Protein Epitope Mimetics: From New Antibiotics to Supramolecular

Jun 1, 2017 - John A. Robinson was born in London and received a Ph.D. from the University ... Epitope mimetics may also have great potential in the f...
0 downloads 0 Views 6MB Size
Download PDF
Article pubs.acs.org/accounts

Protein Epitope Mimetics: From New Antibiotics to Supramolecular Synthetic Vaccines Published as part of the Accounts of Chemical Research special issue “Chemical Biology of Peptides”. Katja Zerbe, Kerstin Moehle, and John A. Robinson* Chemistry Department, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland CONSPECTUS: Protein epitope mimetics provide powerful tools to study biomolecular recognition in many areas of chemical biology. They may also provide access to new biologically active molecules and potentially to new classes of drug and vaccine candidates. Here we highlight approaches for the design of folded, structurally defined epitope mimetics, by incorporating backbone and side chains of hot residues onto a stable constrained scaffold. Using robust synthetic methods, the structural, biological, and physical properties of epitope mimetics can be optimized, by variation of both side chain and backbone chemistry. To illustrate the potential of protein epitope mimetics in medicinal chemistry and biotechnology, we present studies in two areas of infectology; the discovery of new antibiotics targeting essential outer membrane (OM) proteins in Gram-negative bacteria and the design of supramolecular synthetic vaccines. The discovery of new antibiotics with novel mechanisms of action, in particular to combat infections caused by Gram-negative pathogens, represents a major challenge in medicinal chemistry. We were inspired by naturally occurring cationic antimicrobial peptides to design structurally related peptidomimetics and to optimize their antimicrobial properties through library synthesis and screening. Through these efforts, we could show that antimicrobial β-hairpin mimetics may have structures and properties that facilitate interactions with essential bacterial β-barrel OM proteins. One recently discovered family of antimicrobial peptidomimetics targets the β-barrel protein LptD in Pseudomonas spp. This protein plays a key role in lipopolysaccaride (LPS) transport to the cell surface during OM biogenesis. Through a highly selective interaction with LptD, the peptidomimetic blocks LPS transport, resulting in nanomolar antimicrobial activity against the important human pathogen P. aeruginosa. Epitope mimetics may also have great potential in the field of vaccinology, where structural information on complexes between neutralizing antibodies and their cognate epitopes can be taken as a starting point for B cell epitope mimetic design. In order to generate potent immune responses, an effective method of delivering epitope mimetics to relevant cells and tissues in the immune system is also required. For this, engineered synthetic nanoparticles (synthetic virus-like particles, SVLPs) prepared using supramolecular chemistry can be designed with optimal surface properties for efficient dendritic cell-mediated delivery of folded B-cell and linear T-cell epitopes, along with ligands for pattern recognition receptors, into lymphoid tissues. In this way, multivalent display of the epitope mimetics occurs over the surface of the nanoparticle, suitable for cross-linking B cell receptors. In this highly immunogenic format, strong epitope-specific humoral immune responses can be elicited that target infections caused by pathogenic microorganisms. Other potential applications of epitope mimetics in next-generation therapeutics are also discussed.

■

INTRODUCTION An emerging paradigm in the study of biomolecular recognition is the design and synthesis of protein epitope mimetics. This field has been fueled by the huge growth in high-resolution 3D structures of biomacromolecules and the complexes they form. Of special interest are the relationships between epitope structure, dynamics, and recognition. Often, elements of stable secondary structure on the surface of proteins mediate protein−protein interactions (PPIs). In such cases, excised linear or unstructured (i.e., unfolded) peptides representing merely the amino acid sequence but not the folded structure of the epitope will often fail to capture the binding affinity and specificity for a target needed for biological activity in the © 2017 American Chemical Society

nanomolar range. This is not always the case, however, since examples of biomolecular recognition are also known where unfolded, or “nascently folded” epitopes can effectively mediate binding to a target through an induced-fit binding mechanism.1 Synthetic epitope mimetics provide one tool to explore the complex relationships between conformational dynamics and biomolecular recognition across this broad spectrum of binding mechanisms. Such studies may also open new approaches to the design of biologically active molecules and, potentially, new classes of drug and vaccine candidates. Received: March 15, 2017 Published: June 1, 2017 1323

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research

Figure 1. Design of macrocyclic β-hairpin epitope mimetics. A hairpin epitope (pink/green) is transplanted from a target protein (left, here an antibody V domain) onto a template (D-Pro-L-Pro; the type-II′ β-turn in this dipeptide is shown), which acts to nucleate β-hairpin structures in the attached loop, which may be typically from 4 to 20 residues (shown is a 12-residue loop).

versatile approach to the design of macrocyclic β-hairpin mimetics with interesting biological activities, as discussed below.

The application of peptide-based epitope mimetics as inhibitors of PPIs has attracted much attention recently.2 The modular construction of peptide-based probes facilitates the incorporation of diverse side chain chemistries and conformational constraints. The ability of epitope mimetics to adopt the types of folded secondary structure recognized by protein targets allow them to display multiple hot residues for cooperative binding, possibly exploiting networks of stabilizing interactions. If, in addition, robust and efficient methods for synthesis can be established, efficient optimization of initial hits to improve target affinity, specificity, and physical or even druglike (ADMET) properties may become possible.3 Examples from two different areas are described below, to illustrate the great potential of epitope mimetics in studies of chemical biology and as next-generation therapeutics.

