Anti-complement effects of lactoferrin-derived peptides

Abstract

Lactoferrin is an important biological molecule with many functions such as modulation of the inflammatory response, iron metabolism and antimicrobial defense. One effect of lactoferrin is the inhibition of the classical complement pathway. This study reports that antimicrobial peptides derived from the N-terminal region from both human and bovine lactoferrin, lactoferricin H and lactoferricin B, respectively, inhibit the classical complement pathway. No inhibitory effect of these peptides was observed on the alternative complement pathway in an AP50 assay. However, lactoferricin B reduced the inhibitory properties of serum against Escherichia coli in a concentration dependent manner. These results suggest that the N-terminal region of lactoferrin is the important part in the inhibition of complement activation and that these peptides possess other important properties than their antimicrobial effect.

Antimicrobial peptides, Lactoferrin, Lactoferricin, Complement inhibition

1 Introduction

Antimicrobial peptides are important components of the host defense against invading pathogens. These peptides are part of the innate immune system and have been isolated in a vast array of organisms and show antimicrobial activity against most microorganisms [1]. In addition to their antimicrobial properties, several peptides have been shown to be involved in other processes of the immune system such as cytokine production, enhanced antibody response, chemotactic activity and modulation of inflammation [2].

The influence of antimicrobial peptides on the complement system has been established through studies with human neutrophil defensins (HNP) [3–5]. Prohaszka et al. [4] showed that binding of defensins to C1q lead to activation of the classical complement pathway, while van den Berg et al. [6] report that this binding caused an inhibition of the classical complement pathway. Other antimicrobial peptides like the pig neutrophil peptide protegrin-1 and tachyplesin-1 from horseshoe crab amebocytes have also been shown to bind to C1i–C1r–C1s in human serum, but the effect on the complement system by this binding has not been further studied [3].

Lactoferrin is a multifunctional glycoprotein found in milk and exocrine fluids such as saliva, tears, bile, and pancreatic fluid. It is also a major component of polymorphonuclear neutrophils [7]. Lactoferrin isolated from milk and tears have previously been described as an inhibitor of the classical complement pathway [8,9]. Further studies have shown that lactoferrin inhibits the formation of the classical C3 convertase, thus preventing the cleavage of C3 [10,11]. The inhibition of the complement can be reversed by addition of Fe³⁺, but not by Ca²⁺ or Mg²⁺[10].

Gastric pepsin treatment of lactoferrin generates antimicrobial peptides from the N-terminal region. From bovine lactoferrin a 25 amino acid peptide consisting of residues 17–41, lactoferricin B (Lfcin B), is generated, [12] while a 47 amino acid peptide, lactoferricin H (Lfcin H), consisting of residues 1–47 can be generated from human lactoferrin [13]. The peptides are 2–12-fold more active against Escherichia coli compared to their corresponding undigested lactoferrin [14]. Lfcin B is the most potent peptide of the two and has antimicrobial activity against a wide range of microorganisms, including Gram-positive and Gram-negative bacteria [15], viruses [16,17], protozoa [18,19], and fungi [20]. Recently, several anti-inflammatory activities of lactoferrin-derived peptides have been described, suggesting that they have several other properties in the defense against microbes in addition to their antimicrobial activity [21–24].

2 Materials and methods

2.1 Proteins and peptides

Lactoferricin B (Lfcin B) and lactoferricin H (Lfcin H) were prepared by pepsin digestion of bovine and human lactoferrin, respectively, at the Centre for Food Technology, Qld, Australia. Lactoferricin H 18–42 (Lfcin H 18–42) was synthesized by MedProbe, Oslo, Norway. Unsaturated bovine lactoferrin, cecropin P1 and magainin 2 were purchased from Sigma–Aldrich, USA. Bovine lactoferrin was saturated with iron as previously described by Davidson and Lonnerdal [25]. All compounds were dissolved in double distilled water and stored at −20 °C until use. All peptides and proteins had a purity of more than 95%.

