Purification and characterization of tinamou egg white ovotransferrin as an antimicrobial agent against foodborne pathogenic bacteria

Food Research International 54 (2013) 1836–1842

Contents lists available at ScienceDirect

Food Research International journal homepage: www.elsevier.com/locate/foodres

Purification and characterization of tinamou egg white ovotransferrin as an antimicrobial agent against foodborne pathogenic bacteria Orly Varon a, Kevin J. Allen a, Darin C. Bennett b, Lili R. Mesak a, Christine H. Scaman a,⁎ a b

Food, Nutrition and Health Program, Faculty of Land Food Systems, The University of British Columbia, 2205 East Mall, V6T 1Z4, Vancouver, BC, Canada Avian Research Centre, Faculty of Land Food Systems, The University of British Columbia, 2205 East Mall, V6T 1Z4, Vancouver, BC, Canada

a r t i c l e

i n f o

Article history: Received 6 October 2012 Received in revised form 19 February 2013 Accepted 22 February 2013 Keywords: Chilean tinamou Egg white Ovotransferrin Antimicrobial Escherichia coli O157:H7 Staphylococcus aureus

a b s t r a c t Chilean tinamou (Nothoprocta perdicaria) egg white ovotransferrin was purified with anion-exchange Fast Protein Liquid Chromatography and identified by peptide mass fingerprinting. Tinamou egg white ovotransferrin was compared to chicken (Gallus gallus) and emu (Dromaius novaehollandiae) ovotransferrin by SDS-PAGE analysis. Though all three species contained similar molecular weight ovotransferrins, tinamou egg white contained 20% more ovotransferrin than chicken, but 30% less than emu. The effectiveness of tinamou ovotransferrin as a natural antimicrobial agent was compared to chicken ovotransferrin against two food related pathogens, Escherichia coli O157:H7 and Staphylococcus aureus. The bacteriostatic and bactericidal activities of the native, apo and holo forms of the proteins were dependent on the presence of 50 mM bicarbonate and protein iron saturation. Native ovotransferrins were the most effective at a concentration of 10 mg/ml with 50 mM bicarbonate, exhibiting bacteriostatic and bactericidal activities against both pathogens. Holo ovotransferrins exhibited moderate antimicrobial activity in the presence of bicarbonate. Different antimicrobial activities were observed between chicken and tinamou ovotransferrins, suggesting tinamou ovotransferrin may possess unique antimicrobial motifs that warrant further investigation of its amino acids composition. Further, our results suggest that tinamou ovotransferrin may represent a natural antimicrobial agent suitable for use in food matrices or on food preparation surfaces. © 2013 Elsevier Ltd. All rights reserved.

1. Introduction Ovotransferrin, a monomeric 78 kDa glycoprotein, comprises 12% of the chicken egg white and is one of the major egg antimicrobial components. Chicken ovotransferrin is reported to have antibacterial properties against a variety of bacteria (Oratore, D'Andrea, D'Alessandro, Moreton, & Williams, 1990; Tranter & Board, 1982), including two foodborne pathogens, Escherichia coli O157:H7 and Staphylococcus aureus. However, contradictory results have been reported in studies examining ovotransferrin antimicrobial activity against various strains of E. coli and S. aureus. Some studies suggest that these bacteria are ovotransferrin sensitive, while others indicate that they are ovotransferrin resistant (Ibrahim, Iwamori, Sugimoto, & Aoki, 1998; Ko, Mendonca, & Ahn, 2008; Ko, Mendonca, Ismail, & Ahn, 2009; Valenti et al., 1980, 1983). Additionally, the antimicrobial mechanism of ovotransferrin is not clear. Ovotransferrin can reversibly bind two Fe2+ ions per molecule in the presence of bicarbonate (Schalabach & Bates, 1975). It is suggested that to serve as an antimicrobial agent, ovotransferrin should be in the apo, iron-free form which inhibits bacterial growth by restricting iron availability through chelation,

⁎ Corresponding author. Tel.: +1 604 822 1804; fax: +1 604 822 5143. E-mail address: [email protected] (C.H. Scaman). 0963-9969/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodres.2013.02.041

forming iron-saturated ovotransferrin (i.e. holo ovotransferrin) — a process first recognized by Alderton, Ward, and Fevold (1946). However, other studies suggest that ovotransferrin is antimicrobial regardless of the iron saturation state, with antimicrobial effects deriving from bacterial membrane disruption (Ibrahim, Sugimoto, & Aoki, 2000; Ibrahim et al., 1998; Valenti, Visca, Antonini, Orsi, & Antonini, 1987). Ovotransferrin antimicrobial activity against various foodborne pathogens emphasizes its importance as a natural food antimicrobial agent. Due to increasing consumer interest, the use of natural antimicrobials has become popular in the food industry. These may be more acceptable to the public than synthetic compounds, eco-friendly, medically acceptable, and economical to manufacture. Due to consumer consumption of chicken products, most research regarding antimicrobial properties of eggs has focused on chicken eggs; therefore it is of interest to find antimicrobial compounds from other avian sources, such as the Chilean tinamou (Nothoprocta perdicaria). Tinamous are ground dwelling birds endemic to South America. Recently, phylogenetic analysis has placed the tinamous within the ratites, as a sister group to emus (Dromaius novaehollandiae) (Harshmann, Braun, & Braun, 2008). An interesting characteristic of the fresh Chilean tinamou egg white is its distinctive pink hue. It is known that ovotransferrin is pink when iron saturated (Alderton et al., 1946), therefore we hypothesized that tinamou egg white may contain high amounts of iron saturated ovotransferrin.

