Potential of lactic acid bacteria in aflatoxin risk mitigation Sara H. Ahlberg, Vesa Joutsjoki, Hannu J. Korhonen PII: DOI: Reference:

S0168-1605(15)00245-7 doi: 10.1016/j.ijfoodmicro.2015.04.042 FOOD 6907

To appear in:

International Journal of Food Microbiology

Received date: Revised date: Accepted date:

14 January 2015 10 April 2015 25 April 2015

Please cite this article as: Ahlberg, Sara H., Joutsjoki, Vesa, Korhonen, Hannu J., Potential of lactic acid bacteria in aflatoxin risk mitigation, International Journal of Food Microbiology (2015), doi: 10.1016/j.ijfoodmicro.2015.04.042

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT A review: Potential of lactic acid bacteria in aflatoxin risk mitigation 1

1

T

12

Sara H Ahlberg , Vesa Joutsjoki and Hannu J Korhonen

LUKE Natural Resources Institute Finland 31600 Jokioinen Finland

2

ILRI International Livestock Research Institute, P.O. Box 30709-00100 Nairobi, Kenya, Office tel. +254

RI P

1

SC

20 422 3473, Mobile +254 705 949 932 [email protected]

MA NU

[email protected] [email protected]

Word count: 11138

ED

ABSTRACT

PT

Aflatoxins (AF) are ubiquitous mycotoxins contaminating food and feed. Consumption of contaminated food and feed can cause a severe health risk to humans and animals. A novel biological method could reduce the

CE

health risks of aflatoxins through inhibiting mold growth and binding aflatoxins. Lactic acid bacteria (LAB) are commonly used in fermented food production. LAB are known to inhibit mold growth and, to some extent, to

AC

bind aflatoxins in different matrices. Reduced mold growth and aflatoxin production may be caused by competition for nutrients between bacterial cells and fungi. Most likely, binding of aflatoxins depends on environmental conditions and is strain-specific. Killed bacteria cells possess consistently better binding abilities for aflatoxin B1 (AFB1) than viable cells. Lactobacilli especially are relatively well studied and provide noticeable possibilities in binding of aflatoxin B1 and M1 in food. It seems that binding is reversible and that bound aflatoxins are released later on (Haskard et al. 2001; Peltonen et al. 2001). This literature review suggests that novel biological methods, such as lactic acid bacteria, show potential in mitigating toxic effects of aflatoxins in food and feed.

Keywords Lactic acid bacteria, aflatoxins, binding, fungal growth

1

ACCEPTED MANUSCRIPT Chemical compounds studied in this review Aflatoxin B1 (PubChem CID: 14403)

T

Aflatoxin M1 (PubChem CID: 15558498)

RI P

Abbreviations

SC

AF Aflatoxins AFB1 Aflatoxin B1

MA NU

AFM1 Aflatoxin M1 A. Aspergillus Bb. Bifidobacterium CYP Cytochrome P CFU Colony Forming Units

ED

DNA Deoxyribonucleic Acid

EU European Commission

PT

ELISA Enzyme-Linked Immunosorbent Assay

FB1 Fumonisin B1 FB2 Fumonisin B2

CE

FAO Food and Agricultural Organization

AC

FDA US Food and Drug Administration HACCP Hazard Analysis and Critical Control Points HCl Hydrochloric Acid

LAB Lactic Acid Bacteria Lb. Lactobacillus Lc. Lactococcus L. Leuconostoc NaCl Sodium Chloride MIC Minimum Inhibitory Concentration PBS Phosphate Buffered Saline P. Pediococcus

2

ACCEPTED MANUSCRIPT P. Propionibacterium S. Streptococcus W. Weissella

RI P

T

WHO World Health Organization

SC

1. Introduction

Mycotoxins are a group of toxic secondary metabolites produced by fungi which are found ubiquitously in the

MA NU

soil. Elevated temperatures and humid environmental conditions promote the fungal growth and toxin production. Mycotoxins appear in six major classes: aflatoxins, fumonisins, ochratoxins, trichothecenes, zearalenone and ergot alkaloids.

Among mycotoxins, aflatoxins are the most toxic. AFB1 is the most carcinogenic of the naturally occurring

ED

aflatoxins (IARC, 2002) and causes acute and chronic intoxication in humans and animals (Shetty et al. 2007). AFB1 metabolizes to the 4-hydroxy derivate aflatoxin M1 (AFM1) in lactating animals including humans

CE

PT

(IARC, 2002).

Aspergillus species produce aflatoxins and Fusarium species produce fumonisins (IARC, 2002). These two

AC

fungal genera frequently occur simultaneously and it is highly probable that fumonisin B1 (FB1) and AFB1 are co-contaminants of foods (Pizzolitto et al. 2012). FB1 and FB2 constitute up to 70% of the fumonisins found in foods (Niderkorn et al. 2006).

Mycotoxins are present in foods as a result of raw material contamination, or deficiencies in harvesting, storage or processing. Mycotoxins in maize pose a notable health risk to humans and certain farm animals. In East and West African countries where maize is a staple food, dietary exposure to aflatoxins is frequent and may occur at high levels. Exposure to mycotoxins increases the prevalence of bacterial and parasitic infections (IARC, 2002).

