Antibacterial activity of lactic acid bacteria included in inoculants for silage and in silages treated with these inoculants

Authors


N. Gollop, Department of Food Science, The Volcani Center, Bet Dagan 50250, Israel (e-mail: ngollop@agri.gov.il).

Abstract

Aims:  To determine antibacterial activity in lactic acid bacteria (LAB) silage inoculants and in wheat and corn silages which were treated with these inoculants.

Methods and Results:  Wheat and two corn silages were prepared in 0·25 l sealed glass jars. Inoculant treatments were prepared for each type of silage with each of 10 LAB silage inoculants at inoculation rate of 106 CFU g−1. Untreated silages served as controls. Antibacterial activity was determined in the inoculants and in their respective silages with Micrococcus luteus and Pseudomonas aeruginosa. Antibacterial activity was detected in nine of the 10 inoculants whereas such activity in the silages varied. Control silages did not have antibacterial activity.

Conclusions:  Many LAB silage inoculants have antibacterial activity and in some cases this activity is imparted on inoculated silages.

Significance and Impact of the Study:  This study was conducted as part of a broader research objective, which is to find out how LAB silage inoculants enhance ruminant performance. The results of this study indicate that LAB silage inoculants produce antibacterial activity, and therefore, have a potential to inhibit detrimental micro-organisms in the silage or in the rumen.

Introduction

Ensiling is a preservation method for moist crops that is based on natural lactic acid fermentation under anaerobic conditions, whereby epiphytic lactic acid bacteria (LAB) converts water-soluble carbohydrates (WSC) into organic acids. As a result the pH decreases and the forage is preserved.

Inoculants containing principally LAB are used as silage additives in order to improve preservation efficiency. Among the LAB most frequently used are homofermentative species such as Lactobacillus plantarum, Enterococcus faecium and Pediococcus spp. These are used because of their efficient utilization of the crop's WSC, intensive production of lactic acid and their ability to cause a rapid decrease in pH. Other LAB are also included, such as Lactobacillus buchneri, a heterofermentative LAB which produces high concentrations of acetic acid in silage that inhibits fungi, thereby improving the aerobic stability of silage. An inoculation rate of 105–106 viable cells per gram crop is often sufficient for the inoculant LAB to overwhelm the epiphytic LAB and become the predominant population in the silage (Weinberg and Muck 1996).

In some cases an effect of silage inoculation on animal performance has also been observed. Feed intake, live-weight gain, energy feed efficiency and/or milk production were improved in 25–40% of the studies reviewed. The improvement of these parameters ranged from 5 to 11% (Muck 1993).

The cause of improved animal performance is unclear. A considerable number of animal experiments using a single silage inoculant strain, L. plantarum MTD1 were performed in Northern Ireland with grass silage (Keady and Steen 1994, 1995; Keady et al. 1994). The majority of these studies reported improved animal performance with silages inoculated with this strain, regardless of fermentation quality. When the inoculant was added to the silage immediately before feeding, there was no significant effect on dry matter (DM), nitrogen, neutral detergent fibre (NDF) or acid detergent fibre (ADF) digestibility (Keady and Steen 1996). This might suggest that the benefits result from the silage fermentation rather than from a direct effect of LAB's on rumen fermentation. By contrast, in a recent study (Khuntia and Chaundhary 2002), dietary addition of a mixed culture of LAB increased DM intake, live-weight gain and DM digestibility in calves. Rumen pH was lower and lactic acid was higher following LAB feeding. Salawu et al. (2001) found that application of L. plantarum to pea-wheat silage increased the rate of nitrogen and NDF degradation in the rumen. Malik and Sharma (1998) inoculated rumen fluid with various micro-organisms in the presence of wheat straw and concentrates, and showed that Lactobacillus acidophilus improved DM and organic matter digestibility in vitro as compared with an untreated control.

