Plant Cyclic Peptides: Discovery from Viola alpina to Next-Generation Macrocyclic Drug Design

Introduction

The discovery that plants produce cyclic peptides — head-to-tail macrocyclic molecules with no free N- or C-termini — has fundamentally expanded the structural space of bioactive peptides beyond the linear paradigm that dominated peptide chemistry for most of the 20th century. Plant cyclic peptides, particularly the cyclotide family first characterized from Viola alpina (alpine violet) and other Violaceae species, represent an evolutionary solution to the challenge of producing stable, protease-resistant peptide scaffolds capable of surviving the gastrointestinal tract of herbivores. With over 400 cyclotide sequences now characterized across the Violaceae, Rubiaceae, Cucurbitaceae, Fabaceae, and Solanaceae plant families, these macrocyclic peptides have emerged as privileged scaffolds for drug design, offering exceptional thermal stability (T_m > 95 °C for many cyclotides), resistance to enzymatic degradation, and a conserved structural framework amenable to epitope grafting.

Cyclotides: The Archetypal Plant Cyclic Peptide Family

Viola alpina cyclic peptide research pioneered the structural elucidation of the cyclotide family. Cyclotides are defined by their unique topology: a head-to-tail cyclized peptide backbone of 28-37 amino acids reinforced by six conserved cysteine residues that form three interlocking disulfide bonds in a characteristic knotted arrangement known as the cyclic cystine knot (CCK) motif. This CCK topology creates an exceptionally compact and rigid structure — the disulfide knot effectively locks the peptide into a conformation that resists thermal denaturation, chemical reduction, and proteolytic degradation by all major classes of proteases. X-ray crystallography of kalata B1 (the prototypical cyclotide from Oldenlandia affinis) and NMR studies of cycloviolacin O2 from Viola odorata revealed that the CCK framework presents six structurally defined "loops" (loops 1-6) between successive cysteine residues, with loops 2, 3, 5, and 6 forming a solvent-exposed surface that can be extensively modified without disrupting the knotted core — a property exploited in epitope grafting strategies.

Biosynthesis: From Linear Precursor to Macrocyclic Product

Cyclotide biosynthesis proceeds via ribosomal synthesis of a larger precursor protein, followed by proteolytic processing and cyclization — a pathway entirely distinct from non-ribosomal peptide synthesis (NRPS) that produces many bacterial cyclic peptides. The cyclotide precursor contains an ER-targeting signal peptide, an N-terminal pro-region, one to three mature cyclotide domains, and a C-terminal vacuolar sorting determinant. Asparaginyl endopeptidase (AEP), also known as legumain or vacuolar processing enzyme (VPE), catalyzes both the cleavage of the pro-region and the critical transpeptidation reaction that forms the peptide bond between the N- and C-termini. The discovery that the same enzyme performs both the cleavage and cyclization reactions — effected by a subtle pH-dependent shift in active-site geometry — was a landmark in understanding how ribosomal peptide macrocyclization occurs in nature. The six cysteine residues in the mature domain undergo oxidative folding in the ER lumen, catalyzed by protein disulfide isomerase (PDI), to form the CCK motif before vacuolar targeting.

Structure-Activity Relationships and Bioactivities

Cyclotides exhibit a remarkably diverse pharmacological profile that reflects the evolutionary arms race between plants and herbivores/pathogens:

Insecticidal Activity

Kalata B1 and cycloviolacin O2 display potent insecticidal activity against Helicoverpa armigera (cotton bollworm) larvae with LD₅₀ values of 0.8-2.0 μM in artificial diet assays. Mechanistically, cyclotides bind to phosphatidylethanolamine-rich membranes in the insect midgut epithelium, inducing pore formation and cell lysis — a mechanism that exploits the compositional difference between insect and mammalian cell membranes. This membrane-disrupting activity is critically dependent on a surface-exposed patch of hydrophobic residues in loops 2 and 3, with single-point mutations in this region reducing hemolytic activity by >100-fold without affecting CCK stability.