Acute exposure to aflatoxins can cause aflatoxicosis, and in severe hepatotoxicity cases the mortality rate is approximately 25%. Chronic exposure to aflatoxins is associated with hepatocellular carcinoma, especially in

3

ACCEPTED MANUSCRIPT the presence of hepatitis B infection. Other probable health impacts are immunological suppression, impaired growth and nutritional interference (Strosnider et al. 2006). These impacts have been demonstrated in various species of livestock and fish, and while they may have similar effects on humans causal evidence

RI P

T

is still lacking.

Biological, chemical, and physical methods exist to eliminate, inactivate, and prevent fungal growth and toxin

SC

production. Most of these applications have not been adopted, especially in poor countries, due to high costs, loss of nutritional value, altered organoleptic characteristics of the products, practical difficulties, and

MA NU

(un)known effects to human health. On the other hand, traditionally used indigenous food microorganisms may have potential in prevention of mycotoxin-caused health impacts (Bhat et al. 2010; Di Natale et al. 2009).

Lactic acid bacteria (LAB) are commonly used in silage feed production and in fermented food product

ED

processes. LAB are known to inhibit fungal growth and extend the shelf-life of the product (Broberg et al. 2007). This literature review evaluates the potential of different LAB strains to prevent the growth of aflatoxin

PT

forming fungal strains and bind AFB1 and AFM1 in vivo and in vitro and examine the nature of aflatoxin-

CE

binding. The published studies will be discussed critically, and finally some conclusions will be drawn.

AC

2. Aflatoxin in the feed - food chain

Aflatoxin-producing fungi are ubiquitous in the soil. The prevention of contamination and good crop handling practices play a key role in the control of fungal growth in foods and feeds. Maize and milk in different forms comprise a significant part of staple food diet in Kenya and some other parts of the East African region. Climate change is likely to worsen the challenges during harvesting and storage. Aflatoxin contamination is especially severe after a long-term storage in excessive heat and humid conditions. Damage-causing insects and rodents spread the fungal spores allowing proliferation. Even small exposure reductions from several sources in high-exposure areas can have important benefits (Turner et al. 2005).

Fungi live on the surface of the crop and Aspergillus species are dominant in favorable conditions of maize cultivation areas (Kpodo et al. 2000). In areas of seasonal food scarcity spoiled maize is occasionally used

4

ACCEPTED MANUSCRIPT for human consumption, but even more often spoiled maize is fed to dairy animals and poultry. Aflatoxin does not accumulate in muscle meat, but is excreted in milk, urine, feces and is found also in blood.

T

AFB1 is metabolized predominantly by the cytochrome P450 enzyme system (involved in xenobiotics

RI P

metabolism in humans) to produce a range of metabolites (Guengerich, 2001). The toxicity of AFB1 is generally regarded to occur via CYP 3A4, but also by CYP 1A2 by the production of the highly reactive AFB1

SC

- 8, 9-epoxide which forms covalent adducts with macromolecules, such as proteins and DNA. The 8, 9epoxide of AFB1 is short-lived but highly reactive and is capable of causing damage to cells in the liver and at

MA NU

the intestinal interface. Direct damage caused by aflatoxin exposure within the intestine may alter nutrient uptake (Gratz et al. 2007).

According to a WHO estimation (Wu et al. 2011), 25,000 to 155,000 aflatoxin-induced liver cancer cases occur globally each year of which 40% are estimated to be in Africa. Co-occurrence of AFB1 with hepatitis B

ED

increases the liver cancer risk 12-fold. Data from Wu and Tritscher (2011) suggests that children are more

PT

vulnerable to acute hepatotoxicity from ingested aflatoxins than adults.

Aflatoxins are genotoxic carcinogens and their combined and separate levels are subject to regulation in

CE

most countries. Accepted exposure levels are different due to the wide variety of standards, levels, and geographical area of consumption. The international levels are political compromises to promote trade. Little

AC

is known about specific threshold levels or health effects associated with consumption of aflatoxin concentrations between 20 µg/kg and 300 µg/kg (Strosnider et al. 2006).

US Food and Drug Administration (FDA) levels are set at 20 µg/kg for human foods, 0.5 µg/kg for milk, and 20-300 µg/kg for feed. The Codex Alimentarius limit for all foods worldwide is 15 µg/kg (FAO, WHO, 1999). In the European Union the limit for AFB1 in maize products is 5 µg/kg and combined aflatoxins 10 µg/kg, other cereal products 2 µg/kg and combined aflatoxins 4 µg/kg, AFM1 in raw milk 0.05 µg/kg, AFB1 in infant cereal 0.10 µg/kg, and AFM1 in infant milk 0.025 µg/kg (EU, 2006).

For feeds given to milk-producing animals, a level of 5 µg/kg AFB1 is widely supported, as an AFM1 level of 0.05 µg/kg is achievable with such a limit ( FAO, WHO, 1993). The European Commission limit for AFB1 in

5

ACCEPTED MANUSCRIPT feeds is 20 µg/kg (EU, 2002). A summary of different AFB1 and AFM1 limits in different materials is presented in Figure 1.