The results of some of these studies suggest a possible probiotic effect of LAB used in inoculants for silage, the mechanism of which is as yet unclear. One hypothesis is that specific LAB strains interact with rumen micro-organisms to enhance rumen functionality and animal performance. In that context, Weinberg et al. (2003) showed that LAB from silage inoculants can survive in rumen fluid for at least 96 h and they resulted in increase in pH. Another possibility is that LAB, which are used as inoculants for silage, inhibit detrimental micro-organisms in the silage. In this regard, it is well known that LAB produce a variety of antimicrobial substances such as bacteriocins (e.g. Vandenbergh 1993; Muller et al. 1996).

Bacteriocins are ribosomally synthesized antibacterial peptides secreted by many strains of micro-organisms. Bacteriocins are usually active against bacteria closely related to the strain from which they are produced (Tagg et al. 1976). Bacteriocins are found in numerous Gram-positive and Gram-negative bacteria, but those produced by LAB and other micro-organisms used in the food industry have been subjected to intensive research because of their potential use as food preservatives (Stiles 1996).

The purpose of the present study was to determine antibacterial activity in LAB silage inoculants cultures and in silages treated with these inoculants.

Materials and methods

Media

The MRS media was used for Enterococcus spp. and Rogosa SL was used for Lactobacilli (Difco Becton Dickinson and Company, Sparks, MD, USA).

The LB media was used for Micrococcus luteus and Pseudomonas aeruginosa.

Silage fermentation

One type of wheat silage (at the milk ripening stage, at 350 g kg−1 DM, harvested in spring) and two types of corn silages (at early dent and at half milk line, at 195 and 280 g kg−1 DM, respectively, harvested in summer) were treated with and without the same 10 LAB inoculants (separately) which were applied at 106 CFU g−1 each (i.e. 33 treatments in total). The inoculation rate was based on the numbers of CFU g−1 in the inoculants powders which was determined before the experiments had started. The inoculants were applied by suspending an adequate weight of powder in 20 ml of de-ionized water which was sprayed over batches of 2 kg forage and mixed thoroughly. The chopped crops (control and treated) were ensiled in 0·25 l sealed jars. Control silages were prepared at the same time from crops without inoculant addition. There were four jars per treatment for each silage type. The silages were stored for 8 (wheat) and 5 months (corn) at room temperature (25–28°C). Differences in storage duration were because of different harvesting time of the crops (wheat in April, and corn in July). At the end of the fermentation process the silages were analysed for pH, fermentation products and enumeration of LAB. Dry matter was determined by oven-drying for 48 h at 60°C. Lactic acid was determined by a spectrophotometric method according to Barker and Summerson (1941), and volatile fermentation products were determined by gas chromatography as described in Weinberg et al. (2004).

The number of LAB cells in the dry products was determined before the experiments by suspending the inoculants in deionized water and pour plating serial dilutions into the MRS or Rogosa agar. Enumeration of LAB in the silage samples was carried out with pour plates. Plates were incubated at 30°C for 3 days.

Determination of antibacterial activity in the LAB inoculants

Antibacterial activity was determined by an agar diffusion assay (Jack et al. 1995). The LAB silage inoculants were grown to stationary phase in MRS broth and the media were tested for antibacterial activity by applying 5 μl to LB soft agar plates which contained 400 μl of late stationary phase of M. luteus.

Determination of antibacterial activity in silages

Antibacterial activity in silages was determined at the end of the storage period. Five grams of pooled silage samples were blended with 10 ml of deionized sterile water in a Stomacher blender for 3 min at room temperature. The supernatants were separated from the silage by centrifuge 5000 g for 10 min at 4°C, and evaporated to dryness using Speed-Vac (Heto, Gydevang, Denmark). The dry sample was rehydrated with 2 ml of deionized sterile water and tested for the existence antibacterial activity as described above. Micrococcus luteus and P. aeruginosa served as Gram-positive and Gram-negative indicator organisms respectively. The pH of the sample was adjusted to pH 7 prior to use in the activity assay. Antibacterial activity was determined on three separate replicates.

