrates that are specific for furin, such as hemagglutinin, or substrates that are cleaved with poor efficiency by other PCs, can be effectively blocked using nanobodies directed against furin. Further research may evaluate the effectiveness of this approach to treat pathologies associated to furin activity as well as to develop nanobodies specific for other PCs. Figure 2 provides a scheme of this therapeutic approach.

Figure 2.

Furin nanobodies. Furin nanobodies consisting of the variable region of an antibody specific for the P-domain of furin, binds this regulatory region, inhibiting the binding of furin to physiological substrates. Note the change in shape in the catalytic site upon Nb14 binding to the P-site.

5. Furin silencing: strengthening the immune response

Carcinogenesis is characterized in part by the expression of certain genes and the consequent production of new proteins that enable tumor cells to successfully compete with normal cells, leading to their survival. One of the properties that tumor cells acquire is the ability to escape the normal immune mechanisms mounting strong immunosuppressive responses. Although tumor cells exhibit a repertoire of antigens not present in normal cells and tissues, these potentially immunogenic markers are not recognized by the immune system. One of the culprits for immunosuppression is the transforming growth factor beta- TGFβ, both variants 1 and 2. Many tumor cell lines overexpress this extracellular factor, resulting in an extra advantage for survival, especially since TGF-β inhibits the secretion of CD8⁺T cell-dependent Granulocyte-Macrophage Colony Stimulating Factor (GMCSF), and, consequently, the maturation of dendritic cells in the bone marrow [80]. Low levels of dendritic cells translate into a deficient antigen presentation, thwarting the possibility of immune-mediated elimination of cancer cells.

Furin has been identified as the main activator of TGF-β. This growth factor is synthesized in the rER, and after separation of the signal peptide, it transits through the endomembrane system where furin cleaves this protein at the C-terminal end of the basic sequence RHRR A²⁷⁹ [81]. Although the other PCs can process TGF-β, it has been shown that furin is the most efficient among the PCs as demonstrated by restoration of normal levels of TGF-β after incorporating the furin cDNA into the furin-deficient cells LoVo, and PDX-mediated inhibition studies [12]. Furin processing is indispensable for activity since engineered mutants in the PC cleavage sequence (RHRR) produced secreted proteins, albeit inactive, which can act as dominant negative [82]. Hence, blocking furin activity may impair the activation of TGF-β and lead to decreased immune responses and successful elimination of tumor cells.

This approach has been investigated and seems to be promising in the various malignancies. The cDNA from GMCSF and a bidirectional small hairpin RNA interference targeting furin [83]were placed under the control of the strong cytomegalovirus promoter. This construct express elevated GMCSF stimulating the immune response and shRNAi for furin that silence this PC. When transfected to autologous cancer cells, this, these cells produced GMCSF, and efficiently silenced furin. Lack of furin expression led to impairment of TGF-β activation. These transfected, patient-derived tumor cells, were used as cellular vaccines (formerly named FANG, and now called VIGIL in the clinical trials) and reinjected in patients. These tumor cells were unable to effect immunosuppression, hence, becoming sensitive to GMCSF-derived responses. Elevated levels of GMCSF also resulted in increased activation of bone marrow-derived dendritic cells, facilitating the elimination of tumor cells. In a phase I trial, Nemunaitis et al [84–86] produced vaccines from a wide variety of tumor sources, including, small cell lung cancer, breast cancer, colon cancer, liposarcoma, and ovarian cancer [84] This approach is still underway, but anticipates positive outcomes in decreasing the relapse in patients diagnosed with hepatocellular carcinoma [84], metastatic advanced Ewing’s sarcoma [87, 88], advanced stage ovarian cancer [89].

These vaccines represent a safe therapy. Overall these vaccines point to effective treatments of advanced tumors, which frequently relapse, with poorer outcomes with successive relapses. The phase II clinical trial for advanced ovarian cancer showed a significant (P= 0.033) increase in relapse free survival from 481 days (control) to 826 days (VIGIL treatment). Few toxic effects have been reported, most of them at the injection site, including erythema, induration, pain, pruritus, swelling, and tenderness. Other effects might be associated to the therapy itself, such as joint function, back pain, and fatigue; however no definite association between treatment and these effects can be drawn with certainty.

6. Activating prodrugs at the right target: Taking advantage of the PCs cleavage sequence

Specific targeting of tumor cells poses a problem that has required creative approaches. Most drugs, small synthetic drugs or biologicals, are delivered systemically. Drugs may be toxic to normal cells and display unacceptable toxic effects. Prodrugs have provided a partial response to this (and other) problems however, the activation of the prodrug may not be achieved in the target cells. In response to this difficulty, researchers have focused on furin as a potential prodrug activating enzyme [90]. An early approach consisted to deliver the Tumor necrosis factor related apoptosis induced ligand (TRAIL) specific to cancer cells, avoiding TRAIL’s toxic effects on normal hepatocytes [91]. The soluble N-terminal domain of TRAIL was fused to the extracellular domain of a Fms-like tyrosine kinase (hFlex), a highly immunogenic peptide. Between these two genes, a DNA fragment containing the sequence for PCs cleavage was inserted in phase. Transfection of several cancer cell lines with this fusion plasmid (hFlex-furin sequence-soluble TRAIL) resulted in induction of apoptosis and cytolysis. Similar results were observed in vivo, after intratumoral treatment of subcutaneous tumors with this plasmid.