RI P

T

3. Potential of LAB to inhibit fungal growth and reduce toxin production

Aflatoxins are excreted by toxin-producing fungi, but fungal growth does not necessarily entail toxin

SC

production. Even if toxin production is inhibited or reduced, mold growth might be possible (Gourama and Bullerman, 1995). Antifungal and antibacterial compounds produced by LAB are assumed to reduce the

MA NU

toxin production. Gourama and Bullerman (1995) suggested that lactic acid and other metabolites produced by LAB play an important role in the inhibition of aflatoxin synthesis. These inhibiting substances were considered to be low-molecular-weight and heat-stable compounds. A low pH, depletion of nutrients, or

ED

microbial competition cannot alone explain the mechanisms of aflatoxin synthesis inhibition.

Dalié et al. (2010) listed some antifungal compounds produced by different strains: organic acids, phenolic

PT

compounds, hydroxyl fatty acids, hydrogen peroxide, reuterin, and proteinaceous compounds. Especially low-molecular-weight metabolites produced at the beginning of the exponential growth phase of LAB could

CE

have an inhibitory effect on aflatoxin accumulation.

AC

Several studies show that LAB can reduce fungal growth, at least in vitro. Dalié et al. (2010) listed species belonging to Lactococcus and Lactobacillus genera (and Pediococcus and Leuconostoc genera) as recognized for their ability to prevent or inhibit the toxic fungal growth. Dalié et al. (2010) listed several factors influencing the antifungal activity of LAB. These factors are incubation period and temperature, culture medium effect on bacterial metabolism, nutritional compounds delaying and enhancing the production of antifungal compounds, and pH. All these factors are highly dependent on the species.

For example, Lactococcus lactis subsp. diacetylactis produced low amounts of antifungal compounds against A. fumigatus in milk, although these compounds were not detected in the experimental incubation (Dalié et al. 2010). High concentrations of NaCl reduced the production of antifungal compounds. Xylose, casein-hydrolysate, and proteose-peptone delayed the production of toxins. The efficiency of antifungal

6

ACCEPTED MANUSCRIPT compounds and proliferation of Lactobacillus rhamnosus was increased after 1% and 2% glucose addition and in 3% NaCl concentration, when added separately.

RI P

T

Studies on individual LAB species are described in the following paragraphs.

SC

3.1 Lactobacillus strains

Gourama and Bullerman (1995) isolated Lb. casei pseudoplantarum DMS 20008 from commercial silage

MA NU

culture and screened it for inhibitory effect on fungal growth and aflatoxin production with living cultures and cell-free supernatants. The living cultures inhibited fungal growth and aflatoxin production was reduced. The cell-free supernatants did not inhibit fungal growth but significantly reduced aflatoxin production.

Onilude et al. (2005) studied the inhibition ability of two Lb. fermentum strains, against aflatoxin-producing

ED

Aspergillus strains. The inhibition zones of the tested Lb. fermentum strains varied between fungal strains

PT

between 1 to 16 mm and 5 to 12 mm, respectively.

CE

Lb. fermentum L23, a LAB strain with probiotic properties, was tested against 10 different Aspergillus section Flavi fungal strains to assess effects on growth inhibition and reduction of AFB1 production (Gerbaldo et al.

AC

2012). Lb. fermentum L23 inhibited nine of the tested strains but showed no growth inhibition effect against one Aspergillus section Flavi strain. The fungal growth inhibition rate increased from 36 to 50% compared with the control. Aflatoxin production decreased when the fungal growth decreased. Lb. fermentum L23 decreased aflatoxin production between 73 – 99%.

Most of the published mycotoxin inhibition studies with LAB have used Lb. plantarum strains. Gourama and Bullerman (1995) isolated Lb. plantarum NCDO 1752 from commercial silage inoculants and tested the bacterial culture and the cell-free supernatant ability to inhibit mycelia growth and to reduce aflatoxin production. In bacterial culture incubation the mycelia growth was inhibited during 6 days of incubation. Also, toxin production was diminished. When the cell-free supernatant inhibition was tested, the mycelia growth was significantly reduced compared with controls or any other tested bacteria strain or mixture. The AFB1 production was significantly reduced.

7

ACCEPTED MANUSCRIPT

Vanne et al. (2001) tested fungal growth inhibition abilities of Lb. plantarum strains (VTT E-78076 and VTT E-79098) against 15 Penicillium verrucosum and Aspergillus ochraceus strains. The inhibition rate varied

T

between fungal strains from 28 to 50%. Laitila et al. (2002) studied the same Lb. plantarum strains against

RI P

Fusarium species. The growth inhibition against four different Fusarium species varied from 33 to 53% with VTT E-78076 and 30 to 50% with VTT E-79098. Lactic acid alone (0.25%) inhibited at best 20% of the

SC

growth but mainly less than 10%. The results indicate that lactic acid and low pH alone cannot explain the

MA NU

inhibitory action of Lb. plantarum.