O. Varon et al. / Food Research International 54 (2013) 1836–1842

This work determined ovotransferrin content of tinamou egg white compared to chicken and emu egg white. For the first time, tinamou ovotransferrin was purified by anion exchange FPLC and identified by mass spectrometry. Additionally, this work evaluated the antimicrobial activity of native (as naturally accrues in egg white), apo, and holo forms of tinamou and chicken ovotransferrin against E. coli O157:H7 and S. aureus in the absence and presence of bicarbonate. Bacteriostatic and bactericidal activities were determined by turbidity and viability assays, respectively. Both pathogens were susceptible to tinamou ovotransferrin, and some differences were observed in the activity of the tinamou protein compared to its chicken counterpart. The bacteriostatic activity of both chicken and tinamou holo ovotransferrins in the presence of bicarbonate suggest that there is an additional synergist interaction between these two components that is not related to iron complexation. 2. Materials and methods 2.1. Ovotransferrin preparation 2.1.1. Native ovotransferrin Freeze dried chicken ovotransferrin was purchased from SigmaAldrich (St. Louis, Missouri, USA) and a concentration of 20 mg/ml was prepared in sterile, deionized water. Four to six day old Chilean tinamou eggs were obtained from a local producer in British Columbia, Canada. Egg whites were separated from yolks and stored at −20 °C, defrosted prior to use at room temperature, and mixed for 30 min at medium velocity with a magnetic stirring plate. Ovotransferrin was isolated from egg whites by anion exchange FPLC according to the method of Awadé and Efstathiou (1999). Diluted egg whites, with ovomucin removed, were fractioned on a HiTrap Q HP 5/1 column in 500 μl aliquots or on a HR Source 15Q 16/10 (both from GE Healthcare Bio-Sciences, Piscataway, New Jersey, USA) in 50 ml aliquots. Ovotransferrin fractions collected from FPLC were concentrated and desalted using a 3 kDa molecular weight cut-off centrifugal filter device (Amicon Ultra-15, EMD Millipore, Billerica, Massachusetts, USA). Concentrated fractions were freeze-dried and analyzed for purity by SDS-PAGE. All ovotransferrin solutions were sterilized by passing through a 0.22 μm syringe filter (EMD Millipore, Billerica, Massachusetts, USA) for microbial studies. 2.1.2. Apo and holo ovotransferrins To obtain the apo form of the protein, native ovotransferrins were dialyzed against 0.1 M citrate pH 4.5 containing 0.1% EDTA for 24 h at 4 °C, then against an excess of deionized water for two days. An immediate color change to salmon pink corresponding to a distinct peak at ~475 nm was obtained after iron saturation. Holo ovotransferrin was prepared from tinamou and chicken ovotransferrin, respectively, using a 10 molar excess of ferric ion to ovotransferrin. Ovotransferrins were dialyzed against excess 0.1 M FeCl3:6H2O and 50 mM NaHCO3 at 4 °C and then against an excess of deionized water for two days (Ibrahim et al., 1998). The apo and holo ovotransferrins were freeze-dried and stored at −20 °C. 2.2. Ovotransferrin characterization

1837

Bio-Sciences, Piscataway, New Jersey, USA). The molecular weight of each protein band was calculated by the relative mobility of individual bands with respect to the molecular weight marker. This was defined by plotting migration distances versus the logarithm of protein molecular mass (log Mr). The ovotransferrin content of egg whites was calculated by measuring band optical density relative to the optical density for the lane, which was proportional to protein concentration. 2.2.2. Ovotransferrin identification by mass spectrometry Concentrated FPLC fractions prepared as described in Section 2.1.1 were deglycosylated by Peptide:N-glycosidase F (PNGase-F, New England Biolabs Inc., Ipswich, Massachusetts, USA) prior to mass spectrometry. Determination glycosylation extent was made by comparing relative migration of treated and untreated samples on 12.5% SDS-PAGE gels. Major deglycosylated bands were cut out and sent to the Proteomics Core Facility, Michael Smith Labs, UBC, for partial proteomic analysis. Gel slices were digested with trypsin using a Millipore Montage Zip Plate. Peptide mass fingerprinting was performed by ESI MS/MS (PE SCIEX API 300 Triple Quad) with protein identification made using the MASCOT database. An ion score (IS) was assigned to each peptide, and was calculated as 10 ∗ Log(P) where P is the probability that the observed match is a random event. Individual ion scores > 49 indicate identity or extensive homology (p b 0.05). 2.3. Bacterial strains Since we were only able to purify a limited quantity of tinamou ovotransferrin, antimicrobial activity against one Gram negative and one Gram positive foodborne pathogenic bacteria was evaluated, and included the Gram positive and negative reference strains E. coli O157:H7 EDL933 and methicillin resistant S. aureus COL, respectively. Both were kindly provided by Dr. Julian Davies (Department of Microbiology and Immunology, University of British Columbia). From frozen glycerol stocks (−80 °C), both pathogens were cultured on Luria Bertani (LB) agar (Fisher Scientific, Lenexa, Kansas, USA) for 24 h at 37 °C. Single colonies were inoculated into 3 ml of Brain Heart Infusion (BHI) broth (Fisher Scientific, Lenexa, Kansas, USA) and incubated for 24 h at 37 °C, reaching concentrations of 109 CFU/ml. After incubation, samples were diluted to 106 CFU/ml with fresh BHI broth, and serial dilutions of 10−2–105 made. One hundred microliters of each dilution was spread-plated in duplicate on LB agar, incubated as above, and colony forming units were enumerated. 2.4. Determination of ovotransferrins bacteriostatic activity by turbidity assay The bacteriostatic activity of ovotransferrins against E. coli O157: H7 and S. aureus COL in BHI with or without 50 mM NaHCO3 was Table 1 Sample preparation for the turbidity assay used to determine the bacteriostatic activity of chicken and tinamou ovotransferrins against Escherichia coli O157:H7 and Staphylococcus aureus. Sample properties