■

ANTIBIOTICS WITH NOVEL MECHANISMS OF ACTION Cationic antimicrobial peptides (CAMPs) play important roles in defense against microbial infections in many organisms.6 CAMPs display a variety of secondary structures in solution or membrane-mimetic environments, including folded β-hairpin structures stabilized by internal disulfide bridges (Figure 2A).7 One example is protegrin-I (PG-I) from porcine leukocytes, which displays broad-spectrum antimicrobial activity in the low micromolar range. The amphipathic structure of PG-I has a detergent-like effect resulting in lysis of microbial cell membranes.7 Unfortunately, this mechanism-of-action limits the therapeutic value of PG-1 as an antibiotic due to problems with toxicity. Nevertheless, the β-hairpin scaffold is found in CAMPs from many different organisms. From this background, efforts were started to discover structurally related macrocyclic antimicrobial peptidomimetics with more favorable properties. Initially loops of varying size attached to a D-Pro-L-Pro template were explored. Interesting results were obtained with 12-residue loops containing side-chains typically found in naturally occurring CAMPs. Optimization was facilitated by an efficient parallel synthesis method allowing the iterative production and testing of small libraries of macrocycles with diversified sequences. An initial success was the discovery of peptidomimetic L8-1 (Figure 2B), which showed broadspectrum micromolar antimicrobial activity but was largely devoid of membrane lytic activity.8 In further rounds of optimization, a substantial gain in activity against Pseudomonas aeruginosa was achieved at the expense of broad-spectrum activity (e.g., with L19-45). The final round of library synthesis

■

APPROACHES TO FOLDED EPITOPE MIMETICS In this large and rapidly evolving field, several structure-based design approaches can be envisaged. The starting point is usually an epitope of known structure in a naturally occurring target peptide or protein. Both backbone secondary structure and side-chain positions can be used in mimetic design, with diverse chemistries and cross-links (e.g., helix-stabilizing staples) harnessed to optimize folding of the mimetic. Alternatively, the epitope might be grafted onto alternative scaffolds, including non-natural foldamers or scaffolds designed to mimic only side-chain locations rather than backbone secondary structure.2 An approach to β-hairpin epitope mimetics was described in an earlier report, by transplanting a loop from a target protein onto a semirigid template (Figure 1).4 One very convenient template comprises the dipeptide D-Pro-L-Pro, because of its highly constrained type-II′ β-turn conformation, which is ideal for nucleating β-structure in an attached peptide loop.5 This is a 1324

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research

Figure 3. Synthetic peptidomimetic antibiotics. (A) LB-01 has potent antimicrobial activity, and NMR reveals a folded hairpin structure (shown), whereas NMR structures of the inactive LB-02, with an inverted template sequence, show no defined hairpin fold. (B) Sequence and NMR structure of JB-95. Figure 2. (A) CAMPs (five examples are shown) provide the inspiration for a new class of synthetic peptidomimetic antibiotics. (B) Cationic and hydrophobic residues (as in PG-1) are incorporated into a 12-residue β-hairpin mimetic scaffold (gray). Stepwise optimization of initial hits, involving library synthesis and screening for antimicrobial activity, led to a unique pseudomonas-specific family of antibiotics (Dab = L-2,4-diaminobutyric acid).

An important clue to the mechanism of action came from analysis of spontaneous resistant mutants of P. aeruginosa, which occur at very low frequency.9 One mutant contained a six residue tandem duplication in the gene lptD, and this allele conferred protection against the antibiotic when introduced into the wild-type strain. Photolabeling experiments with a derivative of L27-11 containing L-photoproline (L-proline-4spiro-3-(3H-diazirine)) in place of L-Pro and a biotin tag at position-1 (called PAL-1) led to remarkably specific labeling of LptD in P. aeruginosa, thereby demonstrating a direct binding interaction with the protein.9 More recent studies have shown that L27-11 not only binds to LptD, but also inhibits its essential function in the bacterium, which is described below.12 The Gram-negative OM represents an important permeability barrier to lipophilic molecules, due to its asymmetric organization with lipopolysaccharide (LPS) in the external leaflet and phospholipids in the inner leaflet (Figure 4). The phosphate groups in the lipid A and inner core regions of neighboring LPS molecules are bridged by Mg2+ or Ca2+ ions, which is important for OM function and stability. Substantial progress in understanding how this unique bilayer is biosynthesized has been made recently.13 New LPS molecules are assembled at the inner membrane (IM), and then transported to the cell surface by seven essential Lpt (lipopolysaccharide transport) proteins (LptA−G, Figure 4). An oligomeric form of LptA bridges the periplasm and delivers LPS molecules to a complex comprising the β-barrel OM protein LptD and the lipoprotein LptE. The function of LptD/ E is to translocate LPS molecules from the periplasmic side into the outer leaflet of the OM. Crystal structures are now available for several Lpt proteins, including LptD/E from several different Gram-negative

(library 27) was an alanine scan, to identify side chains important for activity. Interestingly, one member of the library showed an improved activity rather than a loss (L27-11; Figure 2B). This compound shows potent antibiotic activity in the nanomolar range specifically against Pseudomonas spp.9 Although the peptidomimetic is active only against Pseudomonas sp., the high potency and novel mechanism of action (vide inf ra) suggest applications in the treatment of infections caused by P. aeruginosa, one of only three pathogens on the 2017 WHO critical list for antibiotic research. Interestingly, the enantiomeric form of L27-11 showed no antimicrobial activity, consistent with a mechanism of action involving a chiral target. The alanine scan also showed that the aromatic side chains of Trp2 and Trp8 are very important for antibiotic activity.10 Both L27-11 and a closely related derivative (LB-01), optimized for NMR studies, showed potent antibiotic activity and a stable βhairpin conformation in aqueous solution (Figure 3A).10 However, a related derivative (LB-02) with the D-Pro-L-Pro template inverted to L-Pro-D-Pro, shows no stable hairpin conformation and the antimicrobial activity was lost (Figure 3A), suggesting that the β-hairpin structure is crucial for interaction with the bacterial target. This conclusion was reinforced in recent studies with an NMe-scan of peptide bonds around the macrocycle.11 1325

