Activity of hydrolysed lactoferrin against foodborne pathogenic bacteria in growth media: the effect of EDTA
P.M. Davidson, Department of Food Science and Technology, PO Box 1071, University of Tennessee, Knoxville, TN 37901–1071, USA (e-mail: firstname.lastname@example.org).
Lactoferrin was hydrolysed with pepsin and the antimicrobial activity of the resulting hydrolysate (HLF) was studied in 1% peptone, 0·05% yeast extract, 1% glucose (PYG) medium and tryptic soy broth (TSB). HLF was effective against Listeria monocytogenes, enterohaemorrhagic Escherichia coli and Salmonella enteritidis in PYG, however, the highest studied concentration (1·6 mg ml−1) did not inhibit growth of any of these organisms in TSB. The addition of EDTA enhanced the activity of HLF in TSB, indicating that the decreased activity of HLF may have been due, in part, to excess cations in the medium.
Antimicrobial substances have been isolated and characterized from a variety of sources including plants, animals and animal products, and micro-organisms. However, many of these antimicrobials have a limited antimicrobial spectrum and are active only at very high concentrations. An example of such a substance is lactoferrin, an iron-binding protein that is found in milk and colostrum, as well as in various other secretions ( Masson et al. 1969 ). The antimicrobial activity of lactoferrin has been attributed solely to its ability to bind essential iron and thus slow growth of micro-organisms ( Law and Reiter 1977). Lactoferrin is generally only bacteriostatic even at high concentrations. Recently, however, potent antimicrobial peptides have been generated by the peptic hydrolysis of lactoferrin ( Tomita et al. 1991 ). Hydrolysed lactoferrin was found to be nine to 25 times more effective than lactoferrin against several Gram-negative and Gram-positive micro-organisms. In addition, the hydrolysate was found to be bactericidal ( Tomita et al. 1991 ), whereas lactoferrin is only bacteriostatic. The active peptides responsible for the antimicrobial activity of the lactoferrin hydrolysate were subsequently identified and sequenced by Bellamy et al. (1992a) and were named lactoferricin H (from human milk) and lactoferricin B (from bovine milk). The antimicrobial spectrum of lactoferricin is wide ( Bellamy et al. 1992b ; Jones et al. 1994 ; Shin et al. 1998 ), inhibiting both Gram-positive and -negative bacteria (including several foodborne pathogens), yeasts, moulds, and even parasites ( Tanaka et al. 1995 ). In addition, the concentrations of lactoferricin necessary to cause inhibition are low, ranging from 0·5 to 500 mg ml−1 for both Gram-positive and Gram-negative bacteria ( Bellamy et al. 1992b ; Jones et al. 1994 ). Lactoferricin has also been shown to be active over a wide pH range ( Bellamy et al. 1992b ) as well as being resistant to heat; for example, lactoferricin hydrolysate maintained activity after autoclaving at 121 °C for 15 min ( Tomita et al. 1991 ). However, the activity of lactoferricin decreased in the presence of cations and salts ( Bellamy et al. 1992b ; Jones et al. 1994 ), so, despite its strong antimicrobial capabilities, it may not be useful as an antimicrobial in foods. In addition, the antimicrobial activity of hydrolysed lactoferrin or lactoferricin has been studied primarily in 1% peptone or in 1% peptone−0·05% yeast extract−1% glucose (PYG) ( Bellamy et al. 1992b ; Jones et al. 1994 ; Shin et al. 1998 ).
The objective of this study was to determine the antimicrobial activity of hydrolysed lactoferrin in tryptic soy broth against foodborne pathogenic bacteria including enterohaemorrhagic Escherichia coli, Salmonella enteritidis, and Listeria monocytogenes. Although lactoferricin was found to be more potent than the lactoferrin hydrolysate, we chose to study the hydrolysate because there are a number of purification steps necessary to obtain lactoferricin. In addition, other researchers ( Dionysius and Milne 1997) have isolated two additional antibacterial peptides other than lactoferricin from hydrolysed lactoferrin. Finally, if there is potential for use of either hydrolysed lactoferrin or lactoferricin as food additives, it is more likely that a less purified form would be more acceptable to the industry and regulatory agencies ( Sofos et al. 1998 ).
