Table 1. Cyclic peptide drugs approved from 2006-2014 by the FDA and/or EMA
| Approved | Generic name | Indication | Mode of action | MW (Da) | Route of administration | Company |
|---|---|---|---|---|---|---|
| 2006 | Anidulafungin | Fungal infections | Fungal 1,3-β-d-glucan synthase inhibitor | 1140 | IV infusion | Vicuron/Pfizer |
| 2007 | Lanreotide | Acromegaly, neuroendocrine tumors | Growth hormone release inhibitor | 1156 | SC, IM | Ipsen |
| 2009 | Telavancin | Complicated skin and skin structure infections (CSSSIs), nosocomial pneumonia | Bacterial cell-wall synthesis inhibitor | 1756 | IV infusion | Theravance |
| 2009 | Romidepsin | Cutaneous T-cell lymphoma (CTCL), peripheral T-cell lymphomas (PTCLs) | Histone deacetylase inhibitor | 541 | IV infusion | Gloucester Pharmaceuticals/Celgene |
| 2012 | Peginesatide | Anemia associated with chronic kidney disease | Erythropoiesis stimulating agent | 4415 + 40 kDa PEG | IV, SC | Affymax/Takeda |
| 2012 | Linaclotide | Constipation-predominant irritable bowel syndrome (IBS-C) and chronic idiopathic constipation (CIC) | Guanylate cyclase 2C receptor activator | 1527 | OR | Forest Labs/Ironwood Pharmaceuticals |
| 2012 | Pasireotide | Cushing’s disease, acromegaly, neuroendocrine tumors | Growth hormone release inhibitor | 1047 | SC, IM | Novartis |
| 2014 | Dalbavancin | Complicated skin and skin structure infections (CSSSIs) | Bacterial cell-wall synthesis inhibitor | 1817 | IV infusion | Durata Therapeutics/Teva |
| 2014 | Oritavancin | Complicated skin and skin structure infections (CSSSIs) | Bacterial cell-wall synthesis inhibitor | 1793 | IV infusion | The Medicines Company |
The three antibacterials on this list, telavancin, dalbavancin and oritavancin, are semi-synthetic cyclic lipoglycopeptides [4]. They belong to the same drug class as the established antibiotics vancomycin and teicoplanin, which all contain a common heptapeptidic core with five fixed residues that serves as the main binding site for the D-Ala-D-Ala target. Binding of these drugs to their target blocks the transpeptidation of peptidoglycan precursors in the bacterial cell wall [5]. These three new antibiotics all contain a lipophilic side chain that is thought to increase the dwell time near the target by anchoring them to the cell membrane and/or destabilize the bacterial membrane. Interactions of the hydrophobic tails with cell membranes and plasma proteins also prolong the plasma half-life. These three drugs are used for the treatment of complicated skin and skin structure infections and nosocomial pneumonia. The small variations in their structures cause subtle differences in their pharmacologic effects by fine-tuning their activities towards different bacterial strains or differing pharmacokinetic properties.
The cyclic peptide anidulafungin is a member of the class of echinocandin antifungals to which caspofungin and micafungin, approved in 2001 and 2005, also belong [6]. All three drugs share a similar peptide core formed by six amino acids, two of which are proline derivatives and two are threonine or threonine derivatives. Anidulafungin, derived from echinocandin B, is a natural fermentation product of Aspergillus nidulans wherein the lineoyl side-chain tail is replaced by a lipophilic alkoxytriphenyl group. Like caspofungin and micafungin, anidulafungin inhibits the 1,3-β-d-glucan synthase responsible for fungal cell wall synthesis. Anidulafungin has a high affinity for human plasma proteins and a slow degradation time, giving it a half-life in the body of around 24 hours, doubling that of caspofungin and micafungin.
The oncology drugs lanreotide and pasireotide are analogues of the successful cyclic peptide drug octreotide, which itself is an analogue of the endogenous cyclic peptide hormone somatostatin [7]. Somatostatin blocks the release of hormones such as the growth hormone by inhibiting G-protein-coupled somatostatin receptors. These new mimetics that have a substantially longer plasma half-life than somatostatin are used to treat acromegaly and endocrine tumors. Lanreotide, like octreotide, is a disulfide cyclized octapeptide wherein four amino acids are identical and the other four are closely related. The two drugs also have similar pharmacologic properties. Pasireotide’s structure was reduced to a hexapeptide cyclized through an amide bond and its sequence conserves only the d-tryptophan-lysine core of somatostatin [8]. Pasireotide differs from octreotide and lanreotide in its substantially higher affinity for certain receptor subtypes, making it more suitable for treating Cushing’s disease and acromegaly in patients unresponsive to the other agents.
The third oncology drug, romidepsin is a natural product isolated from gram-negative Chromabacterium violaceum, approved for the treatment of T-cell lymphomas [9]. It is a depsipeptide composed of five backbone-cyclized residues and a disulfide bond that creates a bicyclic structure. The disulfide-cyclized romidepsin is a pro-drug that is significantly more stable in plasma than the active drug. Once inside a cell, it is reduced to its active form wherein a free thiol group chelates the zinc ion in the active site of intracellular histone deacetylase (HDAC) enzymes. The drug inhibits the removal of acetyl groups from lysine residues of N-terminal histone tails, maintaining a more open and transcriptionally active chromatin state which alters gene expression. Romidepsin is the second HDAC inhibitor that entered the market after vorinostat, which is a small molecule drug.