Most of the antifungal studies are usually performed with pure cultures and in optimal controlled laboratory conditions. Laitila et al. (2002) analyzed Lb. plantarum VTT E-78076 activity against Fusarium fungi in naturally infested steeped barley and malt. Soaking of kernels induced the fungi to grow, hence it was important to add Lb. plantarum at an early stage of soaking and malting. In general, VTT E-78076 reduced

ED

Fusarium growth by 20-50% in barley and by 5-17% in fine malts. The Lb. plantarum effect was very limited

PT

in heavy fungal growth.

Valerio et al. (2004) identified from Lb. plantarum strains a range of antimicrobial compounds. These

CE

compounds include phenyllactic acid and its 4-hydroxy derivate. Also, 3-hydroxydecanoic acid, 2-hydroxy-4methylpentanoic acid, benzoic acid, catechol, hydrocinnamic acid, salicylic acid, 3-phenyllactic acid, 4-

AC

hydroxybenzoic acid, (trans, trans)-3,4-dihydroxycyclohexane-1-carboxylic acid, ρ-hydrocoumaric acid, vanillic acid, hydroferulic acid, ρ-coumaric acid, hydrocaffeic acid, ferulic acid, and caffeic acid were found in grass silage made with Lb. plantarum strains. Other identified antifungal metabolites additional to above from Lb. plantarum strains found in grass silage were catechol and azelaic acid (Broberg et al. 2007). These metabolites and fungal strains were tested separately for a minimum inhibitory concentration (MIC mg/ml) against Pichia anomala, Penicillium roqueforti and Aspergillus fumigatus. Against all three fungi, 3hydroxydecanoic acid had the lowest MIC, 0.1 mg/ml or more. Against A. fumigatus 0.1 mg/ml MIC was reached with 3-hydroxydecanoic acid, hydrocinnamic acid, salicylic acid, and benzoic acid. A MIC of 0.1 mg/ml or above was required for ρ-coumaric acid, ferulic acid, vanillic acid and azelaic acid. Concentrations found in a gram per inoculated grass silage were not enough to reach the minimum inhibitory concentration.

8

ACCEPTED MANUSCRIPT Onilude et al. (2005) isolated Lb. plantarum MW and YO from indigenously fermented cereal gruels and measured the fungal inhibition zone for four Aspergillus flavus and two Aspergillus parasiticus strains. Both

T

Lb. plantarum strains inhibited fungal growth, but the results varied between strains.

RI P

Valerio et al. (2009) isolated Lb. plantarum C21-41 from durum wheat semolina samples, inoculated with Aspergillus niger ITEM5132 and calculated the fungal growth inhibition with respect to the uninoculated

SC

medium. A. niger growth was inhibited by 42% with Lb. plantarum.

MA NU

An antifungal compound, 3,6-bis(2-methylpropyl)-2,5-piperazinedion, was identified from Lb. plantarum AF1, isolated from the fermented vegetable food kimchi (Yang and Chang, 2010). A concentrated Lb. plantarum AF1 culture inhibited fungal Aspergillus flavus growth completely in soya beans for two days. A cell-free supernatant and a combination of organic acids analyzed had better abilities (visually detected) to inhibit fungal growth. Also, lactic acid, at the concentration of 106.05 mg/ml, inhibited fungal growth. The

ED

metabolites oxalic acid (7.4 mg/ml), maleic acid (2.415 mg/ml), acetic acid (1.895 mg/ml), tartaric acid (0.28 mg/ml), citric acid (3.5 mg/ml) and malic acid (0.205 mg/ml) did inhibit fungal growth very little or not at all.

PT

Yang and Chang (2010) listed some antifungal compounds isolated from different LAB and suggested that

CE

the fungal growth inhibition is caused by other compounds in addition to organic acids.

According to Ndagano et al. (2011), the antifungal influence of separate organic acids by LAB and the

AC

inhibitory activity is rather caused by a synergy of the produced compounds. For example, lactic acid alone requires a concentration of 1220 mM to inhibit fungal growth, whereas the production level is around 76 mM. When antifungal activity was tested against pH changes, at pH 4 and 5 the inhibition was nearly 100% after 24 h of incubation with all tested fungi, but dropped to 10-40% at pH 6. In a synergic inhibitory activity test against three Aspergillus strains, a growth inhibition level of over 75% was reached with a mixture of acetic acid, lactic acid, and phenyllactic acid at the concentrations of 33.3, 333, and 2 mM, respectively. The authors suggested that beside organic acids, other antifungal compounds produced by LAB could also contribute to the inhibition activity and act by synergy.

The growth inhibition ability of Lb. rhamnosus L60 against 10 different Aspergillus flavus section Flavi strains was studied by Gerbaldo et al. (2012) with an agar overlay method. Lb. rhamnosus inhibited the growth of all

9

ACCEPTED MANUSCRIPT fungal strains partially or totally, compared with the control, and the growth rate decreased from 77 to 96%. The toxin production of fungi was in pattern with the fungal growth inhibition when Lb. rhamnosus L60 was co-cultured with the fungal strains. The AFB1 production decreased between 96 to 99% compared with the

RI P

T

control.