2.2 Buffers

Gelatin veronal buffered saline (GVBS), pH 7.4, containing 3.1 mM barbituric acid, 1.8 mM sodium barbital, 0.8 mM MgCl₂, 0.15 M NaCl, 0.25 M CaCl₂, and 0.05% gelatin, was used in the classical complement assay. For the alternative complement assay GVBS, pH 7.4, supplemented with 10 mM EGTA and 5 mM MgCl₂ (Mg-EGTA GVBS), was used. All chemicals were purchased from Sigma–Aldrich, St. Louis, USA, except for barbituric acid and sodium barbital, which were obtained from Merck, Darmstadt, Germany.

2.3 Bacterial strain

Escherichia coli ATCC 25922 was used in the time-kill studies. The bacteria were stored at −70 °C until use. An overnight culture was grown in 2% bacto peptone water (BPW) (Difco, MI, USA), pH 6.8, at 37 °C to mid-logarithmic growth phase and diluted in 2% BPW to the desired final concentration (1.0×10⁶ CFU/ml) before use.

2.4 Human serum

Human sera were collected from normal healthy volunteers (2 male and 2 female) by venous puncture. The blood was collected in a vacutainer purchased from Becton–Dickinson, Plymouth, UK and allowed to coagulate for 30 min to 1 h at room temperature before centrifugation at 3000 rpm for 10 min to obtain serum. Inactivation of complement was achieved by incubating serum at 56 °C for 30 min. The prepared serum was pooled and either used fresh or used after storage in aliquots at −70 °C.

2.5 Determination of MIC

The MIC of the peptides against E. coli ATCC 25922 was determined using a standard broth microdilution technique as previously described [26]. Briefly, overnight cultures of bacteria were grown to mid-logarithmic growth phase and diluted in 2% BPW to give a final inoculum of 1×10⁶ CFU/ml. Serial dilutions of peptide (50 µl) and bacteria (50 µl) were added to a 96-well microtiter plate (Nunc, Roskilde, Denmark), and incubated at 37 °C overnight. The MIC was determined as the lowest peptide concentration where no visible growth occurred. The absorbance at 540 nm was also measured as a control. All assays were performed in parallel and performed at least three times.

2.6 Time-kill studies in human serum

Prepared human serum (180 µl) was mixed with different concentrations of peptide (10 µl) in a microtiter plate and incubated at 37 °C for 5 min. Bacteria (10 microliters) in mid-logarithmic phase were added to the wells to give a final inoculum of 1×10⁶ CFU/ml. The plate was incubated at 37 °C and aliquots of 10 µl were taken out at 15, 30 min, 1, 2, and 4 h and transferred to 0.9% saline water. Ten microliter from the dilutions was then plated on to PDM agar plates (Antibiotic Sensitivity Medium II, AB Biodisk, Sweden), and incubated overnight at 37 °C. Colonies were counted to calculate CFU/ml.

2.7 Hemolytic assays

For analysis of the classical pathway of complement activation (CH50 assay), sheep red blood cells (SRBC) were isolated from sheep blood obtained from Statens Serum Institut, Hillerød, Denmark, by centrifugation at 2000 rpm for 10 min at 4 °C. The SRBC were washed 3 times with GVBS. The washed SRBC (1×10⁹ cells/ml) were sensitized with antibodies by adding hemolysin (Dade Behring, Marburg, Germany) to a final dilution of 1:2000 and incubated at 37 °C for 30 min with intermittent agitation. The sensitized SRBC were washed once in GVBS and stored at 4 °C before use. Serial dilutions of peptide (50 µl) in GVBS and diluted serum (50 µl) were added to a round-bottomed microtiter plate (Nunc, Roskilde, Denmark) and incubated at 37 °C for 5 min before sensitized SRBC (50 µl) (final concentration 0.5×10⁷ cells/ml) were added. The plate was then incubated at 37 °C for 60 min with intermittent agitation. For positive control (100% lysis), distilled water was added to the sensitized SRBC, and for negative control (0% lysis), sensitized SRBC were incubated in GVBS. After incubation, the plate was centrifuged at 1000g for 3 min and 100 µl of the supernatant was transferred to an empty flat bottomed microtiter plate (Nunc, Roskilde, Denmark) and optical density (OD) was measured at 414 nm using a Spectramax PLUS (Molecular Devices, CA, USA).