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research

Figure 4. Schematic of the membrane organization in Gram-negative bacteria, and the essential roles played by the Bam complex and Lpt proteins in OM biogenesis (see text). (left) β-Barrel proteins are exported across the IM by the Sec machinery, chaperoned across the periplasm, with folding and insertion into the OM catalyzed by the Bam complex (structure shown is from PDB file 5D0O). (right) The LPS transport pathway comprises LptA−G, with LptD/E (structure from PDB 4Q35) acting in the final step to transport LPS into the outer leaflet of the OM. A typical structure of LPS from P. aeruginosa is shown.

bacteria.13 LptD contains a highly conserved 26-stranded Cterminal β-barrel domain embedded in the OM and an Nterminal periplasmic domain composed of a U-shaped βjellyroll (Figure 4).14,15 LptA also contains a jellyroll, whose hydrophobic groove provides a track for movement of the lipid chains of LPS across the periplasm from LptA to LptD. One side of the LptD β-barrel can transiently accommodate the LPS sugar chain, as it orients vertically on its way to the cell exterior. An external loop on the β-barrel opens transiently to allow the sugar chain to emerge onto the cell surface, while the lipid chains emerge into the outer leaflet through a transient opening in the side of the β-barrel. This model for LPS transport is supported by extensive crystallographic, mutagenesis, and computer modeling studies.16 Upon treatment with L27-11, defects in the morphology of the OM in P. aeruginosa, which are characteristic of a block in LPS transport, can be detected by electron microscopy (EM).9,12 The binding of L27-11 to LptD appears to block the route used by LPS on its way to the cell surface. So far no other small molecules or natural products are known that block the LptD/E transporter in Gram-negative bacteria. To generate a clinical lead, the drug-like properties of L27-11 required optimization, since initial studies showed rapid degradation by trypsin-like enzymes in human serum.9 Scientists at Polyphor AG were able to optimize the ADMET properties, and one

clinical lead (POL7001) containing multiple Arg/Lys-to-Dab substitutions is much more stable toward proteolysis (Figure 2). The closely related clinical candidate (POL7080) called murepavadin, recently successfully completed phase-II clinical tests in hospital patients with life-threatening lung infections (clinical trials identifier NCT02096328). A pivotal phase-III study is now planned on the road to eventual clinical use of this new antibiotic to treat pseudomonas infections. These results also raise the question whether related molecules can be found, perhaps targeting other essential bacterial OM proteins. One promising step was the recent report of a different β-hairpin peptidomimetic called JB-95 (Figure 3B), showing potent antimicrobial activity against Escherichia coli.17 While JB-95 exhibits no lytic activity on human red blood cells, transmission EM and fluorescence studies reveal an ability to selectively disrupt the OM but not the IM of E. coli.17 A fluorescently labeled derivative was used to stain live E. coli cells. Super-resolution fluorescence microscopy revealed a punctated staining pattern, with selective labeling of discrete islands or patches across the cell surface. In a complementary approach, photochemical labeling experiments in E. coli using a derivative of JB-95 containing L-photoproline in place of L-Pro and a biotin-tag in place of Arg9 revealed cross-linking to several β-barrel OM proteins, including BamA and LptD.17 1326

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research Interestingly, recent fluorescence labeling studies have shown that in E. coli many OM proteins (including BamA) are not uniformly distributed across the cell surface but rather associate into islands or spatial clusters in the OM.18 Binding of JB-95 to BamA/LptD in these OM islands would therefore explain both the fluorescence microscopy and photochemical labeling results. An interaction of the antibiotic with BamA is interesting, because this β-barrel OM protein is the main component of the Bam complex, the essential machinery used by Gram-negative bacteria to catalyze folding and insertion of β-barrel proteins into the OM (Figure 4).19 Unfolded nascent OM proteins are exported across the IM using the Sec translocon and escorted across the periplasm bound to dedicated chaperones. Upon delivery to the Bam complex, comprising BamA and four lipoproteins BamB/C/D/E, β-barrel folding and insertion occurs into the OM.20 Indeed, LptD is one of the most important substrates of the Bam complex. It has been estimated that folding, OM insertion, and disulfide oxidation of an LptD molecule into the OM requires about 20 min, 21 an extraordinarily long time given that E. coli cells double during exponential growth every ∼20−40 min. Recently, several crystal structures of the Bam complex have been reported (Figure 4).22 The mechanism(s) of Bammediated β-barrel folding and insertion into the OM are still under intensive investigation.19,23 One hypothesis is that the Bam complex may act as a membrane bilayer “disruptase”, by creating a thinner less densely packed bilayer near to the β1−β16 strand-junction of the BamA β-barrel, thereby destabilizing the membrane to assist folding and insertion of a new β-barrel OM protein.24,25 It is not yet clear whether or how the OM disruptions caused by JB-95 might be linked to its interaction with BamA. However, semiquantitative proteomic studies with E. coli cells treated with JB-95 revealed a substantial depletion of many βbarrel proteins in the OM that might be explained by inhibition or disruption of the Bam complex.17 However, this effect might also be caused by sRNAs, deployed by envelope stress responses, which negatively regulate expression of many OMPs.17 Future in vitro studies might help to clarify this interesting question.