Results

Table 1 lists the LAB silage inoculants used and their respective pH and LAB viable counts in the silages. Table 2 summarizes the silage parameters. The DM content of the fresh crops (g kg−1) was 350 ± 3 (wheat), 195 ± 1 (early dent corn) and 280 ± 2 (half milk line corn). During the storage period the pH of all silages decreased markedly. The major fermentation product was lactic acid, but when the heterofermentative L. buchneri was used it produced more acetic than lactic acid. In the wheat silage treated with Lactobacillus pentosus the pH was low – 3·9 in spite of the low LAB count (only 5·6 log10).

Table 1.  List of inoculants and their pH and LAB viable counts in the silages
InoculantManufacturerWheat silage (DM 350 g kg−1)Corn silage (DM 190 g kg−1)Corn silage (DM 280 g kg−1)
pHLAB counts (log10 CFU/g−1)pHLAB counts (log10 CFU/g−1)pHLAB counts (log10 CFU/g−1)
  4·57·04·38·73·65·8
L. plantarum MTD1Ecosyl, Yorkshire, UK4·06·74·28·63·76·4
P. pentosaceusEcosyl, Yorkshire, UK4·16·34·38·73·66·8
L. plantarumAgri-King, Fulton, IL, USA4·26·14·08·73·66·1
L. pentosusAgri-King, Fulton, IL, USA3·95·64·18·73·66·5
P. pentosaceusAgri-King, Fulton, IL, USA3·96·54·18·63·66·3
E. faecium (C)Agri-King, Fulton, IL, USA4·56·34·47·43·76·7
E. faecium (Q)Agri-King, Fulton, IL, USA4·07·04·27·63·77·2
L. buchneriBiotal Canada Ltd, Calgary, AB, Canada4·27·13·87·03·87·1
L. buchneri 11A44Pioneer Hi-Bred International Inc., Johnston, IA, USA4·36·54·07·03·86·9
L. plantarum and E. faecium 1188Pioneer Hi-Bred International Inc., Johnston, IA, USA3·96·74·17·63·77·2
Table 2.  Analysis of the silages
SilageDry matterpHLactic acidEthanolAcetic acidLAB
  1. Dry matter is expressed in g kg−1; fermentation products, g kg−1 DM; LAB, log10 CFU g−1 DM.

Wheat3503·9–4·55–640–1310–635·6–7·1
Corn I1903·8–4·329–890–1427–537·0–8·7
Corn II2803·5–3·841–844–815–296·1–7·2

Antibacterial activity in the LAB inoculants

Table 3 gives the antibacterial activity of the LAB inoculants against M. luteus. All cultures except Pediococcous pentosaceus showed antibacterial activity but different zone inhibition sizes were observed with the various silage inoculants.

Table 3.  Antibacterial activity in the growth media of LAB silage inoculcants against Micrococcus luteus
InoculantInhibition zone
  1. (−) No clear zone; (+) 0–2 mm inhibition zone; (++) 3–4 mm inhibition zone; (+++) 5–8 mm inhibition zone.

L. plantarum MTD1++
P. pentosaceus+++
L. plantarum+++
L. pentosus+++
P. pentosaceus
E. faecium C+
E. faecium Q+
L. buchneri+
L. buchneri+++
L. plantarum and E. faecium+++

Antibacterial activity in the silages

The silage extracts were tested for antibacterial activity on two different indicator bacteria, a Gram-positive (M. luteus) and Gram negative (Pseudomonas aeruginosa) (Table 4). Although most of the pure cultures showed antibacterial activity, only some of the silages showed antibacterial activity. Although the number of LAB was high in all silages and the inculcated cultures were capable to produce antibacterial substances, the extraction of corn and wheat silages did not always show antibacterial activity. Silages inculcated with P. pentosaceus cultures are in agreement with the results obtained with pure culture of this strain that did not show any antibacterial activity. Silages extracts that showed antibacterial activity against Gram-positive bacteria also had activity against Gram-negative bacteria.

Table 4.  Antibacterial activity in various silages treated with LAB silage inoculants against Micrococcus luteus and Pseudomonas aeruginosa
InoculantInhibition zone
Micrococcus luteusPseudomonas aeruginosa
Wheat silage (DM 350 g kg−1)Corn silage (DM 190 g kg−1)Corn silage (DM 280 g kg−1)Wheat silage (DM 350 g kg−1)Corn silage (DM 190 g kg−1)Corn silage (DM 280 g kg−1)
  1. (−) No clear zone; (+) 0–3 mm inhibition zone; (++) 3–8 mm inhibition zone; (+++) 9–13 mm inhibition zone.