2.2.1. Ovotransferrin characterization by SDS-PAGE Concentrated FPLC fractions prepared as described in Section 2.1.1 and aliquots of 20 μl of diluted egg white samples were subjected to SDS-PAGE. Samples were mixed with 20 μl sample buffer and boiled for 5 min; after centrifugation, a 10 μl aliquot of each sample supernatant was loaded into a well and electrophoresed in a 12.5% acrylamide gel at 150 V for 40 min (Laemmli, 1970). The gel was stained with 0.1% Coomassie brilliant blue solution for 1 h and destained twice (45% methanol, 10% acetic acid). Protein molecular mass markers (Fermentas UAB, Vilnius, Lithuania) were used for molecular weight determination using Image Quant TL software (GE Healthcare

Sample

OTF species

OTF type

OTF (mg/ml)

NaHCO3 (mM)

1 2 3 4 5 6 7 8

Chic, Chic, Chic, Chic, Chic, Chic, Chic, Chic,

Apo, Apo, Nat Apo, Nat Apo, Apo, Apo,

0 0 1 10 1 10 5 5

0 50 0 0 50 50 0 50

Tin Tin Tin Tin Tin Tin Tin Tin

holo, nat holo, nat holo, nat holo, nat holo holo

OTF = ovotransferrin; Chic = chicken; Tin = tinamou; nat = native; DW = deionized water. Each test well was inoculated with 5 μl of culture.

1838

O. Varon et al. / Food Research International 54 (2013) 1836–1842

Fig. 1. SDS-PAGE of duplicates of tinamou (T1, T2), chicken (C1, C2) and emu (E1, E2) egg white samples. Twenty micrograms of protein was loaded in each lane. Molecular weights of bands in the rectangles were calculated by relative mobility and identified by comparing to the literature as (A) ovomucin, (B) 150 kDa protein, (C) ovotransferrin, (D) ovoglobulin, (E) ovalbumin, (F) ovomucoid and (G) lysozyme.

determined by measuring the optical density (OD) at 595 nm with a microplate reader (Bio-Rad labs, Richmond, California, USA) at 0 and 24 h incubation at 37 °C in BHI. Each test sample was prepared directly in a microplate well in a total volume of 200 μl, as presented in Table 1. The percent inhibition of E. coli O157:H7 and S. aureus COL imparted by ovotransferrin exposure with or without NaHCO3 was calculated as follows:  % inhibitation ¼ 100  1−

 ðOD Test Sample−OD Negative ControlÞ : ðOD Positive Control−OD Negative ControlÞ

OD Negative control was the optical density of the sample at 0 h. The OD Positive control was the optical density value of sample containing media only (BHI), or media with 50 mM NaHCO3, as appropriate.

2.5. Determination of ovotransferrin bactericidal activity by viability assay The bactericidal activity was determined by spread plating each sample from the turbidity assay after further dilution with 0.1% peptone water (Difco, Mississauga, Ontario, Canada). In wells where obvious growth of bacteria was indicated, 10-fold dilutions of 10 −5 to 10 −8 for E. coli O157:H7 and 10 −4 to 10 −7 for S. aureus were made. In wells where bacteria growth was low or not obvious, 10-fold dilutions of 10 −2 to 10 −5 were made. From each dilution, 100 μl was spread on duplicate LB plates and colonies were enumerated after incubation for 24 h at 37 °C. 2.6. Statistics For each experiment, three replicates of each of two independent cultures were evaluated. One way ANOVA and Student t-tests were

Fig. 2. Anion-exchange FPLC of egg white proteins. A 500 μl aliquot of 10-fold diluted egg white (0.02 M Tris–HCl, pH 9) was separated using a HiTrap Q HP 5/1 column. Fractions A–G were collected and subjected to SDS-PAGE to confirm identity (data not presented) of the most abundant proteins in each peak as: (A) lysozyme, (B) ovotransferrin and ovomucoid, (C) ovotransferrin, ovalbumin and lysozyme, (D–F) ovoglobulin and ovalbumin, and (G) ovoglobulin.

O. Varon et al. / Food Research International 54 (2013) 1836–1842

1839

The purification of tinamou egg white ovotransferrin was carried out as a small scale proteomics analysis. The egg white proteins were initially separated by a two-step procedure. The first step separated proteins to major fractions by anion-exchange FPLC (Fig. 2), followed by separation of the major FPLC fractions by SDS-PAGE and deglycosylation (Fig. 3). SDS-PAGE of PNGase-F treated tinamou ovotransferrin band showed a 5 kDa reduction in molecular weight, confirming that it is an N-linked glycoprotein. Finally, tentative deglycosylated ovotransferrin band (lane C +, Fig. 3) was excised from the gel and digested with trypsin. Peptide mass fingerprinting, performed by ESI MS/MS with a MASCOT database search for protein identification, confirmed the band identity as ovotransferrin (Fig. 4). 3.2. Antimicrobial activity of native ovotransferrins

Fig. 3. 12.5% SDS-PAGE of PNGase-F treated fraction C before (−) and after (+) deglycosylation. Deglycosylated band (C+) was cut out of the gel and sent for mass spectral analysis, confirming the identity of ovotransferrin as shown in Fig. 4.

performed using Microsoft Office Excel (2007) while Minitab 15 (State College, PA) was used for 3- and 4-way ANOVA. Differences were considered significant at p b 0.05 and compared by Tukey's honest significance test. Mean values and standard deviations were reported.

