Available online at www.sciencedirect.com
Biochimie 91 (2009) 133e140 www.elsevier.com/locate/biochi
Research paper
Bactericidal effect of lactoferrin and lactoferrin chimera against halophilic Vibrio parahaemolyticus Nidia Leon-Sicairos a,b,*, Adrian Canizalez-Roman b, Mireya de la Garza c, Magda Reyes-Lopez c, Jorge Zazueta-Beltran d, Kamran Nazmi f, Bruno Gomez-Gil e, Jan G. Bolscher f a
Departamento de Investigacio´n, Hospital Pedia´trico de Sinaloa, Blvd. Constitucio´n S/N, Culiaca´n, Sin. 80200, Me´xico b Unidad de Investigacio´n, Facultad de Medicina, Universidad Auto´noma de Sinaloa, Av. Cedros y Sauces, Fracc. Los Fresnos, Culiaca´n, Sin. 80246, Me´xico c Departamento de Biologı´a Celular, Centro de Investigacio´n y de Estudios Avanzados-IPN, Apdo. postal 14-740, Me´xico, D.F. 07000, Me´xico d Programa Regional para el Doctorado en Biotecnologı´a, Facultad de Ciencias Quı´mico-Biolo´gicas, Culiaca´n, Sin. 80040, Me´xico e CIAD, A.C. Mazatlan Unit for Aquaculture. AP. 711 Mazatlan, Sinaloa, Mexico f Department of Oral Biochemistry, Academic Centre for Dentistry Amsterdam (ACTA), Amsterdam, The Netherlands Received 15 May 2008; accepted 13 June 2008 Available online 26 June 2008
Abstract Infections caused by Vibrio parahaemolyticus, an halophilic member of the genus Vibrio, have increased globally in the last 5 years. Diarrhea caused by V. parahaemolyticus results from eating raw or undercooked seafood. The aim of this work was to investigate whether lactoferrin and some lactoferrin-peptides have bactericidal activity against Vibrio parahaemolyticus ATCC 17802, the pandemic strain O3:K6, and the multidrug resistant isolate 727, as well as against Vibrio cholerae strains O1 and non-O1. Whereas both peptides lactoferricin (17e30) and lactoferrampin (265e284) did not have bactericidal activity, 40 mM of lactoferrin chimera (a fusion of the two peptides) inhibited the growth of all Vibrio tested to the same extent as the antibiotic gentamicin. The cidal effect of LFchimera showed a clear concentration response in contrast to bovine lactoferrin which showed higher inhibition at 10 mM than at 40 mM. FITC-labeled LFchimera bound to the bacterial membranes. Moreover LFchimera permeabilized bacterial cells and membranes were seriously damaged. Finally, in experiments with the multidrug resistant isolate 727, sub-lethal doses of LFchimera strongly reduced the concentrations of ampicillin, gentamicin or kanamicin needed to reach more than 95% growth inhibition, suggesting synergistic effects. These data indicate that LFchimera is a potential candidate to combat the multidrug resistant pathogenic Vibrio species. Ó 2008 Elsevier Masson SAS. All rights reserved. Keywords: Lactoferrin; Lactoferricin; Lactoferrampin; LFchimera; Vibrio parahaemolyticus
1. Introduction Vibrio parahaemolyticus is a Gram-negative marine enteropathogenic bacterium commonly associated with seafoodborne illness. Outbreaks have appeared in Japan, Taiwan, India, Chile, Me´xico, and other coastal countries [1e4]. The * Corresponding author. Unidad de Investigacio´n, Facultad de Medicina, Universidad Auto´noma de Sinaloa, Av. Cedros y Sauces, Fracc. Los Fresnos, Culiaca´n, Sin. 80246, Me´xico. Tel.: þ52 667 227 8588. E-mail address:
[email protected] (N. Leon-Sicairos). 0300-9084/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.biochi.2008.06.009
organism is phylogenetically close to Vibrio cholerae, the causative agent of cholera. Toxigenic strains of V. cholerae belonging to either the O1 or O139 serogroups, have the ability to produce epidemics and pandemics [5,6]. Until recently, other Vibrio spp. including V. parahaemolyticus, have been associated with sporadic infections or localized outbreaks [7,8]. V. parahaemolyticus causes gastroenteritis linked with the consumption of contaminated raw or undercooked seafood. The symptoms include diarrhea with abdominal cramps, nausea, vomiting, headache and low-grade fever. The illness is self-limiting, of moderate severity and lasts an average of