their own to stimulate strong protective immune responses in humans, a problem that can be partly offset by formulation with immunostimulatory adjuvants. By the start of the 21st century, DNA genome sequencing technologies were being used to identify hundreds of potential exposed surface proteins on bacteria, each of which could be produced as recombinant proteins and tested to identify antigens that could be used as components of a protective vaccine. This “reverse vaccinology” approach has recently succeeded in generating a new four component protein-based vaccine against meningococcus B.27 On the other hand, many promising protein vaccine candidates from pathogenic microorganisms are poorly stable and cannot be easily produced for clinical use. The production of virus-neutralizing antibodies (nAbs) is thought to be important in conferring protection against infection by antiviral vaccines. A new impetus in the field has come from B-cell cloning technologies, which have allowed the isolation of selected B cells excreting nAbs from human B cell repertoires taken from vaccinated or infected persons.28 Single B cell sorting and culturing allows screening of naturally occurring B cells producing nAbs, and sequencing of the corresponding Ig genes. Structural characterization of such nAbs bound to their target antigen allows a structural definition of the protective epitope and provides a starting point for epitope mimetic and vaccine design, an approach also called “structural vaccinology”.29 Most effort has focused on viral pathogens, since viruses typically have only a few proteins on their surface. Nevertheless, some viruses such as HIV-1, present special obstacles for vaccine design, not least because HIV-1 mutates at a very high frequency, and uses conformational mobility (shape-shifting) and glycosylation to mask sensitive epitopes and evade the immune system. Abs are produced by B-cells following activation in the lymph nodes by cross-linking their B cell receptors (BCRs) during binding to cognate pathogen-derived antigens, delivered through lymph-fluid or blood, and presented by macrophages or dendritic cells (DCs).30 Protein antigens are mostly presented to naive B cells in a folded state. Although the entire surface of a protein may be antigenic, the epitopes recognized by protective or nAbs are of special interest.31 Activated B cells then undergo repeated rounds of somatic hypermutation of their BCRs and selection, with help from activated T follicular CD4 helper (TfH) cells in germinal centers in lymphoid tissues. TfH−B cell interactions lead to the production of high affinity, class-switched antibodies following infection or vaccination. One challenge is to design epitope mimetics that activate B and T cells to produce nAbs against the target pathogen. Another is to devise a suitable method to deliver the epitope mimetics to both naive and activated B- and T-cells in the lymphoid tissues.32 To achieve this goal, there is presently considerable interest in formulation using 20−100 nm engineered nanoparticles that can display multiple copies of the folded antigen.33 With suitable surface properties, this size range is compatible with efficient capture by circulating DCs and transport through the lymphatic system. On their way to the lymph nodes, the DCs also process the pathogen, leading to the release of chemical signals that have a powerful activating effect on the DC and on other immune cells, through the actions of secreted cytokines.34 This includes the production by proteolysis of pathogen-derived T cell epitopes in the form of linear peptides (9−15 residues in length). These T cell epitopes are subsequently recycled to the DC surface, bound to MHC-I and MHC-II molecules, where they can be recognized by T-cell

■

PROTEIN EPITOPE MIMETICS AND STRUCTURAL VACCINOLOGY The potential of synthetic peptides as antigens to generate focused immune responses against invading pathogens has long been recognized. But today there are still no vaccines in human use based upon synthetic peptides or epitope mimetics. Recent scientific and technological advances, however, suggest that this reductionist goal might soon become feasible. The history of vaccinology demonstrates that more limited reductionist approaches to vaccine design have already been a great success.26 In the 18th and 19th centuries, Edward Jenner and Louis Pasteur showed that live attenuated and killed microorganisms could elicit protective immune responses in humans against smallpox, rabies, and anthrax, among others. During the early 20th century, immune protection was achieved by subunit vaccines, such as the purified polysaccharide coat from bacteria such as meningococcus and pneumococcus. By the 1980s recombinant DNA technology had been harnessed to produce individual pathogen-derived proteins as subunit vaccines against hepatitis B, human papillomavirus, and others. However, most proteins are insufficiently immunogenic on 1327

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research

Figure 5. Protein epitope mimetics and the virosome delivery system in synthetic (malaria) vaccine design.

Figure 6. (A) Designed synthetic lipopeptide building blocks self-assemble into nanoparticles containing B- and T-cell epitopes as well as TLR ligands. (B) A synthetic V3 loop mimetic on SVLPs designed to elicit anti-HIV-1 antibodies.

receptors on CD8 and CD4 T-cells, respectively.35 Additional signals are provided by pathogen-associated-molecular-patterns (PAMPs), which are recognized, for example, by the Toll-like and NOD-like receptors (TLRs and NLRs).36 Among known PAMPs are lipidated peptides, heterocycles, LPS, and small RNA molecules that are characteristic of the invading pathogen. When these pathogen-derived ligands bind to TLRs, trafficking of the DCs to the lymph nodes is enhanced, and the expression of surface MHC proteins and selected cytokines is up-regulated. The engineered nanoparticles, therefore, should contain multivalent folded B- and linear T-cell epitopes, as well as PAMPs that can stimulate TLRs or NLRs or both. In this way, the nanoparticles incorporate key signals required for activation of the innate and adaptive immune systems. Two examples from recent work illustrate approaches to addressing these design challenges. In our efforts to design a synthetic malaria vaccine, interest focused on the circumsporozoite (CS) protein, an immunodominant surface protein on the sporozoite stage of Plasmodium falciparum, carried by live mosquitoes.37 The central region of this membrane-anchored CS-protein contains tandem repeats comprising (NANP)∼37. Early attempts to exploit synthetic linear (NANP)3 peptides conjugated to tetanus toxin (to supply T-cell epitopes) in alum adjuvant (for immune stimulation) as a vaccine gave disappointing results. Although the crystal structure of the intact CS protein is not available, a short Ac-ANPNA-NH2 peptide could be crystallized in a type-I β-turn conformation.38 Such β-turns might also propagate along the repeat region of the CS protein, so efforts were made to incorporate β-turn-stabilizing elements