Both chicken and tinamou ovotransferrins at 10 mg/ml combined with 50 mM NaHCO3 were bacteriostatic and bactericidal against E. coli O157:H7 (Table 2). Tinamou ovotransferrin without 50 mM NaHCO3 exhibited minor bacteriostatic activity against E. coli O157: H7; however, this was greater than the effect of native chicken ovotransferrin. Native ovotransferrins were more bacteriostatic against S. aureus compared to E. coli O157:H7 at 1 mg/ml with 50 mM NaHCO3 (Table 2). Under these conditions, chicken ovotransferrin was significantly more bacteriostatic and bactericidal (p b 0.05) against S. aureus than the tinamou protein, though the overall percent inhibition was moderate. The most effective antimicrobial treatment was 10 mg/ml native tinamou ovotransferrin with 50 mM NaHCO3 which resulted in 95 ± 1% inhibition of S. aureus and at least a 9 log reduction. While the 10 mg/ml chicken ovotransferrin with 50 mM NaHCO3 resulted in only 61% inhibition, it was also a highly effective bactericidal treatment. 3.3. Antimicrobial activity of apo and holo ovotransferrins

3. Results 3.1. Ovotransferrin characterization Whole chicken, tinamou and emu egg whites were characterized by SDS-PAGE (Fig. 1) and proteins were identified by comparing to molecular weight markers and to the literature. The tinamou egg white profile contained seven major protein bands, whereas chicken egg white contained three and emu egg white two. The comparison between egg white ovotransferrin contents was made using the SDS-PAGE analysis software. The average optical densities measured were 0.32, 0.42 and 0.51 OD for chicken, tinamou and emu ovotransferrins, respectively. Based on optical density ratios, tinamou egg white contained approximately 30% more ovotransferrin than chicken. Emu egg white contained approximately 20% more ovotransferrin than tinamou and approximately 60% more than chicken.

Native ovotransferrins were chemically manipulated to the apo and holo forms. An immediate color change to salmon pink corresponding to a distinct peak at 475 nm for chicken ovotransferrin and 473 nm for tinamou ovotransferrin was obtained after iron saturation (data not shown). There was no statistical difference in the bacteriostatic activity of 5 and 10 mg/ml apo or holo ovotransferrin with or without 50 mM NaHCO3 (p = 0.116); therefore, only the results for 5 mg/ml ovotransferrin are presented. Ovotransferrin from both species combined with 50 mM NaHCO3 inhibited bacterial growth more than ovotransferrin alone, with apo ovotransferrins being more effective (Table 3). Apo ovotransferrin with 50 mM NaHCO3 was also bactericidal against E. coli O157:H7, but less effective than the native ovotransferrins. There was an interaction between ovotransferrin species and bicarbonate, and iron saturation state and bicarbonate (Table 4), but was similar for the apo and holo forms of both species.

Fig. 4. Sequence homology of Chilean tinamou egg white ovotransferrin with chicken egg white ovotransferrin. The deglycosylated protein (Band C+, Fig. 3) was subjected to mass spectrometry using ESI MS/MS and fingerprinting using the MASCOT database. The underlined peptides are homologous with chicken protein sequences, confirming the identity of ovotransferrin. PS = protein scores.

1840

O. Varon et al. / Food Research International 54 (2013) 1836–1842

Table 2 Antimicrobial activity of native chicken and tinamou ovotransferrins against Escherichia coli O157:H7 and Staphylococcus aureus. Treatment

Escherichia coli O157:H7

Staphylococcus aureus

Bacteriostatic activity (% growth inhibition)

1 mg OTF/ml + 50 mM NaHCO3 10 mg OTF/ml + 50 mM NaHCO3 1 mg OTF/ml 10 mg OTF/ml

Bactericidal activity (log reduction CFU/ml)

Bacteriostatic activity (% growth inhibition)

Chicken

Tinamou

Chicken

Tinamou

Chicken

0b 95 ± 1a 0b⁎ 0b⁎

0d 96 ± 1a 8 ± 2c⁎ 16 ± 2b⁎

1 9 0 0

1 9 0 0

68 61 3 3

± ± ± ±

5a⁎ 3a⁎ 1b 1b

Bactericidal activity (log reduction CFU/ml)

Tinamou 14 95 2 2

± ± ± ±

2b⁎ 1a⁎ 1c 1c

Chicken

Tinamou

3* 9 0 0

1 9 0 0

OTF = ovotransferrin; different superscript letters within a column denote significance treatment differences (p b 0.05). Asterisks denote significance differences between chicken and tinamou ovotransferrins (p b 0.05).

4. Discussion 4.1. Ovotransferrin identification and purification The SDS-PAGE profile (Fig. 1) of tinamou, chicken and emu egg whites revealed an 86 kDa band for both chicken and tinamou, and a 92 kDa band for emu. Though the molecular weights are higher than previously reported molecular weight of chicken ovotransferrin, these proteins were tentatively identified as ovotransferrin; For example, Awadé and Efstathiou (1999) determined that the second major FPLC peak of their sample was chicken ovotransferrin and Maehashi (2010) determined that emu egg white SDS-PAGE band of a 78 kDa was ovotransferrin. The SDS-PAGE profile (Fig. 1) was also analyzed to determine the relative optical densities of each of the proteins bands. This analysis demonstrated that tinamou egg white contained approximately 30% more ovotransferrin than chicken. Emu egg white contained approximately 21% more ovotransferrin than tinamou and approximately 60% more than chicken. Chicken ovotransferrin, was found to comprise 12% of the entire egg white, as described in previous reports. Feeney, Anderson, Azari, Bennett, and Rhodes (1960) determined by absorption of the ovotransferrin iron complex that emu egg white contains 10% ovotransferrin. Takeuchi and Nagashima (2010) compared chicken egg white and emu egg white using a 12.5% Coomassie stained SDS-PAGE gel, producing a thicker and darker band for the emu ovotransferrin. Maehashi (2010) did a similar comparison using a 15% gel. They were able to detect a 78 kDa band as the most abundant protein in emu egg white, comprising 33% of the total emu egg white proteins. This band identified as ovotransferrin due to its high N-terminal sequence similarity to that of chicken egg white. Additionally, they determined that emu ovotransferrin had the same molecular mass as chicken ovotransferrin.