134
N. Leon-Sicairos et al. / Biochimie 91 (2009) 133e140
3 days in immunocompetent patients [9]. V. parahaemolyticus has become a worldwide important etiologic agent of diarrhea, since the first pandemic infection in 1996 and their rapid spreading in Asia, Europe and America [2,10e16]. The most dreadful threat for human health is that multidrug resistant Vibrio spp have been recently isolated from the environment [11], in regions were seafood is a main component of diet. Therefore, new compounds are direly needed to control and combat Vibrio spp. Lactoferrin (LF) is an iron-binding protein involved in a large spectrum of biological properties including antimicrobial action [17]. Digestion of LF with pepsin generates the fragment lactoferricin (LFcin), which is more potent against bacteria compared with the native protein [18]. A second antimicrobial LF-derived peptide has been identified and designated lactoferrampin (LFampin) [19]. Both, LF and LF-peptides exert the bactericidal activity after cell membrane penetration and damage [18e20]. Therefore, LF and LF-peptides are a promising class of antimicrobial compounds to fight pathogenic microbes. Linking these two antimicrobial peptides into one molecule resulted in a strikingly stronger bactericidal activity than its constituent lactoferricin and lactoferrampin peptides [21]. Especially the finding that the bactericidal activity of this chimerical peptide was not hampered by repelling effects of increasing salt concentrations makes this component a potential candidate to combat the halophilic bacterium V. parahaemolyticus.
In the present study we have investigated the bactericidal activity of this promising chimera peptide in comparison with the constituent peptides and the native bovine LF. LFchimera displayed strong bactericidal activity against V. parahaemolyticus, including a multidrug-resistant strain, as well as against V. cholerae. Permeabilization and damaged of the plasma membrane, are suggested as the mechanism of action. Moreover sub-lethal doses of the LFchimera increased the sensitivity towards classical antibiotics in a multidrug resistant V. parahaemolyticus. 2. Materials and methods 2.1. Materials Bovine LF (20% iron saturated, was kindly donated by DMV International, USA). Synthetic peptides LFcin17e30, LFampin265e284 and LFchimera corresponding the bovine lactoferrin amino acid sequences were synthesized using F-moc chemistry as described before [21]. 2.2. Bacterial culture conditions The following V. parahaemolyticus strains were used: ATCC 17802, (without tdh toxin gene), O3:K6 (pandemic strain with tdh toxin gene), and multidrug resistant strain727.
Fig. 1. Inhibition of the growth of V. parahaemolyticus and Vibrio cholerae with lactoferrin and lactoferrin-peptides. A: ATCC 17802 culture (DO 0.005 at 660 nm, was incubated in LB medium with or without iron, and with gentamicin or bLF, LFampin265e284, LFcin17e30, or LFchimera all at final concentrations of 10 mg/ ml. Culture optical density was determined every 30 min at 660 nm and results after 2 h incubation are shown. B: Strain O3:K6 in the same conditions that ATCC 17802 strain. C. Vibrio cholerae O1 strain and D. Vibrio cholerae non-O1 strain. All experiments were made three times in triplicate. Percentage of viable cells was calculated relative to viable bacteria obtained in LB. The mean and deviation standard are indicated.