For analysis of the alternative complement pathway (AP50 assay), rabbit erythrocytes (Dade Behring, Marburg, Germany) were washed 3 times in Mg-EGTA GVBS. Serial dilutions of peptide (50 µl) in Mg-EGTA GVBS and diluted serum (50 µl) were added to a round-bottomed microtiter plate and incubated at 37 °C for 5 min before rabbit erythrocytes (50 µl) (final concentration 0.5×10⁷ cells/ml) were added. The plate was then incubated at 37 °C for 60 min with intermittent agitation. For positive control (100% lysis) distilled water was added to rabbit erythrocytes and for negative control (0% lysis) rabbit erythrocytes were incubated in Mg-EGTA GVBS. After incubation the plate was centrifuged at 1000g for 3 min and 100 µl of the supernatant was transferred to an empty flat-bottomed microtiter plate and OD was measured at 414 nm.

Calculation of hemolytic activity (Z) was performed using the formula described by Roos et al. [27]. Z=−ln (1−((OD₄₁₄ (sample)−OD₄₁₄ (negative control))/(OD₄₁₄ (positive control)−OD₄₁₄ (negative control)))). For the classical complement assay serum was diluted 1:100 in GVBS and for the alternative complement assay 1:6 in Mg-EGTA GVBS. The results are presented as relative Z-values.

3 Results

3.1 Antibacterial activity

The antibacterial activity of the peptides and proteins against E. coli ATCC 25922 was determined to show the differences in activity between the different peptides and proteins and to determine the concentrations used in the time-kill assay in serum. Cecropin P1 and magainin 2, from nematodes [28] and the African frog Xenopus laevis[29], respectively, were used throughout the study to investigate if the observed effects of the lactoferrin-derived peptides was common for other antimicrobial peptides.

The MIC of the peptides and proteins against E. coli ATCC 25922 are presented in Table 1. Cecropin P1 was the most active substance with an MIC of 3 µg/ml. Lfcin B and magainin 2 had an MIC of 30 and 40 µg/ml, respectively, while Lfcin H, Lfcin H 18–42 and the different forms of bovine lactoferrin all had MIC values above 100 µg/ml.

Table 1

Amino acid sequence of the antimicrobial peptides and their minimal inhibitory concentration (MIC) against E. coli ATCC 25922

Peptide/protein	Amino acid sequence (single letter code)	MIC (µgml−1)
Lfcin B	F17 KCRRWQWRMKKLGAPSITCVRRAF41	30
Lfcin H	G1 RRRRSVQWCAVSQPEATKCFQWQRNMRKVRGPPVSCIKRDSPIQCI47	>100
Lfcin H 18–42	T18 KCFQWQRNMRKVRGPPVSCIKRDS42	>100
Cecropin P1	SWLSKTAKKLENSAKKRISEGIAIAIQGGPR	3
Magainin 2	GIGKFLHSAKKFGKAFVGEIMNS	40
Lf B		>100
Lf B Fe		>100

Underlined C in the amino acid sequence indicates the cysteines involved in the disulfide bridge.

3.2 Hemolytical assays

Two different assays were used to determine an effect of the antimicrobial peptides on the complement. The CH50 assay was used for the classical pathway, while the AP50 assay was used to determine an effect on the alternative pathway. To exclude hemolysis by the peptides themselves, they were tested against the red blood cells without complement present. None of the peptides showed hemolytic activity against red blood cells in the concentrations used, however cecropin P1 at a concentration of 150 µg/ml was not used due to some intrinsic hemolytic activity against rabbit erythrocytes. (Results not shown.)