Table 3 Apo and holo chicken and tinamou ovotransferrins' antimicrobial activity against Escherichia coli O157:H7. Treatment

Apo OTF

Holo OTF

5 mg OTF/ml + 50 mM NaHCO3 5 mg OTF/ml 5 mg OTF/ml + 50 mM NaHCO3 5 mg OTF/ml

Bacteriostatic activity (% growth inhibition)

Bactericidal activity (log reduction CFU/ml)

Chicken

Tinamou

Chicken

Tinamou

58 ± 8

50 ± 5

3*

1

2±4 25 ± 8

7±4 15 ± 4

0 0

0 0

0

0

0

0

OTF = ovotransferrin. There was an interaction between ovotransferrin species and presence of bicarbonate, and iron saturation state and presence of bicarbonate (Table 4), but overall, the bacteriostatic activity was similar between species and treatments.

Chilean tinamou egg white ovotransferrin quantity has not been previously reported. The observations that tinamou egg white contained more ovotransferrin than chicken egg white, and that chicken and tinamou egg whites contain similar amounts of iron, 0.59 ± 0.16 and 0.43 ± 0.02 (μg iron/ml egg white), respectively, as determined in a previous assay performed by our group (Scaman, unpublished data), contradicted the hypothesis that the fresh Chilean tinamou egg white is pink due to high quantities of iron saturated ovotransferrin. Since iron quantities of chicken and tinamou egg whites were similar, and ovotransferrin quantities were higher for tinamou, Chilean tinamou egg white does not contain a high quantity of iron-saturated ovotransferrin resulting in the pink hue. More research is required in order to ascertain why fresh Chilean tinamou egg white is pink. The next step for ovotransferrin characterization was separating chicken and tinamou egg whites with anion-exchange FPLC. According to Fig. 2, it appeared that peaks B, C, and D contained ovotransferrin. All fractions were subjected to deglycosylation, confirming that the protein was glycosylated, similar to chicken ovotransferrin. The deglycosylated band was confirmed as ovotransferrin by MS. Peptide mass fingerprinting was performed by ESI MS/MS with a MASCOT database search for protein identification as chicken ovotransferrin, as seen in Fig. 4. Though mass spectrometry analysis confirmed ovotransferrin identify, disappointingly, only 9% of the total sequence corresponding to region with an unknown biological function was obtained. At the time this assay was conducted (2011) the emu ovotransferrin amino acid sequence was not available in the MASCOT database. By comparing the recently sequenced emu ovotransferrin (Maehashi et al., 2012) it can be observed that the tinamou ovotransferrin sequence obtained was highly homologous with emu ovotransferrin. A possible next step could be to obtain additional tinamou ovotransferrin amino acid sequence data, in order to determine which species the tinamou protein is more similar to. This could be done via cDNA analysis and sequencing. Knowledge of the tinamou ovotransferrin amino acid sequence will demonstrate whether there are differences between tinamou and chicken ovotransferrins, which can result in functionality differences; for example whether there are differences in the antimicrobial cationic regions in tinamou ovotransferrin compared to chicken ovotransferrin.

Table 4 Three-way analysis of variance for apo and holo chicken and tinamou ovotransferrins' bacteriostatic activity against Escherichia coli O157:H7.

Species (chicken/tinamou) OTF type (apo/holo) 50 mM NaHCO3 (+/−) Species*OTF type Species*NaHCO3 OTF type*NaHCO3 Error OTF = Ovotransferrin.

