The rising incidence of severe fungal infections, due largely to increased numbers of immunocompromised patients, has made the discovery of new compounds that have effective antifungal activity or that support the activity of currently used antifungal agents more important than ever. Natural products are a significant source of novel antimicrobial compounds, and an increasing number are being identified from mammalian sources including milk proteins. One example is the multifunctional iron-binding glycoprotein, lactoferrin (Lf), and its various derived peptides. This article will focus on the spectrum of antifungal activity shown by Lf when used alone and in combination with other antifungal drugs, and will review currently proposed mechanisms of antifungal action for Lf and its derived peptides.

First identified almost 80 years ago in bovine milk (Sorensen and Sorensen, 1939) and later isolated and purified from humans and cows in 1960 (Groves, 1960), Lf is produced by various mammals including pigs, goats, dogs, mice, and horses (Lönnerdal and Iyer, 1995; Garcia-Montoya et al., 2012). In humans, it is a major protein secreted by the epithelial cells into exocrine fluids such as tears, saliva, vaginal fluids, gastrointestinal fluids, bile, urine, and sweat (Masson et al., 1968; van der Strate et al., 2001; González-Chávez et al., 2009; Park et al., 2011) but is present at the highest concentrations in human colostrum and breast milk (Lönnerdal and Iyer, 1995). Lf is involved in various physiological functions including iron regulation and the innate immune response, and has been demonstrated to have anti-inflammatory, anti-oxidant, anti-allergic, and anti-cancer properties (Conneely, 2001; Kruzel et al., 2006; Actor et al., 2009; Iigo et al., 2009; Burrow et al., 2011; Ogasawara et al., 2014) in addition to antimicrobial activity against a wide range of fungal, bacterial, viral, and parasitic pathogens (Jenssen and Hancock, 2009; Leboffe et al., 2009; Berlutti et al., 2011; Giansanti et al., 2013). The ability to bind and sequester iron is an important component of the antimicrobial activity of Lf (Jenssen and Hancock, 2009).

In addition to being an active intact protein, Lf is a rich source of antimicrobial peptides that are cleaved from the polypeptide chain by various proteolytic enzymes. As many of these enzymes are present in various body sites in humans, it is expected that a number of the cleaved peptides contribute to the normal functions of Lf in the body (Sinha et al., 2013). An increasing number of Lf-derived peptides are also being generated and modified by chemical synthesis (van der Kraan et al., 2005a). Many Lf peptides exhibit a broad spectrum of antimicrobial activity, have various different structural elements and modes of action, and appear to perform numerous functions including modulation of the inflammatory response, and stimulation of apoptosis (Farnaud et al., 2004; Sánchez-Gómez et al., 2008; van der Does et al., 2012).

The structure of Lf consists of a single polypeptide chain containing ~700 amino acids folded into two globular lobes. These lobes are highly homologous to each other and are connected by an α-helical residue providing flexibility to the Lf molecule. Each lobe consists of two domains forming a single iron binding site, allowing each Lf molecule to strongly bind two ferric ions (Anderson et al., 1989). The open, iron-free form of Lf is known as apo-Lf while the closed, iron-rich form is holo-Lf, which differs in tertiary structure and is more resistant to proteolysis than apo-Lf (Jameson et al., 1998). In addition to binding Fe³⁺, Lf has also been observed to bind a range of other compounds including lipopolysaccharides, heparin, DNA, and metal ions including Cu²⁺, Zn²⁺, and Mn²⁺ (Baker, 1994).

Lf is most commonly purified from human (hLf) or bovine (bLf) milk but it has also been obtained from the milk and colostrum of several other mammals. Recombinant Lf (rLf) is increasingly being generated from microbial expression systems and through transgenic plants and animals (Lönnerdal and Iyer, 1995). Lf molecules from across species are highly homologous but have varying antifungal activity owing to small differences in structure. Furthermore, the level of activity of Lf with similar purities can vary substantially between different producers, ranging from reasonably active to inactive in the same test organism. bLf in particular has well-documented activity against a range of human and plant fungal pathogens including yeasts and molds (Table 1). Most widely studied is its activity on members of the Candida species, with Candida tropicalis, Candida krusei, and Candida albicans exhibiting the highest level of susceptibility while Candida glabrata has the lowest (Nikawa et al., 1993; Xu et al., 1998).