Two of the lactoferrin-derived peptides, Lfcin B and Lfcin H, inhibited the classical complement pathway in a concentration dependent manner (Fig. 1A). The bovine peptide was somewhat more effective than the human peptide. The inhibitory effect of the peptides was not as strong as for unsaturated bovine lactoferrin (Lf B). Lf B saturated with iron showed no inhibitory activity. In contrast to Lfcin H, a shorter derivative of human lactoferrin (Lf H), Lfcin H 18–42, exerted no significant effect on the classical complement pathway. Cecropin P1 and magainin 2 showed no effect on the classical complement pathway.

Figure 1

Anti-complement activity of antimicrobial peptides. (A) CH50 assay, classical complement pathway, (B) AP50 assay, alternative complement pathway. The results are presented as relative Z. Bars indicate ±SEM. For the CH50 assay serum was diluted 1:100 in GVBS and for the AP50 assay 1:6 in Mg-EGTA GVBS.

Neither Lfcin B, Lfcin H nor Lf B showed an effect on the alternative pathway. However, Lfcin H 18–42, cecropin P1 and magainin 2 at high concentrations seemed to enhance the effect of the alternative complement pathway (Fig. 1B).

3.3 Time-kill studies in human serum with active and heat-inactivated complement

To assess if the inhibitory effect of the antimicrobial peptides on the classical complement pathway had an effect on the inhibition of bacteria by pooled human serum, E. coli was incubated in serum with active or heat-inactivated complement in the presence of peptide. Concentrations of 15 µg/ml (0.5× MIC) to 300 µg/ml (10× MIC) of Lfcin B inhibited the effect of serum on E. coli in a concentration dependent manner (Fig. 2A). After 4 h the CFU/ml in serum without peptide is reduced by 73% (9.5×10⁵–2.5×10⁵ CFU/ml). However in the presence of 300 µg/ml of Lfcin B, the CFU/ml after 4 h are increased by 96% (from 1.0×10⁶ to 2.76×10⁷ CFU/ml, respectively) (Fig. 2A). In contrast to Lfcin B, neither Lfcin H nor Lfcin H 18–42 did have an effect on the inhibitory properties of serum against E. coli (data not shown). None of the lactoferrin-derived peptides had an effect on E.coli in heat-inactivated serum (Fig. 2B).

Figure 2

Time-kill curves of E. coli incubated in pooled human serum in the presence of lactoferrin-derived peptides. (A) Serum with active complement and different concentrations of Lfcin B, (B) serum with heat-inactivated complement and Lfcin B, Lfcin H and Lfcin H 18–42. Bars indicate ±SEM.

In contrast to the lactoferrin-derived peptides magainin 2 and cecropin P1 showed an effect on E.coli in serum with active complement (Fig. 3A and B). After 4 h incubation with concentrations corresponding to 10× the MIC no bacteria were detected (Fig. 3A and B). Like the lactoferrin-derived peptides neither magainin 2 nor cecropin P1 at high concentrations showed any antimicrobial effect on E. coli in serum with heat-inactivated complement (Fig. 3C).

Figure 3

Time-kill curves of E. coli incubated in human serum in the presence of two none lactoferrin-related peptides, cecropin P1 and magainin 2. (A) Serum with active complement and different concentrations of cecropin P1, (B) serum with active complement and different concentrations of magainin 2, (C) serum with heat-inactivated complement and cecropin P1 and magainin 2. Bars indicate ±SEM.

4 Discussion

The complement system is an important part of the host defense against invading microorganisms. It consists of at least 30 proteins in plasma and on cell surfaces that work in a cascade, where the activation of one component results in the activation of the next. The activation cascade results in the formation of a membrane attack complex that creates pores in the plasma membrane of the target cell [30,31].

Lactoferrin is involved in numerous biological functions like modulation of the inflammatory response, iron metabolism, recruitment of polymorphonuclear leukocytes and antimicrobial defense [7]. Human lactoferrin from tears inhibits activation of the classical complement pathway through inhibition of formation of the classical C3 convertase [10,11]. In this study we have shown that N-terminal peptides derived from both human and bovine lactoferrin have anti-complement properties.