Degrees of freedom

p-Value

1 1 1 1 1 1 41

0.048 0.000 0.000 0.243 0.000 0.000 –

O. Varon et al. / Food Research International 54 (2013) 1836–1842

4.2. Antimicrobial activity of native ovotransferrins The results of the turbidity and viability assays indicated that both tinamou and chicken native ovotransferrins were most effective against E. coli O157:H7 and S. aureus at a concentration of 10 mg/ml with 50 mM NaHCO3. Under these conditions, both tinamou and chicken ovotransferrins significantly inhibited and moreover, inactivated the bacterial population. In the absence of bicarbonate, native tinamou ovotransferrins exhibited minor bacteriostatic activity against E. coli O157:H7, while chicken ovotransferrins had no bacteriostatic activity at all. It is possible that tinamou ovotransferrin possesses different amino acid compositions conferring unique antibacterial motifs responsible for this difference. It should be noted that 50 mM NaHCO3 alone had no bacteriostatic effect on either E. coli or S. aureus (data not shown). Tinamou ovotransferrin exhibited bacteriostatic and bactericidal effects against S. aureus that was concentration dependent, and was more bacteriostatic than chicken ovotransferrin at 10 mg/ml with 50 mM NaHCO3. Despite the difference between the growth inhibition of S. aureus with chicken and tinamou ovotransferrins, both treatments resulted in at least a 9-log CFU/ml reduction. The chicken ovotransferrin antimicrobial activity against S. aureus may be more time-dependent than tinamou ovotransferrin; therefore future assays should involve longer incubation times and more frequent OD measurements. As well, the limitations of the bacteriostatic and bactericidal assays may cause inconsistent results between the two assays. The bacteriostatic turbidity assay is less time consuming and more cost efficient, but does not distinguish between live and un-lysed-dead bacteria; therefore it is possible that a treatment was lethal but cell lysis failed to occur, resulting in high absorbance values. Alternatively, the bactericidal activity may not readily detect viable but not culturable cells, though whether this phenomenon is induced by ovotransferrin was not investigated in our study. The iron deprivation mechanism as the mode of antimicrobial activity of ovotransferrin has been controversial for both species of bacteria. Ibrahim et al. (1998, 2000) determined that a 92 amino acid ovotransferrin peptide had bactericidal activity against E. coli K-12, without NaHCO3. OTAP-92, a short cationic sequence located in the N-terminal domain of ovotransferrin, can directly interact with the outer membranes of microorganisms. This sequence showed high similarity to a peptide region in insect defensins, which are active against Gram positive and negative bacteria by blocking voltage-dependent K + channels. There are studies suggesting some S. aureus strains are resistant to chicken ovotransferrin. Valenti et al. (1980) reported that 50 strains of S. aureus were resistant to native ovotransferrin combined with 50 mM NaHCO3. They argued that S. aureus has an efficient iron transport system that can overcome iron deficient conditions. This was determined by observing a similar bacterial growth rate in a media containing ovotransferrin with a medium where iron was precipitated with chromium salts. Different conclusions regarding S. aureus sensitivity to ovotransferrin may result from different researchers using different S. aureus strains, as S. aureus pathogenesis is known to be influenced significantly by iron uptake and metabolism mechanisms. Though Kadurugamuwa, Anwar, Brown, Shand, and Ward (1987) suggested that the iron requirement of S. aureus is relatively low, iron limitation has been studied as an antimicrobial mode of action against S. aureus. For example, some studies have investigated the antimicrobial activity of human transferrin iron binding against S. aureus (Lin, Mason, Woodworth, & Brandts, 1994; Lindsay, Riley, & Mee, 1995; Schade & Caroline, 1944). These studies determined that S. aureus can extract iron bound to human transferrin, which has a stronger affinity for iron than ovotransferrin. Additionally, Friedman et al. (2006) showed that S. aureus is able to alter its gene expression upon changes in iron sources. Further, they demonstrated that regulated overproduction of acidic end-products brought on by iron starvation decreases local pH resulting in the release of iron from

1841

transferrin. Overall, it seems that more research has to be done in order to determine the iron metabolism of S. aureus in the presence of ovotransferrin, as a first step towards understanding how the antimicrobial mechanism of ovotransferrin is related to its iron binding capacity. 4.3. Antimicrobial activity of apo and holo ovotransferrins To further investigate the mechanism of ovotransferrin antimicrobial activity and the role of bicarbonate, apo and holo ovotransferrins, with and without NaHCO3, were evaluated against E. coli O157:H7. Since the native ovotransferrins showed dramatically different effects at 1 mg/ml (no inhibition) and 10 mg/ml (95% inhibition), 5 and 10 mg/ml apo and holo ovotransferrin were evaluated. Due to limitations in the amount of tinamou ovotransferrin available, S. aureus was not used in these experiments. Based on the “iron-binding” mechanism, only apo ovotransferrin with 50 mM NaHCO3 was expected to exhibit bacteriostatic or bactericidal effects. No significant difference was observed between 5 and 10 mg/ml ovotransferrin under all conditions, suggesting that ovotransferrin antimicrobial activity may reach maximum effect at 5 mg/ml. Therefore, the antimicrobial mechanism might involve other modes of action apart from iron binding. The iron content of the BHI media used was reported to be 0.44 μg/ml (Ahn, Park, Oh, Sun, & Shin, 2004). According to Valenti, Antonini, Rossi Fanelli, Orsi, and Antonini (1982), ovotransferrin iron binding capacity is 1 μg iron per 40 mg ovotransferrin. A simple calculation shows that in order to bind all iron in the media the ovotransferrin concentration should be approximately 17 mg/ml. Therefore, removal of all available iron is likely not the reason for the same inhibition being exhibited by the 5 and 10 mg/ml ovotransferrin treatments. The antimicrobial activity of ovotransferrins and bicarbonate has often been associated with the apo protein, with the bicarbonate serving as a bridging ligand between ovotransferrin and iron, promoting iron sequestration (Aisen, 1980). As expected, the bacteriostatic activity against E. coli O157:H7 was bicarbonate-dependent for both apo ovotransferrins. Chicken and tinamou apo ovotransferrins at concentrations of 5 and 10 mg/ml with 50 mM NaHCO3 reduced the bacterial population by 3 or 1 log (CFU/ml), respectively. Similar results were obtained by Ko et al. (2008), who demonstrated that 20 mg/ml apo ovotransferrin with 50 mM NaHCO3 reduced E. coli O157:H7 growth by 1 log (CFU/ml). Those results were different from our previous assay with native ovotransferrins, where ovotransferrins at 10 mg/ml with 50 mM NaHCO3 reduced the bacteria population by at least 9 log (CFU/ml). It is not clear why the antimicrobial activity of apo ovotransferrins differed so dramatically from native ovotransferrins. One explanation may be that the chemical manipulations used to obtain the apo and holo forms caused structural changes that reduced the antimicrobial activity. Ovotransferrin antimicrobial activity is affected by its structure and may be altered by denaturation (Fraenkel-Conrat, 1950). Another speculation is that the E. coli O157:H7 cultures were in different growth stages when native ovotransferrins were evaluated, compared to the holo and apo forms. In the native ovotransferrin experiment, stationary phase E. coli O157:H7 were diluted 1000-fold with BHI, and 5 μl immediately added to each test well. Cells in stationary phase are in a σS-induced state, rendering cells more tolerant to stress (Hengge-Aronis, 2000). With the apo and holo ovotransferrins, after 1000-fold dilution with BHI, 1 h passed before adding 5 μl to each test well, potentially allowing cells to switch to an adaptive or early exponential phase, both of which are associated with lower levels of σ S. In the stationary phase, E. coli O157:H7 was expected to be more resistant to the iron depleted conditions, since it produces iron-stress response factors, such as siderophores; however, this does not coincide with our observations. Interestingly, holo ovotransferrins were also inhibitory in the presence of 50 mM NaHCO3, which suggests that bicarbonate contributes to the antimicrobial activity by a mechanism other than as a bridging