Antimicrobial Properties

Several cyclotides, including circulin A and cycloviolacin O2, exhibit broad-spectrum antimicrobial activity against Gram-negative (E. coli, MIC 1-4 μM), Gram-positive (S. aureus, MIC 0.5-2 μM), and fungal (C. albicans, MIC 4-8 μM) pathogens. The antimicrobial mechanism involves electrostatic interaction with anionic bacterial membrane phospholipids followed by toroidal pore formation, with selectivity ratios (therapeutic index) of 10-50 between bacterial and mammalian cell membranes. Notably, the disulfide-stabilized cyclic backbone confers resistance to bacterial proteases including subtilisin and thermolysin, distinguishing cyclotides from linear antimicrobial peptides that are rapidly degraded in the infection microenvironment.

Anti-HIV and Anti-Cancer Activities

Cycloviolacin O2 and kalata B1 inhibit HIV-1 replication in CEM-SS cells with EC₅₀ values of 40-130 nM, acting at the viral entry stage by disrupting the viral envelope membrane. In cancer models, cyclotides engineered with a tumor-targeting RGD (Arg-Gly-Asp) motif grafted into loop 6 demonstrated selective cytotoxicity against αvβ3 integrin-expressing melanoma and glioblastoma cell lines, with IC₅₀ values of 2-8 μM and no toxicity to normal fibroblasts at 50 μM.

Chemical Synthesis: Total Synthesis and Epitope Grafting

From Viola alpina to the Drug Development Pipeline

The translational trajectory of plant cyclic peptides spans fundamental discovery in ethnobotany to engineered therapeutic candidates in clinical development:

Phase 1: Ethnobotanical Discovery (1960s-1990s)

The uterotonic activity of Oldenlandia affinis tea, used by Congolese women to accelerate childbirth, led to the isolation of kalata B1 by Lorents Gran in 1973. The cyclic nature of kalata B1 was not recognized until 1995 when Craik and colleagues solved its 3D solution structure by NMR, revealing the unprecedented CCK topology. Systematic screening of Violaceae species subsequently identified Viola alpina and related alpine violets as rich sources of cyclotide diversity, yielding 15-25 distinct cyclotide sequences per species.

Phase 2: Scaffold Engineering (2000s-2010s)

The recognition that cyclotide loops could be extensively modified without disrupting the CCK core launched the field of cyclotide grafting. In a seminal 2008 study, Craik and colleagues grafted a 10-residue peptide antagonist of pro-angiogenic angiopoietin-2 into loop 6 of kalata B1, creating a grafted cyclotide (MCoTI-II variant) that inhibited Tie2 receptor signaling with an IC₅₀ of 2 nM — essentially identical potency to the parent antibody but with oral stability. Subsequent work demonstrated successful grafting of protease inhibitory loops, VEGF-A antagonists, and melanocortin receptor ligands into cyclotide scaffolds, establishing a modular platform for converting linear bioactive peptides into stable, orally active macrocycles.

Phase 3: Clinical Candidates and Beyond (2020s-)

Several cyclotide-based therapeutics have entered preclinical development. SZN-043, a cyclotide-grafted Wnt surrogate agonist developed by Surrozen, demonstrated efficacy in murine models of inflammatory bowel disease with oral bioavailability of 5-8%. Cyclotide-grafted alpha-4-beta-7 integrin antagonists (Protagonist Therapeutics) achieved sustained receptor occupancy >90% after oral dosing in non-human primates, a milestone that validates the cyclotide scaffold as an oral delivery vehicle for peptide therapeutics — a goal that has eluded the pharmaceutical industry for decades. Meanwhile, the solved structures of biosynthetic enzymes including OaAEP1 (the cyclase from O. affinis) are enabling the enzymatic synthesis of designer cyclotides with non-natural loop sequences at preparative scale.

Conclusion

The journey from the alpine meadows where Viola alpina grows to the cutting-edge of macrocyclic drug design illustrates how fundamental natural products discovery can spawn entirely new therapeutic modalities. Plant cyclic peptides have evolved over millions of years to solve precisely the problems that have historically limited peptide therapeutics: proteolytic instability, poor oral bioavailability, and conformational flexibility. The cyclotide scaffold — with its genetically encodable sequence, chemical synthesizability, and tolerance of extensive loop engineering — now stands as one of the most promising platforms for transforming biologically active but pharmacokinetically fragile linear peptides into drug-like molecules. As enzymatic cyclization technologies mature and structure-guided grafting methods improve, the prospect of orally administered peptide drugs for indications ranging from inflammatory bowel disease to oncology is increasingly within reach — all tracing their conceptual roots to a small alpine violet.