SC

3.2 Weissella strains and mixtures of LAB

Gourama and Bullerman (1995) used a commercial silage inoculant, which contained a dried mixture of Lb.

MA NU

plantarum, Lb. bulgaricus, and Lb. acidophilus. Actively growing cells of the Lactobacillus strains inhibited the growth of A. flavus better than a cell-free supernatant. The fungal spore viability remained relatively constant for the first 8 hours in the presence of both the bacteria cells and supernatant. Compared with the control, the supernatant postponed the fungal growth and after 16 hours the inhibition was 82%. When the Lactobacillus strains culture was introduced to the fungal culture, the aflatoxin production at the end of the

ED

fungal growth period was inhibited more than when the fungal culture was introduced to the Lactobacillus

PT

culture.

CE

Gourama and Bullerman (1995) tested aflatoxin production inhibition with viable Lactobacillus species in a dialysis sack of different molecular weight-cutoffs. When the smallest molecules (cut-off 1000) were cut out,

AC

the inhibition level of AFB1 was 76%. When the cut-out molecular weight was 6000 to 8000 and 12,000 to 14,000, the aflatoxin B1 inhibition was 98% and 92%, respectively.

A silage inoculant did not inhibit mycelia growth after six days of incubation equally with Lb. casei strains, but was still better than the controls. With the inoculant aflatoxin production was reduced more than by single strain. Cell free supernatant from the silage inoculant reduced mycelia growth slightly whereas aflatoxin levels were significantly reduced.

Ismaiel et al. (2011) tested fungal growth inhibition and aflatoxin production abilities of the concentrated dairy product kefir. Kefir is an acid-alcohol fermented milk product, which is produced by inoculating milk with grains of kefir. Kefir grains consist of complex flora of LAB, acetic acid bacteria, and yeast. The exact microbial composition of kefir grains was not determined, but based on previous studies, it was estimated to

10

ACCEPTED MANUSCRIPT have a mixed culture of various yeasts (Kluyveromyces, Candida, Saccharomyces, and Pichia) and a mixture of various LAB (Lactobacillus, Lactococcus, and Leuconostoc) and acetic acid bacteria. The fungal growth of A. flavus, AFB1 production and sporulation decreased with increased kefir concentrations. The

T

relative inhibition reduction of the kefir concentrate was strongest at 1% v/v, 4.80 g/L of fungal dry weight. At

RI P

10% concentration the fungal growth and aflatoxin production were totally inhibited, and sporulation stopped

SC

at a kefir concentration level of 7.0%.

Ndagano et al. (2011) tested the antifungal activity of Weissella, W. cibaria FMF4B16 and W.

MA NU

paramesenteroides LC11, against Aspergillus. W. paramesenteroides LC11 had a stronger antifungal activity than Lb. plantarum or W. cibaria FMFB16 in the same study, although the latter two inhibited growth over 80%. In a pH test both Weissella strains had nearly 100% inhibition activity against Aspergillus at pH 4 and 5. The growth inhibition dropped to less than 25% at pH 6. An antifungal organic acid alone produced by

ED

Weissella was also insufficient to explain the inhibition activity.

PT

4. Potential Aflatoxin B1 binding by LAB

CE

Owing the ability of certain LAB strains to bind and neutralize mycotoxins, LAB are suggested as a potential novel biological method to reduce the toxicity of mycotoxins or prevent their absorption into the human body.

species.

AC

Most of the binding studies are done in optimal laboratory conditions with several strains of different LAB

Aflatoxin binding levels are generally analyzed from pelleted bacteria-aflatoxin complex supernatants. Cultured bacteria are centrifuged to pellets, suspended in solution, generally PBS (phosphate buffered saline) with added aflatoxin, incubated and centrifuged again into bacteria pellets with bound aflatoxins and supernatant with unbound aflatoxin. ELISA and HPLC are most used analyzing methods for the aflatoxin levels in the supernatants. (Haskard et al. 2001; Peltonen et al. 2000; Peltonen et al. 2001)

4.1 Lactobacillus

11

ACCEPTED MANUSCRIPT El-Nezami et al. (1998a) analyzed AFB1 removal by Lb. acidophilus ATCC 4356. During 72 hours incubation Lb. acidophilus bound up to 77% of AFB1. Binding was instant but decreased during the first 48 hours, then

T

increasing again back to the starting level.

RI P

Haskard et al. (2001) tested the binding ability of AFB1 by Lb. acidophilus, bacteria treatments and the effect of washing on the binding stability. Acid and heat treatments improved the initial binding and binding

SC

adhesion after repetitive aqueous washing. Lb. acidophilus strains showed good potential in binding and retaining the complex. Lb. acidophilus LC1 bound 59.7%, 74.7% and 84.2% while Lb. acidophilus ATCC

MA NU

4356 bound 48.3%, 69.7% and 81.3% as viable, heat treated and acid treated, respectively.

Peltonen et al. (2001) studied the AFB1 binding abilities of two different Lb. acidophilus strains, E-94507 and CSCC 5361. The binding abilities of both strains were relatively low; E-94507 bound 18.2% and CSCC 5361

ED

20.7%, respectively.