into synthetic NPNA-repeats, for example, by substituting the Pro for L-α-methylproline.39 Further optimization led to a 20residue macrocyclic derivative, called UK-40 (Figure 5), which proved to have structural and antigenic similarity to the repeat region of the CS protein on sporozoites.40 For delivery to the immune system, the macrocyclic NPNAmimetic was coupled to phosphatidylethanolamine (UK-39, Figure 5) and loaded onto the surface of immunopotentiating reconstituted influenza virus-like particles or virosomes.41 Virosomes are ∼100 nm unilamellar liposomes, prepared by removal of excess lipid from a mixture of natural and synthetic phospholipids and influenza surface glycoproteins. Upon reconstitution in the presence of the synthetic lipopeptide (UK-39), the epitope mimetic is displayed on the surface of the virosome. In this way, delivery of UK-39 to primed mice and rabbits was shown to elicit high titers of antigen-specific and sporozoite cross-reactive antibodies, which could inhibit invasion of human hepatocytes by P. falciparum sporozoites.40 An effective malaria vaccine is likely to be a multicomponent, multistage vaccine. So a second lipopeptide was also developed to elicit antibodies against the apical membrane antigen 1 (AMA-1), a membrane protein located within the apical complex of the merozoite surface of the liver-stage parasite. A disulfide cross-linked synthetic peptide based upon a semiconserved loop in AMA-1 was identified and shown to induce asexual blood-stage parasite-growth inhibitory antibodies using the virosome delivery system.42 In collaboration with the Swiss Tropical and Public Health Institute and the industry partner Pevion, the UK-39 and AMA-1 mimetics were subsequently combined in a virosome formulation that entered clinical trials 1328

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research in Europe and Africa.43,44 In phase I/II clinical studies, the vaccine was well tolerated and both synthetic components elicited strong antigen-specific antibody responses in all immunized volunteers. A promising observation made during clinical trials in Africa was that the incidence of clinical malaria episodes in children receiving the vaccine was about half that of the control group.45 These encouraging results, albeit with only a small number of patients, must be extended to larger patient cohorts. Current efforts are also directed toward to the identification of additional vaccine components, targeting other proteins involved in parasite infection, which might further boost the efficacy of this malaria vaccine. One disadvantage of the virosome technology is the requirement for influenza viral membrane glycoproteins, which are expensive and poorly stable. Therefore, an alternative fully synthetic nanoparticle delivery system for epitope mimetics was developed. This delivery system comprises stable, homogeneous nanoparticles in the 20−30 nm size range that can be generated from designed synthetic lipopeptide building blocks.46 The lipopeptide building blocks contain a hyperstable parallel trimeric coiled-coil motif, fused to a CD4 T-helper epitope, and on one terminus residues that influence the surface properties of the nanoparticle and provide an anchor for conjugation to a B-cell epitope mimetic (Figure 6A). At the alternate end of the peptide chain, a lipid moiety such as Pam2Cys or Pam3Cys can be attached, which are potent activators of TLR2. Such lipopeptide building blocks can be produced efficiently by a convergent synthesis procedure. A key property of the lipoprotein building blocks is their ability to spontaneously self-assemble into supramolecular nanoparticles in aqueous buffers.47−49 The self-assembly process includes formation of parallel trimeric helical bundles through the coiled-coil motif (Figure 6A). Association of multiple helical bundles is then driven by sequestration of the clustered lipid chains into the core of a micelle-like nanoparticle.47,48 Each nanoparticle presents an array of about 70− 80 copies of the B- and T-cell epitopes. In one study, porcine DCs were shown to rapidly bind the nanoparticles, which were then internalized using multiple endocytic routes, before being slowly processed by proteolysis.50 These nanoparticles contain multiple signals required to stimulate strong adaptive immune responses (without external adjuvant) and have been called synthetic virus-like particles (SVLPs).46 SVLPs were produced by coupling the macrocyclic malaria epitope mimetic (N-deacetyl-UK40) (Figure 5) to lipopeptide building blocks containing a promiscuous CD4 T helper epitope from the CS protein.47 After three rounds of immunization, these SVLPs elicited high titers of antigenspecific, parasite cross-reactive antibodies in rabbits. Several follow-on studies have confirmed the broad applicability of the SVLP delivery system for synthetic vaccine design. One application focused on an immunodominant hairpin loop, called the V3 loop, found in the HIV-1 envelope glycoprotein gp120. In this case, a macrocyclic β-hairpin V3 loop mimetic, comprising the V3 loop attached to the D-Pro-L-Pro template, was coupled to SVLPs (Figure 6B).49 These V3-SVLPs elicited high titers of V3-specific antibodies in rabbits, which also recognized recombinant gp120 by ELISA and showed HIV-1 neutralizing activity against laboratory tier-1 strains.49 Finally, in a proof-of-concept study for a pneumococcal vaccine with the industry partner Virometix, an SVLP-based immunogen was shown recently to elicit a protective humoral immune response

against a highly virulent Streptococcus pneumoniae strain in a mouse infection model.51