The mode of interference by the peptides on the complement is not known, but since lactoferrin inhibits the classical pathway through inhibition of formation of the C3 convertase [10,11] the peptides might act in the same manner. Other antimicrobial cationic peptides, like human neutrophil defensins, protegrin-1 from pig neutrophils and tachyplesin-1 from horseshoe crab amebocytes have been shown to bind to C1i–C1r–C1s in normal human serum [3]. Like the lactoferrin-derived peptides, the defensins also inhibits the classical complement pathway but not the alternative complement pathway [5] and a similar mechanism of inhibition is possible.

Since Lfcin B inhibited the effect of serum on E. coli, the lectin pathway and the activation of the alternative pathway might also have been inhibited. Activation of the lectin pathway leads to formation of the same C3 convertase as the classical pathway [32]. An interaction of the peptides with components involved in both the classical and the lectin pathway i.e. C2 or C4 may explain this. It has previously been shown that Lfcin B can bind lipopolysaccharide (LPS) from Gram-negative bacteria and that it can inhibit LPS induced cytokine response in human monocytic cells [24,26,33]. LPS is an activator of the alternative pathway and binding of LPS by Lfcin B may explain why there is no effect seen by the alternative complement pathway on E. coli in serum.

Like Lfcin B, Lfcin H inhibited the classical complement pathway, however no inhibitory activity of the effect of serum on E. coli was observed using Lfcin H. The reason for this remains unknown, however Lfcin H might be more susceptible to degradation or protein bound than Lfcin B in serum. This would have a more significant effect in the time-kill assay where the sera are almost undiluted compared to the CH50 assay where the sera are diluted 1:100. Lfcin B might also interact directly with E. coli rendering the bacteria, less susceptible to the effect of the complement.

Since the peptides studied here are derived from the N-terminal part of their respective lactoferrin, the N-terminal seems to be important in the anti-complement effect of lactoferrin. The anti-complement activity of lactoferrin is lost when the protein is iron saturated but the N-terminal region of lactoferrin is equally exposed in both unsaturated and iron saturated lactoferrin suggesting that the N-terminal is not important. However, Baker et al. [34] propose that the N-terminal region of unsaturated lactoferrin is more flexible which may explain the effect of iron saturation.

None of the lactoferrin-derived peptides themselves showed any antimicrobial activity in serum with active or heat-inactivated complement (Fig. 2B). Cecropin P1 and magainin 2 at high concentrations showed activity in serum but no activity was observed in serum with heat-inactivated complement (Fig. 3). This suggests that components in serum inactivate the antimicrobial activity of the peptides. It has previously been reported that peptides loose their antimicrobial activity in heat-inactivated serum [35,36]. It has been hypothesized that blocking or inhibitory factors in serum are activated during heat treatment or that potentiating factors in serum are inactivated [35]. It is also possible that Lfcin B and Lfcin H are still bound to the complement protein(s) they inhibit in heat-inactivated serum and thereby losing the antimicrobial activity.

In conclusion these results show that the N-terminal derived peptides from lactoferrin are important in the inhibitory effect on the classical complement pathway and that the peptides alone can exert the same activity as the protein itself in terms of anti-complement activity. In addition Lfcin B can inhibit the activity of complement in serum. The exact mechanism on how the lactoferrin-derived peptides inhibits the complement remains to be elucidated and should be investigated in the future along with experiments on the effect on the lectin pathway. However, these findings show that lactoferrin in addition to having antimicrobial activity in vivo through iron-sequestering also has properties located in the N-terminal of the protein that can suppress the inflammatory effects caused by bacteria. The lactoferrin-derived peptides can be useful templates for generation of new anti-inflammatory agents.

Acknowledgements

We appreciate the help from Tom Eirik Mollnes for constructive discussion and manuscript preparations. This research has been supported financially by the Norwegian Research Council and Alpharma AS, Oslo, Norway.