Control
L. plantarum MTD1++
P. pentosaceus++++++++++++++++
L. plantarum++++++++
L. pentosus+++
P. pentosaceus
E. faecium++++++++
E. faecium+++++
L. buchneri++++++
L. buchneri++++++++++
L. plantarum and E. faecium+++++++++++++++++

Discussion

The current study was conducted as part of a broader research objective, which is to find out how LAB silage inoculants enhance ruminant performance. One hypothesis for this phenomenon is that specific LAB strains interact with rumen micro-organisms to enhance rumen functionality (fibre digestibility) and animal performance. Another possibility is that LAB, which are used as inoculants for silage, inhibit detrimental micro-organisms in the silage by producing antibacterial substances. It is well known that LAB produce a variety of antimicrobial substances such as bacteriocins (e.g. Vandenbergh 1993; Muller et al. 1996).

So far in that research, it was shown that LAB included in silage inoculants can survive in rumen fluid for at least 96 h (Weinberg et al. 2003). Recently, Weinberg et al. (2004) showed that LAB passed from silage into rumen fluid in vitro and survived there for at least 48 h.

The current results indicate that LAB silage inoculants and many of the silages treated with these inoculants have antibacterial activity. It is possible that the antibacterial activity is related to bacteriocin production, a matter which warrants further investigation. There was a good agreement between the activity in the inoculant powder and the activity in the silages treated with the LAB inoculants. The antimicrobial activity of the silage inoculants was not consistently expressed among the silages (Table 4): for example, L. buchneri (inoculant 8) exhibited antimicrobial activity on the moist corn silage, but not on the other two silage types, whereas for another strain of L. buchneri (inoculant 9) the opposite held true. Our assumption is that antibacterial activity expressed by the inoculant LAB strains depends on environmental conditions prevailing in the silage, such as water activity, pH and presence of volatile fatty acids, which are specific for each LAB strain. The various silage treatments contained different concentrations of lactate and acetate, according to the fermentation pattern, i.e. homofermentation vs heterofermentation (data not shown). In the silage, these acids result in decrease in pH, and in addition, inhibit various micro-organisms by mechanism other than just pH decrease (e.g. acetic acid is known to inhibit yeasts and moulds). However, in the current study it was not possible to correlate lactate or acetate levels in the silages with the antibacterial properties of the silage juice. It should be stressed that uninoculated control silages did not show antibacterial activity. In control silages, unlike in inoculated ones, the ensiling fermentation is carried out by a variety of the epiphytic LAB which apparently do not possess detectable antibacterial activity. Therefore, treating silages with LAB inoculants might have advantages of imparting antibacterial activity to the silages which might inhibit detrimental micro-organisms in the silage or even in the rumen. This might have a beneficial effect on ruminant performance. In order to maximize such beneficial effects, a selection process should be employed to select inoculants that only inhibit detrimental micro-organisms in the silage or in the rumen, because it is equally possible that these inoculants inhibit beneficial bacteria as well. For example, Ohmomo et al. (2000) isolated a strain of E. faecium from fermented vegetables that produced a bacteriocin which inhibited the pathogen Listeria monocytogenes but also various beneficial LAB.

Further research on the potential probiotic effects of LAB silage inoculants on ruminants should include their interactions with rumen micro-organisms and their effects of fibre digestibility.

In conclusion, many LAB silage inoculants have antibacterial activity and in some cases this activity is imparted on inoculated silages. This activity might inhibit detrimental micro-organisms in the silage and in the rumen and thus, enhance the animal health and performance.

Acknowledgements

This research was a contribution from the Agricultural Research Organization, the Volcani Center, Bet Dagan, Israel, no. 401/03-E, 2003 series. The research was supported by Research Grant No. IS-3297–02 from BARD, The United States-Israel Binational Agricultural Research and Development Fund.

Ancillary