Linaclotide is a 14-amino acid cyclic peptide derived from heat-stable enterotoxins, cyclic peptides produced by various Escherichia coli strains which are a frequent cause of diarrhea [10••]. The drug provides well-tolerated relief in patients with chronic constipation and irritable bowel syndrome. Linaclotide contains three disulfide bonds that constrain the conformation of the peptide, providing high proteolytic stability. The drug acts by binding to guanylate cyclase C on the surface of intestinal epithelial cells [11]. Activation of the cyclase triggers a signaling pathway that leads to chloride and fluid secretions into the lumen, increasing colonic transit. The location of this target in the intestine allows for linaclotide to be orally administered while not orally available, as the drug does not have to pass through the gastrointestinal lining [12].
The final cyclic peptide approved in the last ten years is peginesatide, an agonist of the erythropoietin receptor developed for severe anemia in which erythropoietin cannot be used [13]. This cyclic peptide was discovered through phage display panning against the erythropoietin receptor [14, 15•]. It was then chemically dimerized with a polyethylene glycol (PEG), increasing the affinity, potency and circulation time in blood. It was voluntarily withdrawn from the market, however, one year after its approval due to safety concerns not experienced during the clinical trials. Unlike all other previously approved cyclic peptide drugs, peginesatide was the first to be developed de novo.
Taken together, eight of the newly approved nine drugs have a mechanism of action similar to established drugs and only one, linaclotide, is a first-in-class drug. Besides being highly innovative and serving a large group of patients, linaclotide is also a highlight in terms of sales. It generated revenues of 460 million USD, improving upon the already impressive sales of 140 and 300 million USD in the two first years after approval, and it is likely to become a blockbuster drug within the next five years [16]. Lanreotide and romidepsin are the other two drugs approved in the last ten years to have noteworthy sales, earning 440 and 70 million USD last year, respectively.
Table 2. Cyclic peptides developed by de novo design or in vitro evolution that are currently under clinical evaluation
| Phase | Name | Indication | Mode of action | Discovery platform | Company |
|---|---|---|---|---|---|
| 2 | POL7080 | Pseudomonas aeruginosa infections, gram-negative infections | LptD protein homolog inhibitor, inhibits outer-membrane biogenesis | Iterative modification and screening of peptide libraries using cationic antimicrobial peptide protegrin I as starting point | Polyphor |
| 2 | POL6326 | Metastatic breast cancer, acute myocardial infarction, hematopoietic stem cell mobilization from donors | Chemokine receptor CXCR4 antagonist prevents the binding of stromal derived factor-1 (SDF-1) to mobilize stem cells | Designed protein epitope mimetic (PEM) based on natural CXCR4 inhibitor T22 | Polyphor |
| 2 | APL-2 | Age related macular degeneration, paroxysmal nocturnal hemoglobinuria | Complement factor C3 inhibitor, blocks complement activation, PEGylated | Phage display screening of cyclic peptide library | Apellis |
| 1/2 | ALRN-6924 | Acute myeloid leukemia, hematological leukemia hematological malignancies, myelodysplastic syndromes, solid tumors | Inhibitor of p53-MDM2/MDMX interaction, restores p53-mediated apoptotic activity in tumors | Peptide isolated by phage display and modified into stapled peptide | Aileron |
| 1 | RA101495 | Paroxysmal nocturnal hemoglobinuria | Complement factor C5 inhibitor, blocks complement induced hemolysis | mRNA display screening of cyclic peptide libraries containing unnatural amino acids | Ra Pharma |
A look at the techniques used to develop these clinical candidates shows that all depended heavily on peptide library screening. In the case of the two peptides POL7080 and POL6326, natural peptides served as starting points but were extensively modified over several rounds of amino acid substitution and activity testing. As described above, POL7080 was so changed in its sequence that a new mechanism of action had evolved. The development of ALRN-6924 was encouraged by successes with stapled peptides that were rationally designed from the α-helices of p53 located at the p53-MDM2/MDMX binding interface. The template that was eventually used for the development of ALRN-6924 was derived from a phage display peptide library. Finally, the drug candidates APL-2 and RA101495 were developed by in vitro evolution through large random peptide library screening using either phage display or mRNA display. These and other powerful in vitro evolution techniques are now being broadly applied to develop ligands for a wide range of targets. It can be expected that many more de novo cyclic peptides with activities for interesting or previously unexplored targets will be fished out of random libraries by phage or mRNA display, and will soon be evaluated in clinical studies.
Future challenges and potential solutions
Several important challenges remain in the development of cyclic peptide therapeutics, the two most important ones being oral availability and cell permeability. Innovative approaches applied in recent years address one or both of these challenges, including the use of cell penetrating peptides [
24
], the stabilization of peptides in α-helical conformations with hydrocarbon linkers [
25
], or the in vitro evolution of N-methylated peptides [
26
]. In order to develop cyclic peptides that have a good oral availability and that efficiently enter cells by passive diffusion, it might be necessary to consider molecules with smaller molecular weights and fewer peptide bonds. One could imagine that future macrocycles will contain only one or a few amino acids with an additional polyketide type component in the backbone. A good role model is the orally available and cell permeable drug tacrolimus which contains a backbone based on one amino acid and a polyketide chain wherein the amino acid forms key interactions with the target. To develop such chimeric peptide/polyketide macrocycle ligands, it will be crucial to create efficient chemistries and strategies that allow the synthesis and screening of large combinatorial libraries.