MATERIALS and METHODS
Escherichia coli O157:H7 ATCC 43895, E. coli O104:H21 MPHL 94–56815, Salmonella enteritidis ATCC 13076, Listeria monocytogenes Scott A and L. monocytogenes ATCC 19115 were obtained from the University of Idaho Department of Food Science and Toxicology culture collection. All cultures were maintained on trypticase soy agar (TSA) slants and transferred each month to maintain viability. Working cultures were obtained by inoculating a loopful of culture into 10 ml tryptic soy broth (TSB) and incubating at the optimal temperature for each strain for 18–24 h.
Lactoferrin was hydrolysed according to the procedure described by Tomita et al. (1991) . Bovine lactoferrin (17% iron saturated, ICN Biochemicals, Costa Mesa, CA, USA) was dissolved in distilled water to a concentration of 50 mg ml−1. The pH was adjusted to 3·0 with 1·0 mol l−1 HCl. Pepsin (activity: 10 600 NF U mg−1; Sigma, St Louis, MO, USA) was added to a final concentration of 3% (weight per weight of substrate). The mixture was incubated at 37 °C for 4 h. The reaction was terminated by heating at 80 °C in a water bath for 20 min. The pH of the solution was readjusted to 7·0 with 1·0 mol l−1 NaOH. Insoluble solids were removed by centrifugation at 15 000 g for 30 min. To confirm complete hydrolysis, sodium dodecyl sulphate-PAGE was performed using 16·5% polyacrylamide gel. A premade molecular weight marker (Life Technologies,Rockville, MD, USA) was used to determine molecular weight. The supernatant was lyophilized and stored in a freezer before use.
A checkerboard micro-assay ( Barry 1976) was used to study antimicrobial activity of hydrolysed lactoferrin (HLF) alone and in combination with EDTA. HLF was dissolved in de-ionized water and filter sterilized while disodium EDTA (Sigma) was dissolved in distilled water and sterilized by autoclave (121 °C, 15 min). Further dilutions were made in sterile tryptic soy broth (TSB, Difco, Detroit, MI, USA) or 1% peptone, 0·05% yeast extract, 1% glucose (PYG) media. HLF was studied in concentrations of 0·05, 0·1, 0·2, 0·4, 0·8 and 1·6 mg ml−1. When EDTA was combined with HLF, it was added in final concentrations of 0·05, 0·1, 0·2, 0·4 and 0·5 mg ml−1 against Gram-negative strains, and in concentrations of 0·05 and 0·1 mg ml−1 against Gram-positive strains.
Microtitre wells had a capacity of 300 µl. In the first assay, where HLF activity was studied alone in PYG, each well contained 120 µl of diluted HLF and 120 µl of diluted inoculum (4·0 log cfu ml−1). In the second assay, where HLF activity was studied in combination with EDTA in TSB, each well was inoculated with 60 µl of HLF, 60 µl of EDTA, and 120 µl of inoculum diluted in TSB (4·0 log cfu ml−1). Plates were incubated at 37 °C and absorbance at 620 nm was recorded at 24 and 48 h (Titertek Multiscan MC, Labsystems, Helsinki, Finland). Minimum inhibition concentrations (MIC) at 24 and 48 h were defined as the lowest concentration of antimicrobial that completely inhibited growth. Growth inhibition was defined as the concentration in which absorbance of the test wells minus the absorbance of the control wells was ≤ 0·05. To determine the minimum bactericidal concentrations (MBC), 100 µl was taken from all wells where no growth was detected, spread on TSA and incubated at optimum temperature for 48 h. MBCs were designated as wells where there was ≥ 99·9% loss of viable cells ( Barry 1976).
In PYG broth, HLF inhibited growth of all of the foodborne pathogens tested for up to 48 h and was bactericidal against all of the strains tested except for E. coli O104:H21 ( Table 1). HLF inhibited L. monocytogenes strains at 0·2 mg ml−1. However, L. monocytogenes did not grow well in PYG broth and in the second assay growth could not be detected even in the positive control wells, thus preventing MIC determinations. Plating samples from these wells showed that there were still viable cells in the media after 48 h.