N. Leon-Sicairos et al. / Biochimie 91 (2009) 133e140
Besides Vibrio cholerae strains O1 and non-O1were used. Bacteria were maintained in LuriaeBertani medium (LB) (Difco, Becton Dickinson, USA) to which 3% NaCl was added, placed on a shaker and grown at 37 C for 16e18 h. To remove the iron LB medium was incubated with Chelex100 resin (5 g/l) at 4 C in constant agitation. After 16 h the resin was removed by filtration and the medium sterilized. All glass materials were treated with 6 M HCl to remove iron traces. 2.3. Growth inhibition in the presence of lactoferrin and lactoferrin-derived peptides To test the bactericidal activity of LF and LF peptides a bacteria culture with an OD660 nm of 0.005, in LB medium containing 0.2 M NaCl was used at the beginning of experiments. Bacteria were transferred in 1 ml of LB medium with or without iron (control of growth), and in LB medium containing bLF, LFcin17e30, LFampin265e284 and LFchimera at final concentrations of 1, 10, 20 and 40 mM. Cultures were incubated at 37 C with constant agitation. OD660 nm was measured every 30 min until 2 h of incubation. In similar experiments, the effect of temperature (10, 27, 37 and 42 C), and cations (Fe3þ and Mg2þ used at a final concentrations of 1, 2.5 and 5 mM) were tested. In all assays we have microscopically examined for agglutination and never found any indication of agglutination. Percentage of viable cells was determined in relation to control cultures in LB medium. All experiments were repeated at least three times in triplicate.
135
of 2 mM in PBS. This suspension was added in triplicate to the same volume of diluted LFchimera starting with 1 mM in PBS. After 1 h of incubation at 37 C, the microtiter plates were centrifuged and the absorbance in the supernatant was measured at 450 nm. Complete release of hemoglobulin was achieved by adding 1% of Tween 20. Also, permeabilization on erythrocytes treated with 1, 10, 20 and 40 mM of LFchimera during 4 h was tested by the PI exclusion as before has been indicated for V. parahaemolyticus. Furthermore, the effect of LFchimera on cultures of primary liver cells from Fischer 344 rats was tested by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)formazan method. Briefly, cultured rat hepatocytes were treated with 1, 20, 30 and 40 of LFchimera during 4 h and then were washed and incubated with 4 mg/ml MTT during 30 min. Supernatants where removed and the reaction was stopped with 200 mM of DMSO. Absorbance was measured at 540 nm.
2.4. Bacterial killing assay by measuring cell permeabilization V. parahaemolyticus strain O3:K6 (pandemic) culture in LB without iron was washed three times, set to a suspension of approximately 108 cells/ml and incubated with 40 mM LFchimera during 1.5 h at 37 C. After that, bacteria were washed again and incubated with the vital dye propidium iodide (PI) (10 mg/ml) for 10 min at 4 C. Then, bacterial suspension was washed five times and finally fixed and processed to quantify PI exclusion under FACScan (Fluorescence Associated Cell Scanner; Becton Dickinson, USA). Control experiments were carried out in bacterial cells either without the addition of peptide (viability control), with 0.005% Tween 20 (which disrupts lipids in bacterial membrane), or 5 mM ampicillin (which acts as a competitive inhibitor of the transpeptidase, needed for bacterial cell-wall synthesis). 2.5. Hemolytic, permeabilization and MTT assays The hemolytic activity of peptides was determined in essence as described before [23]. Human erythrocytes from donors who gave their informed consent were centrifuged at 75 g for 15 min, and washed three times with PBS. The cell suspension was normalized to a hemoglobulin value
Fig. 2. Concentration depending bactericidal effect of bLF and LFchimera. V. parahaemolyticus culture ATCC 17802 strain under conditions as in Fig. 1 were incubated with 10 mM of gentamicin, or with bLF or LFchimera, at final concentrations of 1, 10, 20, and 40 mM. Culture optical density was determined every 30 min at 660 nm and results at 2 h incubation are shown. All experiments were made three times in triplicate. Percentage of viable cells was calculated relative to viable bacteria obtained in LB. The mean and deviation standard are indicated.