■

FUTURE OUTLOOK In this Account, we have attempted to illustrate how synthetic approaches to epitope mimetics can lead to new families of biologically active molecules. The scope of this approach is not limited to research in antibiotics and molecular immunology, since in principle, wherever protein (peptide)−protein interactions occur, opportunities will arise to exploit structural information for epitope mimetic design. This approach should, therefore, remain a rich avenue for biomolecular research and discovery for many years to come.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail address: [email protected]. ORCID

John A. Robinson: 0000-0001-7857-8556 Funding

This research was supported by the Swiss National Science Foundation and the Swiss Committee for Technology and Innovation. Notes

The authors declare the following competing financial interest(s): J.A.R. is an investor in UZH start-up companies Polyphor AG and Virometix AG. Biographies Katja Zerbe was born in Mannheim and received her Ph.D. from the University of Zurich. Since 1996, she has been a senior scientist at UZH working with Professor Robinson on various topics in the field of microbiology. Her current work is focused on the mechanisms of action of new antibiotics. Kerstin Moehle was born in Brandenburg and received her Ph.D. from the University of Leipzig. She joined the group of Professor Robinson in 1999 as a senior scientist. Her specialist area is computational/theoretical chemistry, and her current research is focused on structural studies of peptidomimetics. John A. Robinson was born in London and received a Ph.D. from the University of Cambridge. Since 1989, he has been on the Chemistry faculty of the University of Zurich, where he is Professor of Bioorganic Chemistry. He directs research on various topics in the field of chemical biology, with a current focus on antibiotics and synthetic vaccines.

■

REFERENCES

(1) Wright, P. E.; Dyson, H. J. Linking folding and binding. Curr. Opin. Struct. Biol. 2009, 19, 31−38. (2) Gopalakrishnan, R.; Frolov, A. I.; Knerr, L.; Drury, W. J.; Valeur, E. Therapeutic Potential of Foldamers: From Chemical Biology Tools To Drug Candidates? J. Med. Chem. 2016, 59, 9599−9621. (3) White, A. M.; Craik, D. J. Discovery and optimization of peptide macrocycles. Expert Opin. Drug Discovery 2016, 11, 1151−1163. (4) Robinson, J. A. β-Hairpin peptidomimetics: design, structures and biological activities. Acc. Chem. Res. 2008, 41, 1278−1288. (5) Favre, M.; Moehle, K.; Jiang, L.; Pfeiffer, B.; Robinson, J. A. Structural mimicry of canonical conformations in antibody hypervariable loops using cyclic peptides containing a heterochiral diproline template. J. Am. Chem. Soc. 1999, 121, 2679−2685.

1329

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research (6) Brown, K. L.; Hancock, R. E. Cationic host defense (antimicrobial) peptides. Curr. Opin. Immunol. 2006, 18, 24−30. (7) Nguyen, L. T.; Haney, E. F.; Vogel, H. J. The expanding scope of antimicrobial peptide structures and their modes of action. Trends Biotechnol. 2011, 29, 464−472. (8) Shankaramma, S. C.; Athanassiou, Z.; Zerbe, O.; Moehle, K.; Mouton, C.; Bernardini, F.; Vrijbloed, J. W.; Obrecht, D.; Robinson, J. A. Macrocyclic hairpin mimetics of the cationic antimicrobial peptide protegrin I: A new family of broad-spectrum antibiotics. ChemBioChem 2002, 3, 1126−1133. (9) Srinivas, N.; Jetter, P.; Ueberbacher, B. J.; Werneburg, M.; Zerbe, K.; Steinmann, J.; Van der Meijden, B.; Bernardini, F.; Lederer, A.; Dias, R. L. A.; Misson, P. E.; Henze, H.; Zumbrunn, J.; Gombert, F. O.; Obrecht, D.; Hunziker, P.; Schauer, S.; Ziegler, U.; Kach, A.; Eberl, L.; Riedel, K.; DeMarco, S. J.; Robinson, J. A. Peptidomimetic Antibiotics Target Outer-Membrane Biogenesis in Pseudomonas aeruginosa. Science 2010, 327, 1010−1013. (10) Schmidt, J.; Patora-Komisarska, K.; Moehle, K.; Obrecht, D.; Robinson, J. A. Structural studies of β-hairpin peptidomimetic antibiotics that target LptD in Pseudomonas sp. Bioorg. Med. Chem. 2013, 21, 5806−5810. (11) Vetterli, S. U.; Moehle, K.; Robinson, J. A. Synthesis and antimicrobial activity against Pseudomonas aeruginosa of macrocyclic β-hairpin peptidomimetic antibiotics containing N-methylated amino acids. Bioorg. Med. Chem. 2016, 24, 6332−6339. (12) Werneburg, M.; Zerbe, K.; Juhas, M.; Bigler, L.; Stalder, U.; Kaech, A.; Ziegler, U.; Obrecht, D.; Eberl, L.; Robinson, J. A. Inhibition of Lipopolysaccharide Transport to the Outer Membrane in Pseudomonas aeruginosa by Peptidomimetic Antibiotics. ChemBioChem 2012, 13, 1767−1775. (13) Okuda, S.; Sherman, D. J.; Silhavy, T. J.; Ruiz, N.; Kahne, D. Lipopolysaccharide transport and assembly at the outer membrane: the PEZ model. Nat. Rev. Microbiol. 2016, 14, 337−345. (14) Dong, H. H.; Xiang, Q. J.; Gu, Y. H.; Wang, Z. S.; Paterson, N. G.; Stansfeld, P. J.; He, C.; Zhang, Y. Z.; Wang, W. J.; Dong, C. J. Structural basis for outer membrane lipopolysaccharide insertion. Nature 2014, 511, 52−56. (15) Qiao, S.; Luo, Q. S.; Zhao, Y.; Zhang, X. J. C.; Huang, Y. H. Structural basis for lipopolysaccharide insertion in the bacterial outer membrane. Nature 2014, 511, 108−111. (16) Li, X. J.; Gu, Y. H.; Dong, H. H.; Wang, W. J.; Dong, C. J. Trapped lipopolysaccharide and LptD intermediates reveal lipopolysaccharide translocation steps across the Escherichia coli outer membrane. Sci. Rep. 2015, 5, 11883. (17) Urfer, M.; Bogdanovic, J.; Lo Monte, F.; Moehle, K.; Zerbe, K.; Omasits, U.; Ahrens, C. H.; Pessi, G.; Eberl, L.; Robinson, J. A. A Peptidomimetic Antibiotic Targets Outer Membrane Proteins and Disrupts Selectively the Outer Membrane in Escherichia coli. J. Biol. Chem. 2016, 291, 1921−1932. (18) Rassam, P.; Copeland, N. A.; Birkholz, O.; Toth, C.; Chavent, M.; Duncan, A. L.; Cross, S. J.; Housden, N. G.; Kaminska, R.; Seger, U.; Quinn, D. M.; Garrod, T. J.; Sansom, M. S. P.; Piehler, J.; Baumann, C. G.; Kleanthous, C. Supramolecular assemblies underpin turnover of outer membrane proteins in bacteria. Nature 2015, 523, 333−336. (19) Plummer, A. M.; Fleming, K. G. From Chaperones to the Membrane with a BAM! Trends Biochem. Sci. 2016, 41, 872−882. (20) Noinaj, N.; Rollauer, S. E.; Buchanan, S. K. The β-barrel membrane protein insertase machinery from Gram-negative bacteria. Curr. Opin. Struct. Biol. 2015, 31, 35−42. (21) Lee, J.; Xue, M. Y.; Wzorek, J. S.; Wu, T.; Grabowicz, M.; Gronenberg, L. S.; Sutterlin, H. A.; Davis, R. M.; Ruiz, N.; Silhavy, T. J.; Kahne, D. E. Characterization of a stalled complex on the betabarrel assembly machine. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 8717−8722. (22) Gu, Y. H.; Li, H. Y.; Dong, H. H.; Zeng, Y.; Zhang, Z. Y.; Paterson, N. G.; Stansfeld, P. J.; Wang, Z. S.; Zhang, Y. Z.; Wang, W. J.; Dong, C. J. Structural basis of outer membrane protein insertion by the BAM complex. Nature 2016, 531, 64−69.