In a study of Oluwafemi and Da-Silva (2009) Lb. acidophilus was inoculated with artificially contaminated

PT

maize grains. An autoclaving treatment of Lb. acidophilus did not seem to have a significant effect on the

CE

detoxification of aflatoxins. During 72 hours incubation detoxification seemed stable, being 35-40%.

In a later study of Oluwafemi et al. (2010) Lb. acidophilus, isolated from fermented maize, was tested for the

AC

detoxification of aflatoxin in maize grains contaminated artificially with different levels of aflatoxin. When aflatoxin levels increased, binding share declined from 54% to 23%.

Pizzolitto et al. (2012) studied AFB1 binding of Lb. acidophilus 24 in the presence of fumonisin B1. Despite different fumonisin levels, Lb. acidophilus bound 15 – 24% of added aflatoxin. At a higher aflatoxin level, binding was slightly less, 15-17% than at a lower level, being 19 – 24%. Different AFB1 levels did not have any effect on fumonisin binding, thus the binding sites are not likely competing.

In a study of Peltonen et al. (2001) two Lb. amylovorus strains (CSCC 5197 and CSCC 5160) bound 57.8% and 59.7% of the added AFB1, respectively. Different incubation times showed that the binding was rapid, being 52.6% and 66.5% at 0 hours for CSCC 5197 and CSCC 5160, respectively. The effect of the

12

ACCEPTED MANUSCRIPT incubation time to the binding was not clear as CSCC 5160 improved binding during 72 hours whereas the CSCC 5197 binding levels were not significantly different at 48 and 72 hours in comparison with 0 hours. In washing test with PBS both Lb. amylovorus strains released bound aflatoxin during washing. Lb. amylovorus

T

CSCC 5197 bound initially approximately 68% and after 5 washes only 17% remained. Lb. amylovorus

RI P

CSCC 5160 released most of the bound aflatoxins, decreasing from an initial level of approximately 51% to

SC

below 5%.

Zinedine et al. (2005) isolated Lb. brevis 1 from sourdough. The strain bound remarkably poorly having only

MA NU

4.46% reduction in the AFB1 level.

Oluwafemi and Da-Silva (2009) studied the detoxification ability of Lb. brevis, isolated from a traditionally fermented maize product, using viable and heat-killed cells. No binding was detected at 0 hours. The binding level increased during the incubation from 33% at 24 hours to 75% at 48 and 72 hours for heat treated cells.

ED

After 24 hours incubation with viable cells the binding rate was stable at 33%. Heat-treated cells bound

PT

better than viable cells.

Oluwafemi et al. (2010) tested the effect of different aflatoxin levels on binding by Lb. brevis. Linear pattern

CE

was not found on the toxin level and binding. At the lowest tested level, 50 ng/g of AFB, the binding was 42%

AC

and at highest level, 500 ng/g, 23%, respectively.

Figure 2 illustrates AFB1 binding results from analyses with Lactobacillus acidophilus, Lb. amylovorus and Lb. brevis.

El-Nezami et al. (1998a) tested the incubation time effect on AFB1 binding with Lb. casei Shirota YIT 9018 in a liquid culture. Initial binding at 0 hours was 33.2% and increased significantly during 72 hours incubation up to 57.6%.

Haskard et al. (2001) washed Lb. casei Shirota YIT901 repetitively. Lb. casei Shirota could not retain the bound AFB1 after five washes with water in viable or acid treated bacteria cells. The heat treated cells

13

ACCEPTED MANUSCRIPT retained 25% of the initial level after five washes. Initial binding was relatively low; 21.8% with viable cells, 41.5% with heat treated and 32.3% with acid treated cells.

T

Zinedine et al. (2005) studied the ability of Lb. casei strains, isolated from sourdough, to remove AFB1 at

RI P

different temperatures and pH levels. When the pH was adjusted to 4.5 and 3, the binding decreased considerably to 3% and 2%, respectively. Incubation at 15 °C decreased the binding level below 5%. Binding

SC

was highest at initial pH 6.5 at 30 °C being 22%. At 25 °C binding remained approximately at the same level

MA NU

and at 30 °C, when the initial pH was 5.5, the binding rate was 20%.

Hernandez-Mendoza et al. (2009a) screened eight different Lb. casei strains for their ability to bind AFB1. After four hours incubation Lb. casei L30 bound the highest percentage, 49% of the added AFB1. After three washes, 42% of the added AFB1 remained bound in the cells. Lb. casei 7R1 and AFB1 complex had the greatest stability, the binding rate being 25% initially and after three washes. For Lb. casei L30, during

ED

repetitive three washes, the loosely bound AFB1 was released during the first wash and only a slight release

PT

was detected after the second wash.

All the Lb. casei strains in the same study were tested for their ability to bind AFB1 in the presence of

CE

different concentrations of bile salts. The presence of bile salts significantly improved the ability of the bacteria to bind AFB1 and the previous differences between strains were reduced. The highest binding levels

AC

were obtained for Lb. casei 21/1 and DPC 3968, corresponding to 80% binding at bile salts presence of 0.05% and 90% binding at 0.1% of bile salts, respectively. (Not presented in figures).