136
N. Leon-Sicairos et al. / Biochimie 91 (2009) 133e140
2.6. Peptide binding or internalization analysis by confocal laser-scanning fluorescence microscopy V. parahaemolyticus (108 cells/ml) was incubated in LB without iron but with 1 mM FITC-labeled LFchimera for 30 min, with the purpose to observe whether this peptide is translocated over the bacterial membrane. Cells were centrifuged (5 min, 10,000 g), resuspended and fixed (4% paraformaldehyde, pH 7.4 at 37 C during 30 min), washed twice and processed to be analyzed under confocal microscopy. To search whether LFchimera is recognized by the bacterial membrane, bacterial cells were fixed, washed twice and incubated with 1 mM FITC-labeled LFchimera for 30 min. After that, bacteria were washed, processed and analyzed. All samples were visualized by using a confocal laser-scanning microscope (Leica, Heidelberg, Germany). 2.7. Electron microscopy V. parahaemolyticus O3:K6 (1.0 109 cells/ml) culture was incubated in LB without iron with 40 mM LFchimera for 1.5 h at 37 C. Cells were harvested, resuspended in PBS (pH 7.4) and fixed with 4% paraformaldehyde plus 0.5% glutaraldehyde. Subsequently, bacteria were rinsed with distilled water and mounted on bare 200-mesh copper grids. Next, phosphotungstic acid (1%, pH 5.5, 30 s) was added, replicas were dehydrated and studied with a transmission electron microscope (IEM2000Ex), operated at 100 kV. 2.8. Assay of synergic effect between LFchimera and antibiotics V. parahaemolyticus 727 (a multidrug resistant strain) was used to determine whether LFchimera is effective against this bacterium, and whether LFchimera synergically increases bactericidal effect of common antibiotics. First, in order to test the resistance level to common antibiotics, this strain was grown with or without gentamicin (2e100 mg/ml), kanamicin (2.5e20 mg/ml) or ampicilin (2.5 mge3.2 mg/ml); bactericidal activity of LFchimera (1, 5, 10, 20 and 40 mM) was tested in parallel. Synergism with ampicillin or gentamicin was tested by growing bacteria in medium with ampicillin (2.5e3200 mg/ ml), with kanamicin (2.5e20 mg/ml), or with gentamicin (2.5e10 mg/ml) in the presence of LFchimera (5 mM). The OD660 nm of culture was measured between 30 and 90 min. Percentage of viable cells was determined in relation to cultures without added peptides or antibiotics. All experiments were repeated at least twice in triplicate. 3. Results 3.1. Bactericidal effect of bLF and LFchimera in V. parahaemolyticus and V. cholerae strains The capacity of bLF, LFcin17e30, LFampin265e284 and LFchimera, to inhibit the growth of V. parahaemolyticus strains ATCC 17802 (Fig. 1A) and O3:K6 (Fig. 1B) was tested
in LB medium without iron. Both bacteria grow equally well in medium without iron as in medium with iron. LFchimera at a concentration of 10 mM effectuated more than 50% of growth inhibition. This was comparable to that found with the native bLF at equimolar concentrations. LFcin17e30 and LFampin265e284 induced only 10e15% or growth inhibition. Also a mixture of both LFcin17e30 and LFampin265e284 was not effective in inhibiting V. parahaemolyticus (data not shown). Gentamicin was used as a positive control, which exerted about 95% of growth inhibition. Both strains appeared to be comparable sensitive to bLF and the LFpeptides. Of the two strains of Vibrio cholerae the non-O1 serotype was more sensitive that the O1 serotype (Fig. 1C,D). A concentration of 10 mM of LFchimera inhibited around 60% of the growth of the O1 serotype of V. cholerae while more than 80% of the non-O1 serotype was inhibited. The same concentration of LF inhibited more than 90% of both cultures. Again, LFcin17e30 and LFampin265e284 had only weak bactericidal activity in both strains (Fig. 1C,D). Next, higher concentrations of LFchimera were tested to find out what concentration would equal the capacity of gentamicin. LFchimera showed a concentration-dependent effect and 40 mM was needed to reach a comparable inhibition value as gentamicin (Fig. 2B). These results indicate that LFchimera is potential candidate to combat infections caused by these Vibrio spp. surprisingly, higher concentrations of the native bLF resulted in an abolishing of the activity rather than an increase in the capacity to inhibit the growth, indicating that other mechanism of action are involved (Fig. 2A). Similar results were obtained with the V. parahaemolyticus strain O3:K6 (data not shown). 