1842

O. Varon et al. / Food Research International 54 (2013) 1836–1842

ligand between ovotransferrin and iron, such as membrane disruption. While Ko et al. (2008) demonstrated that holo ovotransferrin had little or no inhibitory activity against E. coli O157:H7, it has been shown that E. coli possess a variety of iron stress-response mechanisms and therefore may not be sensitive to an iron deprivation state (Earhart, 1996; O'Brien & Gibson, 1970). Recently, Dorschner, Lopez-Garcia, and Peschel (2006) showed that bicarbonate alone can change cell wall structure and gene expression of E. coli, without any change in growth rate. These changes caused E. coli to be more susceptible to antimicrobial peptides. A global alteration in gene expression was seen with more than 300 gene transcripts altered greater than two-fold. Among the genes with significantly decreased expression was the global regulator fliA, which was confirmed by quantitative RT-PCR analysis. Therefore, the presence of bicarbonate in this work may have increased the susceptibility of the E. coli to the holo (and apo) ovotransferrins. 5. Conclusions Characterization of egg white proteins by SDS-PAGE demonstrated that Chilean tinamou and chicken egg whites are most similar in terms of protein content and composition, even though phylogenetically, tinamou and emu are both ratites, and classify remotely to chicken. Proteomic analysis demonstrated that ovotransferrin was the common protein among all three species. Though all egg whites contained ovotransferrins with similar molecular weights, ratite species were found to contain the highest concentrations. It is possible that ovotransferrin has an essential antimicrobial function as one of the major avian egg white proteins, and therefore its presence is conserved among distinct species. Chicken and tinamou egg white ovotransferrins were bacteriostatic and bactericidal activities against E. coli O157:H7 and S. aureus at concentrations of 5 mg/ml or higher in the presence of bicarbonate. Native ovotransferrins were the most effective, followed by apo, and holo forms of the protein. Holo ovotransferrins exhibited moderate antimicrobial activity in the presence of bicarbonate; therefore it is possible that bicarbonate contributes to the antimicrobial activity of ovotransferrin by a mechanism other than as a bridging ligand between ovotransferrin and iron, such as membrane disruption caused by a protein conformational change or by modulating the bacteria stress-related gene expression, increasing susceptibility to ovotransferrin. Since all ovotransferrin types exhibited some level of antimicrobial activity, it seems likely that more than one mechanism contributes to the observed effects. In the absence of bicarbonate, tinamou ovotransferrin exhibited minor antimicrobial activity, while chicken ovotransferrin had no antimicrobial activity. It is possible that tinamou ovotransferrin may possess unique antibacterial motifs compared to the chicken protein due to differences in the amino acid composition; however, this remains to be verified. Overall, tinamou ovotransferrin shows potential as a natural antimicrobial agent for use in food matrices or on food preparation surfaces and should be further investigated. Acknowledgments This research was supported by funds from the BC Ministry of Agriculture and Lands (funds administered by the Specialty Birds Research Committee). We thank Mr. Lorne Stobbe (Chilliwack, British Columbia) and Dwayne Harder (TryHarder Farms, Denholm, Saskatchewan) for the tinamou and emu eggs, respectively. We also thank Camille Bertolo for technical assistance. References Ahn, Y. J., Park, S. K., Oh, J. W., Sun, H. Y., & Shin, S. H. (2004). Bacterial growth in amniotic fluid is dependent on the iron availability and the activity of bacterial iron-uptake system. Journal of Korean Medicine Science, 19, 333–340.