In another study of Hernandez-Mendoza et al. (2009b) Lb. casei Shirota and Lb. casei defensis increased the AFB1 binding share during the incubation period of four hours from an initial share of 15% to 68% and 35% to 60%, respectively. The binding share decreased in both strains after 12 hour incubation. After four PBS washes, Lb. casei Shirota retained bound aflatoxins better than Lb. casei defensis. The effect of pH on binding was tested and the initial binding share at 0 hours was weakest at all three pH levels (6.0, 7.2 and 8.0). After 4 hours incubation at pH 7.2, Lb. casei Shirota bound almost 70% of the added AFB1.

14

ACCEPTED MANUSCRIPT Oluwafemi and Da-Silva (2009) studied the effect of Lb. casei, isolated from a traditional fermented maize product, on detoxification of maize grains artificially contaminated with an aflatoxigenic strain of A. flavus. Lb. casei did not exhibit rapid binding at 0 hours but improved binding during the incubation period. Heat treated

RI P

T

Lb. casei cells bound better than the non-heat treated cells after 72 hours, 63% and 40%, respectively.

In another study Oluwafemi et al. (2010) evaluated the detoxifying ability of Lb. casei against different levels

SC

of AFB1. Artificially contaminated maize samples were inoculated with Lb. casei, isolated from a maize ferment. The removal of AFB1 was assessed after five days of incubation. The highest AFB1 removal level,

MA NU

44.5%, was detected in the maize sample contaminated at the lowest level of toxin (50 ng/g).

Figure 3 illustrates the binding results from the studies done with Lactobacillus casei and AFB1.

Peltonen et al. (2000) tested supernatants from two strains of Lb. crispatus, M247 and MU5, for their AFB1

ED

binding abilities. Lb. crispatus M247 bound only 5% and MU5 20% of the added aflatoxin. Interestingly, MU5

PT

is a spontaneous non-hydrophobic mutant of M247.

Haskard et al. (2001) studied the surface binding of AFB1 by Lb. delbrueckii subsp. bulgaricus. Repetitive

CE

aqueous extractions were used to evaluate the stability of the complexes formed between AFB1 and Lb. bulgaricus, both viable and nonviable (heat- or acid-treated). Most AFB1 was bound and retained by non-

AC

viable bacteria. Acid treated bacterial cells bound 75.8% and retained 35.8% whereas viable cells bound only 15.6% and retained 1.9% of the AFB1.

The binding ability of Lb. delbrueckii was tested in the study of Peltonen et al. (2001). Lb. delbrueckii bound 17.3% of the added AFB1 during 24 hours incubation at 37 °C.

Oluwafemi and Da-Silva (2009), compared heat-killed and non-heat treated Lb. delbrueckii cells for their binding abilities. From the added amount of 80 ng/g of AFB1 the viable Lb. delbrueckii cells bound 50% whereas the heat killed bacteria performed slightly better. Killed cells bound 56% of the aflatoxin during 72 hours incubation.

15

ACCEPTED MANUSCRIPT Oluwafemi et al. (2010) studied the effect of different AFB1 levels on binding. Lb. delbrueckii bound best at lowest levels (43%) but at all tested levels, the binding was 29% or above.

RI P

Peltonen et al. (2001). Lb. fermentum bound 22.6% of the AFB1.

T

The ability of Lb. fermentum to bind AFB1 from a contaminated solution was assessed in the study of

SC

El-Nezami et al. (1998a) assessed the ability of Lb. gasseri to remove AFB1 from liquid media. Initial binding at 0 hours was 58.1%, decreasing to 41.7% after 48 hours incubation and then increased to 67.7% after 72

MA NU

hours incubation.

Haskard et al. (2001) assessed viable, heat-treated and acid treated Lb. helveticus cells for their initial binding capacity and ability to retain aflatoxin after five repetitive washes. Only acid treated cells showed

ED

both binding and retaining abilities of AFB1, shares being 58.1% and 26.5%, respectively.

Peltonen et al. (2001) evaluated Lb. helveticus ability to bind AFB1. The strain bound 34.2% during 24 hours

PT

incubation at 37 °C.

CE

In the study of Peltonen et al. (2000) Lb. johnsonii was the best binder of AFB1 after 24 hours incubation with a binding share of 38.8%. In the further study of Peltonen et al. (2001) Lb. johnsonii (CSCC 5142) bound

AC

30.1% of the added AFB1.

Hernandez-Mendoza et al. (2009b) evaluated the aflatoxin binding abilities and complex stability of Lb. johnsonii. Initial binding at 0 hours was 35% and decreased during the 12 hours incubation to 25%. After four washes, 20% or more remained bound during the incubation period of 12 hours.

Zinedine et al. (2005) isolated two Lb. lactis strains (Lb5 and Lb8) from sourdough. These strains were assessed for their AFB1 binding abilities separately at different pH levels and temperatures. Binding at 30 °C in pH 6.5 was 16.8% for Lb5 and 20.2% for Lb8, respectively. Decreasing pH had a direct decreasing effect on binding. The best binding share for both strains was at higher pH. In any of the circumstances, the binding did not exceed 20%.