Table 1. Minimum inhibitory concentrations and minimum bactericidal concentrations of hydrolysed lactoferrin in PYG broth at 37 °C *
|Escherichia coli O157:H7||0·4||0·4||0·8|
|Escherichia coli O104:H21||0·4||0·8||> 1·6|
|Salmonella enteritidis 13076||0·4||0·4||0·8|
|Listeria monocytogenes Scott A||0·1||0·2||0·4|
|Listeria monocytogenes 19115||0·1||0·2||0·2|
HLF activity was much lower in TSB than in PYG. After 24 h, HLF did not inhibit any of the Gram-negative bacteria at the highest concentration (1·6 mg ml−1) in TSB and the MICs against L. monocytogenes Scott A and ATCC 19115 increased from 0·1 mg ml−1 in PYG to 1·6 and 0·8 mg ml−1, respectively, in TSB. The addition of subinhibitory concentrations of EDTA enhanced the activity of HLF in TSB against foodborne pathogens. After 48 h, an additional 0·5 mg ml−1 EDTA was required for inhibition of E. coli O157:H7 at the same concentration of HLF (0·4 mg ml−1) that inhibited the strain in PYG ( Table 2). Addition of 0·4–0·5 mg ml−1 EDTA was necessary for E. coli O104:H21 to be inhibited by the same MIC as in PYG (0·8 mg ml−1) ( Table 2). Addition of EDTA at up to 0·5 mg ml−1 was insufficient to reduce the inhibitory concentration of HLF to the same level that inhibited Salm. enteritidis ATCC 13076 in PYG (0·4 mg ml−1) (data not shown). Inhibition was achieved only with 0·8 mg ml−1 HLF and 0·4 mg ml−1 EDTA. Similar results were demonstrated for both L. monocytogenes strains ( Table 3). While L. monocytogenes was inhibited by 0·2 mg ml−1 HLF in PYG broth, the highest concentration of EDTA (0·1 mg ml−1) only reduced the 48 h MIC of HLF in TSB to 1·6 mg ml−1.
Table 2. Inhibition of two strains of pathogenic Escherichia coli in the presence of HLF and EDTA in tryptic soy broth at 37 °C
|Escherichia coli |
|O157:H7 ATCC 43895|| || || || |
|Escherichia coli |
|O104:H21|| || || || |
Table 3. Inhibition of Listeria monocytogenes Strains in the Presence of HLF and EDTA in Tryptic Soy Broth at 37 °C
|Listeria monocytogenes |
|Scott A|| || || || |
| 0·1||+||+||+||+||+ –|
|Listeria monocytogenes |
|ATCC 19115|| || || || |
In TSB, EDTA at ≥0·2 mg ml−1 and > 0·8 mg ml−1 HLF were bactericidal to E. coli O157:H7, while ≥ 0·4 mg ml−1 EDTA and 1·6 mg ml−1 HLF were bactericidal to E. coli O104:H21 and Salm. enteritidis ATCC 13076 (data not shown). No combination of HLF and EDTA was bactericidal to either L. monocytogenes Scott A or L. monocytogenes ATCC 19115.