Early studies with Candida species attributed the antifungal effect of Lf to its ability to sequester iron resulting in a fungistatic effect (Kirkpatrick et al., 1971), and inhibition of growth of Candida and Cryptococcus can be rescued by the addition of iron (Al-Sheikh, 2009; Lai et al., 2016). In Aspergillus fumigatus, the sequestration of Fe³⁺ by apo-Lf was shown to be important for host defense (Zarember et al., 2007) indicating that iron sequestration can play a role in the antifungal action of Lf in vivo. More recent studies now suggest that the main antifungal mechanism of Lf is iron-independent and occurs through a direct interaction of Lf with the fungal cell surface, leading to cell membrane damage and leakage. Supernatant protein assays and propidium iodide staining have shown that Lf alters cell surface permeability in C. albicans, C. krusei, and Cryptococcus neoformans, leading to cell death (Nikawa et al., 1993; Wakabayashi et al., 1996; Kondori et al., 2011). Alterations to the cell surface itself including protein leakage, the formation of surface blebs, swelling, and cell collapse have been seen in Candida isolates through scanning electron microscopy (Nikawa et al., 1993; Wakabayashi et al., 1996; Xu et al., 1998; van der Kraan et al., 2005b).

The cellular release of potassium ions, cytosolic acidification, changes in membrane potential, intracellular ROS accumulation, and chromatin condensation have all been detected in C. albicans following treatment with Lf indicating the induction of an apoptotic phenotype (Viejo-Díaz et al., 2004; Andres et al., 2008). The details by which Lf induces apoptosis-like processes in fungi remain largely unclear, however, recent studies have examined this mechanism in significant yeast species. In Saccharomyces cerevisiae, it was shown that hLf induces cell death in a mitochondrial and caspase-dependent way that is characterized by caspase activation, ROS accumulation and cytochrome c release (Acosta-Zaldivar et al., 2016). Work by the same group in C. albicans indicated that the proton-translocating ATPase Pma1p, which is a primary contributor to pH regulation in yeasts, was the target of hLf, inducing lethal mitochondrial dysfunction (Andres et al., 2016).

The combined use of Lf with other antifungal drugs has been increasingly studied in recent years. When combined with fluconazole, various sources of Lf have been observed to significantly enhance inhibitory activity and decrease the minimum inhibitory concentration (MIC) for several Candida species, including wild type and clinical strains, and ergosterol biosynthesis and azole-resistant mutants (Wakabayashi et al., 1998; Kuipers et al., 1999; Naidu et al., 2004; Venkatesh and Rong, 2008; Kobayashi et al., 2011). The mechanism of synergy between Lf and fluconazole is not yet well-understood and does not appear to result from an increased intracellular uptake of fluconazole, as the intracellular concentration of radiolabelled fluconazole was not affected by the addition of bLf (Kobayashi et al., 2011).

Several other azole drugs, including itraconazole, clotrimazole, and ketoconazole, as well as 5-fluorocytosine have been demonstrated to function synergistically with Lf in C. albicans (Wakabayashi et al., 1996; Kuipers et al., 1999). The polyene drugs amphotericin B and nystatin exhibited no interactions with bLf when tested in Candida species (Wakabayashi et al., 1996) although amphotericin B and bLf have been found to act synergistically in Cryptococcus neoformans, Cryptococcus gattii, and S. cerevisiae. The addition of exogenous iron did not rescue growth of Cryptococcus treated with amphotericin B and bLf, indicating that the mechanism of synergy is not solely due to iron chelation (Lai et al., 2016).