Aisen, P. (1980). Iron transport and storage proteins. Annual Review of Biochemistry, 49, 357–393. Alderton, G., Ward, W. H., & Fevold, H. L. (1946). Identification of the bacteria-inhibiting iron-binding protein of egg white as conalbumin. Archives of Biochemistry, 11, 9–13. Awadé, A. C., & Efstathiou, T. (1999). Comparison of three liquid chromatographic methods for egg-white protein analysis. Journal of Chromatography B, 723, 69–74. Dorschner, R. A., Lopez-Garcia, B., & Peschel, A. (2006). The mammalian ionic environment dictates microbial susceptibility to antimicrobial defense peptides. Journal of the Federation of American Societies for Experimental Biology, 20, 35–42. Earhart, C. F. (1996). Uptake and metabolism of iron and molybdenum. In F. C. Neidhardt, R. CurtissIII, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, & H. E. Umbarger (Eds.), Escherichia coli and Salmonella typhimurium: Cellular and molecular biology (pp. 1075–1090). ASM Press. Feeney, R. E., Anderson, J. S., Azari, P. R., Bennett, N., & Rhodes, M. B. (1960). The comparative biochemistry of avian egg white proteins. The Journal of Biological Chemistry, 235, 2307. Fraenkel-Conrat, H. (1950). Comparison of the iron-binding activities of ovotransferrin and of hydroxylamido proteins. Archives of Biochemistry, 28, 452–463. Friedman, D. B., Stauff, D. L., Pishchany, G., Whitwell, C. W., Torres, V. J., & Skaar, E. P. (2006). Staphylococcus aureus redirects central metabolism to increase iron availability. PLoS Pathogens, 2, 10. Harshmann, J., Braun, E. L., & Braun, M. J. (2008). Phylogenomic evidence for multiple losses of flight in ratite birds. Proceedings of the National Academy of Sciences of the United States of America, 105, 13462–13467. Hengge-Aronis, R. (2000). The general stress response in Escherichia coli. In G. Storz, & R. Hengge-Aronis (Eds.), Bacterial stress responses (pp. 161–178). ASM Press. Ibrahim, H. R., Iwamori, E., Sugimoto, Y., & Aoki, T. (1998). Identification of a distinct antibacterial domain within the N-lobe of ovotransferrin. Biochimica et Biophysica Acta (BBA) — General subjects, 1401, 289–303. Ibrahim, H. R., Sugimoto, Y., & Aoki, T. (2000). Ovotransferrin antimicrobial peptide (OTAP-92) kills bacteria through a membrane damage mechanism. Biochimica et Biophysica Acta (BBA) — General subjects, 1523, 196–205. Kadurugamuwa, J. L., Anwar, H., Brown, M. R., Shand, G. H., & Ward, K. H. (1987). Media for study of growth kinetics and envelope properties of iron-deprived bacteria. Journal of Clinical Microbiology, 25, 849–855. Ko, K. Y., Mendonca, A. F., & Ahn, D. U. (2008). Influence of zinc, sodium bicarbonate and citric acid on the antibacterial activity of ovotransferrin against Escherichia coli O157:H7 and Listeria monocytogenes in model systems and ham. Poultry Science, 87, 2660–2670. Ko, K. Y., Mendonca, A. F., Ismail, H., & Ahn, D. U. (2009). Ethylenediaminetetraacetate and lysozyme improves antimicrobial activities of ovotransferrin against Escherichia coli O157:H7. Poultry Science, 88, 406–414. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227, 680–685. Lin, L. N., Mason, A. B., Woodworth, R. C., & Brandts, J. F. (1994). Calorimetric studies of serum transferrin and ovotransferrin. Estimates of domain interactions, and study of the kinetic complexities of ferric ion binding. Biochemistry, 33, 1881–1888. Lindsay, J. A., Riley, T. V., & Mee, B. J. (1995). Staphylococcus aureus but not Staphylococcus epidermidis can acquire iron from transferrin. Microbiology, 141, 197–203. Maehashi, K. (2010). Primary structure of potential allergenic proteins in emu (Dromaius novaehollandiae) egg white. Journal of Agriculture and Food Chemistry, 58, 12530–12536. Maehashi, K., Matano, M., Irisawa, T., Uchino, M., Kashiwagi, Y., & Watanabe, T. (2012). Molecular characterization of goose and chicken-type lysozyme in emu (Dromaius novaehollandiae): Evidence for extremely low lysozyme levels in emu egg white. Gene, 492, 244–249. O'Brien, I. G., & Gibson, F. (1970). The structure of enterochelin and related 2,3dihydroxybenzoylserine conjugates from Escherichia coli. Biochimica et Biophysica Acta (BBA) — General subjects, 215, 393–402. Oratore, A., D'Andrea, G., D'Alessandro, A. M., Moreton, K., & Williams, J. (1990). Binding and iron delivering of monoferric ovotransferrins to chick-embryo red blood cells (CERBC). Biochemistry International, 22, 111–118. Schade, A. L., & Caroline, L. (1944). Raw hen egg white and the role of iron in growth inhibition of Shigella dysenteriae, Staphylococcus aureus, Escherichia coli and Saccharomyces cerevisiae. Science, 100, 14–15. Schalabach, M. R., & Bates, G. W. (1975). The synergistic binding of anions and Fe3+ by transferrin. The Journal of Biological Chemistry, 250, 2182–2188. Takeuchi, J., & Nagashima, T. (2010). Chemical and physical characterization of Dromaius novaehollandiae (emu) eggs. Food Science & Technology Research, 16, 149–156. Tranter, H. S., & Board, R. G. (1982). The influence of incubation temperature and pH on the antimicrobial properties of hen egg albumen. The Journal of Applied Bacteriology, 56, 53–61. Valenti, P., Antonin, G., Von Hunolstein, C., Visca, P., Orsi, N., & Antonini, E. (1983). Studies of the antimicrobial activity of ovotransferrin. International Journal of Tissue Reactions, 1, 97–105. Valenti, P., Antonini, G., Rossi Fanelli, M. R., Orsi, N., & Antonini, E. (1982). Antibacterial activity of matrix-bound ovotransferrin. Antimicrobial Agents and Chemotherapy, 21, 840–841. Valenti, P., Stasio, A. D., Seganti, L., Mastromarino, P., Sinibaldi, L., & Orsi, N. (1980). Capacity of Staphylococci to grow in the presence of ovotransferrin or CrCl3 as a character of potential pathogenicity. Journal of Clinical Microbiology, 11, 445–447. Valenti, P., Visca, P., Antonini, G., Orsi, N., & Antonini, E. (1987). The effect of saturation with Zn+2 and other metal ions on the antibacterial activity of ovotransferrin. Medical Microbiology & Immunology, 176, 123–130.