(23) Rollauer, S. E.; Sooreshjani, M. A.; Noinaj, N.; Buchanan, S. K. Outer membrane protein biogenesis in Gram-negative bacteria. Philos. Trans. R. Soc., B 2015, 370, 20150023. (24) Noinaj, N.; Kuszak, A. J.; Gumbart, J. C.; Lukacik, P.; Chang, H. S.; Easley, N. C.; Lithgow, T.; Buchanan, S. K. Structural insight into the biogenesis of beta-barrel membrane proteins. Nature 2013, 501, 385−390. (25) Fleming, K. G. A combined kinetic push and thermodynamic pull as driving forces for outer membrane protein sorting and folding in bacteria. Philos. Trans. R. Soc., B 2015, 370, 20150026. (26) Delany, I.; Rappuoli, R.; De Gregorio, E. Vaccines for the 21st century. EMBO Mol. Med. 2014, 6, 708−720. (27) De Gregorio, E.; Rappuoli, R. From empiricism to rational design: a personal perspective of the evolution of vaccine development. Nat. Rev. Immunol. 2014, 14, 505−514. (28) Rappuoli, R.; Bottomley, M. J.; D’Oro, U.; Finco, O.; De Gregorio, E. Reverse vaccinology 2.0: Human immunology instructs vaccine antigen design. J. Exp. Med. 2016, 213, 469−481. (29) Dormitzer, P. R.; Grandi, G.; Rappuoli, R. Structural vaccinology starts to deliver. Nat. Rev. Microbiol. 2012, 10, 807−813. (30) De Silva, N. S.; Klein, U. Dynamics of B cells in germinal centers. Nat. Rev. Immunol. 2015, 15, 137−148. (31) Corti, D.; Lanzavecchia, A. Broadly Neutralizing Antiviral Antibodies. Annu. Rev. Immunol. 2013, 31, 705−742. (32) Beitelshees, M.; Li, Y.; Pfeifer, B. A. Enhancing vaccine effectiveness with delivery technology. Curr. Opin. Biotechnol. 2016, 42, 24−29. (33) Narasimhan, B.; Goodman, J. T.; Vela Ramirez, J. E. Rational Design of Targeted Next-Generation Carriers for Drug and Vaccine Delivery. Annu. Rev. Biomed. Eng. 2016, 18, 25−49. (34) Iwasaki, T.; Medzhitov, R. Control of adaptive immunity by the innate immune system. Nat. Immunol. 2015, 16, 343−353. (35) Roche, P. A.; Furuta, K. The ins and outs of MHC class IImediated antigen processing and presentation. Nat. Rev. Immunol. 2015, 15, 203−216. (36) Maisonneuve, C.; Bertholet, S.; Philpott, D. J.; De Gregorio, E. Unleashing the potential of NOD- and Toll-like agonists as vaccine adjuvants. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 12294−12299. (37) Schwenk, R. J.; Richie, T. L. Protective immunity to preerythrocyte stage malaria. Trends Parasitol. 2011, 27, 306−314. (38) Ghasparian, A.; Moehle, K.; Linden, A.; Robinson, J. A. Crystal structure of an NPNA-repeat motif from the circumsporozoite protein of the malaria parasite Plasmodium falciparum. Chem. Commun. 2006, 174−176. (39) Bisang, C.; Weber, C.; Inglis, J.; Schiffer, C. A.; Van Gunsteren, W. F.; Jelesarov, I.; Bosshard, H. R.; Robinson, J. A. Stabilization of Type-I beta-Turn Conformations in Peptides Containing the NPNARepeat Motif of the Plasmodium falciparum Circumsporozoite Protein by Substituting Proline for (S)-alpha-Methylproline. J. Am. Chem. Soc. 1995, 117, 7904−7915. (40) Okitsu, S. L.; Kienzl, U.; Moehle, K.; Silvie, O.; Peduzzi, E.; Mueller, M. S.; Sauerwein, R. W.; Matile, H.; Zurbriggen, R.; Mazier, D.; Robinson, J. A.; Pluschke, G. Structure-activity-based design of a synthetic malaria peptide eliciting sporozoite inhibitory antibodies in a virosomal formulation. Chem. Biol. 2007, 14, 577−587. (41) Pfeiffer, B.; Peduzzi, E.; Moehle, K.; Zurbriggen, R.; Glück, R.; Pluschke, G.; Robinson, J. A. A virosome-mimotope approach to synthetic vaccine design and optimization: Synthesis, conformation, and immune recognition of a potential malaria-vaccine candidate. Angew. Chem., Int. Ed. 2003, 42, 2368−2371. (42) Müller, M. S.; Renard, A.; Boato, F.; Vogel, D.; Naegeli, M.; Zurbriggen, R.; Robinson, J. A.; Pluschke, G. Induction of parasite growth-inhibitory antibodies by a virosomal formulation of a peptidomimetic of loop I from domain III of Plasmodium falciparum apical membrane antigen I. Infect. Immun. 2003, 71, 4749−4758. (43) Okitsu, S. L.; Mueller, M. S.; Amacker, M.; Vogel, D.; Westerfeld, N.; Robinson, J. A.; Zurbriggen, R.; Pluschke, G. Preclinical profiling of the immunogenicity of a two-component 1330