Other approaches to specifically deliver toxins to tumor cells have employed fusion proteins that include a peptide containing the PCs cleavage sequence acting as a linker between two structurally and functionally different proteins [92, 93] One of the most popular therapies using fusion proteins consisted to a fragment of the variable chain of the HER-2 antibody (e23sFv) fused to an internalization sequence, such as the Pseudomonas endotoxin A(PEA), linked to a proapoptotic or toxic proteins though a peptide containing the PCs cleavage sequence [93]. Several tumors, including up to 30% of breast cancers, colon, and lung cancers express the protein HER-2 [94–97] which is localized to the plasma membrane. The easy access to this protein facilitates the binding of the variable chain of HER-2 antibody tethering the fusion protein specifically to cancer cells expressing HER-2. When fused to the translocation domain of the Pseudomonas endotoxin A (PEA), the fusion protein containing the e23sFv fragment internalizes to the endomembrane system [98–100]. Once in the endomembrane system, the fusion protein is cleaved, releasing the proapoptotic or toxic protein. One of the pro-apopototic proteins fused to the HER-2, tBID (truncate BH3 interacting domain death agonist) was used in many cells and animal systems. This protein is transported to the mitochondria and stimulates the release of proapototic proteins such as Smac/DIABLO and AIF [101, 102], specifically eliminating cancer cells.

7. Conclusion

PCs play an important role in several physiological and pathological processes from embryonic development to cancer, viral infections to inflammation, apoptosis to angiogenesis. Although some PCs exhibit a restrictive tissue distribution, PC1 and PC2 in neuroendocrine tissues, and PC4 in testis, others, like furin, PACE4, PC5, and PC7 are ubiquitously expressed. The latter PCs are overexpressed in a series of malignancies. In addition, PCs are crucial for some virus to infect host cells, bacterial pathogenicity, and cardiovascular diseases. The development of PC inhibitors has shown great potential to cure or improve certain pathological conditions. Newer approaches are being continuously developed. Several clinical trials are currently being conducted, or initiated (table 2). It is hoped that future inhibitors may surpass the effectiveness of the ones already at hand. The possibility to mitigate and eliminate some of the most common lethal diseases of mankind, such as cancer, with the use of PC inhibitors and other anticancer approaches seems quite realistic at this time.

Table 2.

Ongoing clinical trials designed to evaluate therapies based on furin inhibition.

Clinical trial	PC targeted	Target	Phase	NCT number
Bi: shRNA-Furin GMCSF	Furin (PCSK-3)	Ewin Sarcoma Non small lung cancer Liver cancer	I	01061840
FANG™ vaccine + Carboplatinum	Furin (PCSK-3)	Stage III and stage IV Ovarian cancer	II	01867086
FANG™ vaccine + Bevacizumab	Furin (PCSK-3)	Stage III and stage IV Ovarian cancer	II	0155145
FANG™ vaccine	Furin (PCSK-3)	Advanced melanoma	II	01453361
VIGIL + atezolizumab	Furin (PCSK-3)	Gynecological cancers	II	03073525

Abbreviations

AIF

Apoptosis-inducing factor

cDNA

complementary DNA

DIABLO

Direct IAP binding protein with low pI

EGF

Epidermal growth factor

GMCSF

Granulocyte Macrophage colony-stimulating factor

HER-2

Human epidermal growth factor receptor 2

IAP

Inhibitor of apoptosis proteins

LDL

Low density lipoprotein

LDLRL

Low-density lipoprotein receptor

MMP-9

Matrix metalloproteinase 9

MT1-MMP

Membrane Type 1 metalloproteinase

PEA

Pseudomonas endotoxin A

PC

Proprotein convertases

PCSK

Proprotein convertases subtilisin and kexin type

PDX

Alpha1- antitrypsin, var Portland

PM

Plasma membrane

shRNAi

Short hairpin interference ribonucleic acid

Smac

Second mitochondria-derived activator of caspases

SREBP-1

Sterol regulatory element binding protein 1

TGN

Trans Golgi network

TGF-β

Transforming growth factor beta

tBID

Truncated BH3 interacting domain death agonist

TRAIL

Tumor necrosis factor related apoptosis induced ligand

VLDL

Very low density lipoproteins

VHH

Variable region of the heavy chain of antibodies