Lactoferricin (Lfcin) is found in the human gut as a natural breakdown product and was first generated by Bellamy et al. (1992) through pepsin hydrolysis of intact Lf. Bovine lactoferricin (bLfcin), comprising residues 17–41 from the N-terminal region of Lf, is simpler in structure but much more potent than human lactoferricin (hLfcin), which comprises residues 1–47. Both Lfcins possess an 18-residue loop region with a disulphide bridge and many positively charged and hydrophobic residues (Gifford et al., 2005).

The three dimensional structures of hLfcin and bLfcin differ from each other and from the homologous regions of the intact Lf protein, and it is thought these differences determine their different antifungal efficacies. While bLfcin forms a β-sheet confirmation in solution that contains a group of aligned hydrophobic residues well-suited to interactions with biological membranes, hLfcin forms a coiled structure in solution lacking the aligned residues and thus having weaker interactions with the target cell (Farnaud et al., 2004; Hunter et al., 2005; Alexander et al., 2012). Lfcin shows strongly enhanced antimicrobial activity compared to Lf and the antifungal efficacy of bLfcin, in particular, has been observed across a large range of various types of fungi (Table 1).

The primary antifungal mechanism of action of Lfcin appears to be through direct interaction with the fungal surface and disruption of the fungal membrane (Bellamy et al., 1993). Using confocal scanning laser microscopy, it was observed that bLfcin molecules were internalized by C. albicans cells within minutes (van der Kraan et al., 2005b) and that treatment with Lfcin resulted in dissipation of the proton gradient across the cell membrane (Gifford et al., 2005). In C. albicans and in the dermatophyte Trichophyton mentagrophytes, changes in ultrastructural features and an aggregation of cytoplasmic materials observed following treatment with Lfcin were traced to stimulation of ATP synthesis and extracellular secretion, resulting in pore formation in the plasma membrane (Bellamy et al., 1994; Ueta et al., 2001). Lfcin has also been shown to upregulate host defense via the induction of reactive oxygen species that stimulate the fungicidal activities of polymorphonuclear leukocytes (Kullberg et al., 1999). This ability to mount a combined assault on microbial pathogens is a common feature of natural host-derived products, most likely resulting from evolutionary pressures to maintain efficacy and prevent pathogens from developing resistance (Xu et al., 2009; Carter et al., 2016).

Synergy between bLfcin and azole antifungal drugs has been observed in C. albicans. In azole-resistant strains, the addition of low levels of bLfcin greatly increased inhibition by fluconazole and itraconazole and was seen to inhibit the growth of fungal hyphae (Wakabayashi et al., 1998). bLfcin has also been observed to function synergistically with clotrimazole and ketoconazole and it significantly reduced the formation of biofilms when used in combination with voriconazole and amphotericin B (Wakabayashi et al., 1996). The mechanistic basis of these synergistic actions remains unknown.

Lf and its major naturally cleaved peptide, Lfcin, have proven broad-spectrum antifungal action, with Lfcin exhibiting greater potency than the intact protein. The less well-characterized peptides including Lfampin and Lf(1–11) have also been seen to possess increased antifungal potency in Candida species but have not been tested extensively beyond this. Lf and Lfcin function synergistically primarily with azole antifungal drugs, however, knowledge of the spectrum and mechanism of these synergistic interactions is lacking, and synergy is largely unexplored for the other Lf-derived peptides.

Mechanistically, Lf and its peptides have been shown to function primarily through membrane destabilization, with immunomodulation and iron sequestration playing secondary roles. Although many aspects of Lf-derived peptides are quite conserved, differences in amino acid composition and tertiary structure contribute to their different functional properties. Novel alternative peptides that have been generated or modified from the original sequence of Lf provide insight into the relationship between structure and function. Lf and its derived peptides have great potential as leads for future antifungal drug development, however, considerable research is still required to understand the antifungal actions of these molecules as well as their potential to support current antifungals.

This article was written by KF and was reviewed and edited by DC.

Work on Lf in the Carter Lab has been supported by the Australian National Health and Medical Research Council (Grant Number APP1021267). KF is financially supported by an Australian Postgraduate Award.