These results confirm that pepsin-hydrolysed lactoferrin has antimicrobial activity against a variety of foodborne pathogenic bacteria in PYG broth. The MICs for E. coli and Salm. enteritidis are similar to those reported by Tomita et al. (1991) . In addition, the MICs against enterohaemorrhagic E. coli correspond exactly with the results obtained by Shin et al. (1998) against four strains of E. coli O157:H7. Listeria monocytogenes was found to be more resistant to the hydrolysate than shown by Tomita et al. (1991) , who reported an MIC of 0·08 mg ml−1 for L. monocytogenes. This may be due to either strain differences or to differences in the assay conditions. For example, Tomita et al. (1991) recorded MICs after 16–20 h while in the present study MICs were determined after 24 and 48 h. Tomita et al. (1991) reported that hydrolysed lactoferrin was bactericidal to E. coli at 0·15, 0·30 and 0·50 mg ml−1. In contrast, we found that the pepsin hydrolysate was only bactericidal to E. coli at concentrations ≥ 0·8 mg ml−1. Again, this may be due to differences in strain, methodology or media composition. Tomita et al. (1991) studied the bactericidal effect by determining viable cell numbers over a 2-h period or after 14 h. While a bactericidal effect could certainly be seen in 2 h, there is the possibility that with low concentrations of the hydrolysate, some cells may survive and grow after 24 and 48 h. In addition, when these researchers determined viable cells after 14 h, their detection limit was 4·0 log cfu ml−1 with an initial inoculum of 6·0 log cfu ml−1. Although there may have been a 2-log destruction of cells, there could have been at least 3·0 log cfu ml−1 survivors which would not have been detected. Finally, Tomita et al. (1991) determined bactericidal activity in 1% peptone. Again, this may help to explain the differences in MBCs between the two studies.
The active peptide in HLF, lactoferricin, is known to lose its activity in the presence of calcium and magnesium ions as well as other salts ( Wakabayashi et al. 1992 ). Subinhibitory concentrations of EDTA (determined in previous analyses) were combined with HLF to determine if the antimicrobial activity of the latter might be restored in TSB ( Tables 2 and 3). The activity of HLF was enhanced in TSB by the presence of EDTA, especially against E. coli and Salm. enteritidis. These results indicate that decreased activity of hydrolysed lactoferrin may have been due, in part, to the presence of excess cations in TSB and that chelation of these by EDTA restored the inhibitory activity of HLF. There are other substances in TSB that may also interfere with HLF activity, such as protein digests (20 mg ml−1) and glucose (2·5 mg ml−1). However, in a study with the purified peptide lactoferricin, it was found that various sugars up to 10 mg ml−1 had no effect on the activity of lactoferricin against L. monocytogenes ( Wakabayashi et al. 1992 ). The effect of proteins on lactoferricin is still not completely understood. One study showed that 10 mg ml−1 BSA and gelatine slightly increased the activity of lactoferricin against L. monocytogenes ( Wakabayashi et al. 1992 ), while another reported that the same concentration of BSA and gelatine had no effect on the activity of lactoferricin against E. coli ( Jones et al. 1994 ).
Another reason for the discrepancy between HLF activity in PYG compared with TSB may simply be that PYG does not support growth as well as TSB, which may give HLF the opportunity to be more effective against slow-growing cells. As previously stated, L. monocytogenes did not grow well in PYG broth. In fact, none of the strains studied grew as well in PYG as in TSB. The mean optical density of E. coli after 48 h in PYG was 0·141, while in TSB, the mean was 0·913 indicating the latter supported better growth. Therefore, HLF may be more effective when bacterial growth rate is reduced. Since EDTA is a bacteriostatic agent, its function in enhancing the activity of HLF in the richer medium, TSB, may have been in slowing the growth rate of the strains. Although we used EDTA concentrations that were much lower than the MICs for each strain, there may have been enough EDTA present to slow the growth rate without completely inhibiting growth.
Our results, coupled with the results of other researchers, have shown that the pepsin hydrolysate of lactoferrin, as well as the active peptide in the hydrolysate lactoferricin, are active inhibitors of many foodborne pathogens from both the Gram-positive and -negative genera. However, while HLF was active in laboratory media such as PYG, it was not as effective in a more complex medium (TSB). This raises the question of whether HLF would be effective in food systems, which are often very complex. Recently, Venkitanarayanan et al. (1999) showed that, while 50 or 100 µg ml−1 lactoferricin B reduced viable E. coli O157:H7 in 1% peptone, it was much less effective as an antimicrobial in ground beef. Our results indicate that HLF may be useful in food products that either are low in cations or which contain EDTA or other chelators. Although HLF was not very active in TSB, we were able to show that its effectiveness was restored with the addition of EDTA, which may have acted by either chelating excess cations, or by slowing the growth of bacteria and allowing HLF to act. Therefore, while use of HLF or lactoferricin as sole antimicrobials in foods may not be practical, they may have potential for use in foods utilizing a combination of antimicrobials.