3.2. Effect of magnesium, iron, and temperature on the inhibition of the growth due to bLF and LFchimera LF and LFchimera were able to affect the V. parahaemolyticus growth in spite of the high salt concentrations present in medium of culture. To explore whether the antimicrobial activity is due to the interaction with the negatively charged bacterial surface, magnesium chloride was added to the culture medium. Indeed, magnesium chloride abolished the LFchimera-induced growth inhibition (Fig. 3A). To a lesser extent also the effect of bLF was decreased (Fig. 3A). In contrast the addition of Ferric iron only reduced the bLF-mediated growth inhibition and had a no effect on the bactericidal activity of LFchimera. This suggests again difference in mechanisms of action of bLF and LFchimera, the first one at least partially by sequestering the bacteria from iron and the latter one by affecting the bacterial membrane. Lowering the temperatures below 10 C reduced the inhibitory effect of both bLF and LFchimera (Fig. 3B). 3.3. Membrane permeabilization and damage due to LFchimera The effect of LFchimera on the integrity of the bacterial membrane was studied with the dye PI (that only penetrates
N. Leon-Sicairos et al. / Biochimie 91 (2009) 133e140
Fig. 3. The effect of magnesium, iron and temperature on the growth inhibition by bLF and LFchimera. A: V. parahaemolyticus culture under conditions as in Figure 1 were incubated with 10 mM gentamicin, bLF or LFchimera, with or without either 5 mM MgCl2 or 5 mM FeCl3. B: similar experiment with bLF or LFchimera incubated at 10 or 37 C. Culture optical density was determined every 30 min at 660 nm and results after 2 h incubation are shown. All experiments were made three times in triplicate. Percentage of viable cells was calculated relative to viable bacteria obtained in LB. The mean and deviation standard are indicated.
dead cells), quantified by flow cytometry. Fig. 4A shows that LFchimera permeabilized the bacterial membrane since virtually all cells had taken up the dye fluorescence, similar to treatment with the detergent Tween 20, or with ampicillin, which inhibits the cell-wall endopeptidase action. By using negative staining in electron microscopy, it was found that LFchimera caused severe damage to the cell membrane such as vesicles, protrusions and filamentation (Fig. 4C). These results suggest that LFchimera destabilize and permeabilize the bacterial membrane structure. Moreover, in experiments using confocal microscopy, we observed that FITC-LFchimera was mainly bound to bacterial membrane (Fig. 5). 3.4. Synergic effect between LFchimera and ampicillin or gentamicin, in a multidrug resistant strain V. parahaemolyticus 727 is a multidrug resistant strain which was isolated from the sea environment. In comparison with the ATCC 17802 strain, which is sensitive to antibiotics (Fig. 5A), this strain exhibits resistance to ampicillin kanamicin, and gentamicin (Fig. 6B), LFchimera was as active against the multidrug resistant strain 727 as against strain ATCC 17802 (Fig. 6) resulting in a more than 95% inhibition with 40 mM of the LFChimera. This strain was used to study the effect of LFchimera when used in combination with antibiotics. Concentrations of 2000 and 3200 mg/ml of ampicillin
137
Fig. 4. Membrane permeabilization and ultraestructural effect. A: membrane permeabilization of V. parahaemolyticus strain O3:K6. Bacteria (DO 0.005 at 660 nm) were incubated in medium LB without iron, with 40 mM of LFchimera, during 1.5 h at 37 C. As control membrane effectors 0.005% Tween 20 or 5 mg/ml ampicillin were used. Cells were washed and incubated with 10 mg/ml propidium Iodide (PI) for 10 min at 4 C, washed again and fixed. Samples were analyzed by flow cytometry. B, and C, electron microscopic photographs of V. parahaemolyticus (109 cells/ml) incubated in medium LB without iron during 1.5 h at 37 C. B: without treatment, C: with 40 mM LFchimera. After incubation, cells were washed and fixed with 4% paraformaldehyde and 4% glutaraldehyde for 30 min at 37 C. Samples were processed to be analyzed in an electron microscope. Bar 200 nm.
were necessary to inhibit 50 or 95% of the growth of V. parahaemolyticus 727 respectively; however the same percentage of inhibition was obtained with only 2.5 and 5 mg/ml of ampicillin if combined with 5 mM of LFchimera (Table 1). This concentration of 5 mM of LFchimera only resulted partially inhibition (