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331

Article

Accounts of Chemical Research subunit malaria vaccine candidate based on virosome technology. Hum. Vaccines 2008, 4, 106. (44) Thompson, F. M.; Porter, D. W.; Okitsu, S. L.; Westerfeld, N.; Vogel, D.; Todryk, S.; Poulton, I.; Correa, S.; Hutchings, C.; Berthoud, T.; Dunachie, S.; Andrews, L.; Williams, J. L.; Sinden, R.; Gilbert, S. C.; Pluschke, G.; Zurbriggen, R.; Hill, A. V. S. Evidence of Blood Stage Efficacy with a Virosomal Malaria Vaccine in a Phase IIa Clinical Trial. PLoS One 2008, 3, e1493. (45) Cech, P. G.; Aebi, T.; Abdallah, M. S.; Mpina, M.; Machunda, E. B.; Westerfeld, N.; Stoffel, S. A.; Zurbriggen, R.; Pluschke, G.; Tanner, M.; Daubenberger, C.; Genton, B.; Abdulla, S. Virosome-Formulated Plasmodium falciparum AMA-1 & CSP Derived Peptides as Malaria Vaccine: Randomized Phase 1b Trial in Semi-Immune Adults & Children. PLoS One 2011, 6, e22273. (46) Boato, F.; Thomas, R. M.; Ghasparian, A.; Freund-Renard, A.; Moehle, K.; Robinson, J. A. Synthetic virus-like particles from selfassembling coiled-coil lipopeptides and their use in antigen display to the immune system. Angew. Chem., Int. Ed. 2007, 46, 9015−9018. (47) Ghasparian, A.; Riedel, T.; Koomullil, J.; Moehle, K.; Gorba, C.; Svergun, D. I.; Perriman, A. W.; Mann, S.; Tamborrini, M.; Pluschke, G.; Robinson, J. A. Engineered Synthetic Virus-Like Particles and Their Use in Vaccine Delivery. ChemBioChem 2011, 12, 100−109. (48) Perriman, A. W.; Williams, D. S.; Jackson, A. J.; Grillo, I.; Koomullil, J. M.; Ghasparian, A.; Robinson, J. A.; Mann, S. Synthetic viruslike particles and hybrid constructs based on lipopeptide selfassembly. Small 2010, 6, 1191−1196. (49) Riedel, T.; Ghasparian, A.; Moehle, K.; Rusert, P.; Trkola, A.; Robinson, J. A. Synthetic Virus-Like Particles and Conformationally Constrained Peptidomimetics in Vaccine Design. ChemBioChem 2011, 12, 2829−2836. (50) Sharma, R.; Ghasparian, A.; Robinson, J. A.; McCullough, K. C. Synthetic Virus-Like Particles Target Dendritic Cell Lipid Rafts for Rapid Endocytosis Primarily but Not Exclusively by Macropinocytosis. PLoS One 2012, 7, e43248. (51) Tamborrini, M.; Geib, N.; Marrero-Nodarse, A.; Jud, M.; Hauser, J.; Aho, C.; Lamelas, A.; Zuniga, A.; Pluschke, G.; Ghasparian, A.; Robinson, J. A. A Synthetic Virus-Like Particle Streptococcal Vaccine Candidate Using B-Cell Epitopes from the Proline-Rich Region of Pneumococcal Surface Protein A. Vaccines 2015, 3, 850− 874.

1331

DOI: 10.1021/acs.accounts.7b00129 Acc. Chem. Res. 2017, 50, 1323−1331