16

ACCEPTED MANUSCRIPT

In the study of Peltonen et al. (2000) a probiotic strain of Lb. paracasei bound almost 30% of the added AFB1

T

during 24 hours incubation.

SC

gasseri, Lb. helveticus, Lb. johnsonii, Lb. lactis and Lb. paracasei.

RI P

Figure 4 illustrates binding results of AFB1 from strains of Lb. crispatus, Lb. delbrueckii, Lb. fermentum, Lb.

In the study of Haskard et al. (2001), neither viable nor heat- or acid-treated Lb. plantarum could not retain

MA NU

the bound aflatoxins very well. In line with other strains and studies, acid treated Lb. plantarum bound best (62.7%). The viable cells reached 29.9% binding but could scarcely retain the bound aflatoxins. After 5 washes, acid treated cells retained only 16.0% of the bound aflatoxins.

ED

Lb. plantarum (E-79098) bound 28.4% of the added AFB1 (Peltonen et al. 2001).

Two Lb. plantarum strains (Lb7 and Lb9), both isolated from sourdough, were evaluated for their AFB 1

PT

binding abilities (Zinedine et al. 2005). Both strains performed poorly, binding only 2.1% and 5.2%,

CE

respectively.

Oluwafemi and Da-Silva (2009) assessed a Lb. plantarum strain, isolated from fermented maize, for its

AC

binding ability as viable and heat killed cells. Heat treated Lb. plantarum cells demonstrated notable binding after 48 hours incubation, 95% of the added aflatoxins were removed from the artificially inoculated maize grains. Also viable cells demonstrated 75% binding after 48 hours incubation.

In a further study of Oluwafemi et al. (2010) Lb. plantarum strain was assessed at different levels of aflatoxin. The strain performed relatively reliably throughout different AFB1 levels, the binding being from 35 to 45%.

Lb. reuteri bound relatively well in the study of Hernandez-Mendoza et al. (2009b). An initial binding was almost 60% at 0 hours and increased up to 80% during 12 hours incubation at pH 7.2. After four washes, the binding remained at remarkable levels and thus Lb. reuteri was further tested at two different pH levels. The

17

ACCEPTED MANUSCRIPT best binding was detected at pH 7.2. The binding abilities at pH 6.0 did not exceed 40% decreasing from an initial 40% down to 25% during 12 hours incubation.

T

El-Nezami et al. (1998a) tested aflatoxin binding abilities of Lb. rhamnosus GG (ATCC 53103) and LC705

RI P

during 72 hours incubation. Both strains bound well throughout the incubation period, the shares remaining relatively stable. Lb. rhamnosus GG bound up to 81% and Lb. rhamnosus LC705 up to 82%, and both were 10

CFU/ml,

SC

further tested. Different bacterial concentrations were evaluated and at a concentration of 2x10

the AFB1 level was reduced to less than 0.1% with Lb. rhamnosus GG and 13% with Lb. rhamnosus LC705. 9

MA NU

Significant removal required a concentration of 2x10 CFU/ml. In line with other studies, heat killed bacteria removed aflatoxins better, 81% and 82% for GG and LC705, respectively. El-Nezami et al. (1998a) found that both temperature and bacterial concentration affected the efficiency of removal. A maximal removal was obtained at temperature of 37 °C. Freeze-drying reduced the binding capacity for both strains. Heat killed

ED

bacteria removed toxin better (over 80%) than the viable or freeze dried bacteria.

El-Nezami et al. (1998b) evaluated Lb. rhamnosus GG (ATCC 53103) and LC-705 for their binding abilities

PT

after different chemical and physical treatments. Without any treatments both Lb. rhamnosus GG and Lb. rhamnosus LC-705 bound over 50% of the AFB1. Acid and heat treatments significantly increased binding

CE

abilities of both strains. Alcohol treatment did not increase binding. Alkali treatment reduced the binding significantly. The most effective treatment was HCl/ H2O incubation at pH 2 and the temperature of 37 °C:

AC

binding reached 99.9% for GG and 96.6% for LC705.

Haskard et al. (2000) studied the binding abilities of lyophilized Lb. rhamnosus GG in the late exponential early stationary phase as viable, heat and acid treated. In the presence of NaCl, binding increased. In the presence of CaCl2 the binding was approximately 80%. Decreased pH did not affected on the AFB1 binding but the AFB2 binding improved significantly. This phenomenon was similar for all viable and pre-treated bacteria. (Not presented in figures).

In the study of Haskard et al. (2001), Lb. rhamnosus strains GG (ATCC 53103) and LC-705 (DSM 7061) were tested for their binding abilities. Both strains had significantly better binding abilities as viable, heat and acid treated cells compared with any other strain tested in the same study. These strains also retained the

18

ACCEPTED MANUSCRIPT bound aflatoxin significantly better than other strains. Up to 71% of the total AFB1 was found to remain bound after fifth extraction in acid treated cells. In addition, around 90% of the bound AFB1 was recovered by a solvent extraction. Alone a change at pH (2 to 10) or temperature (4 to 37 °C) did not show any effect on the

RI P

T

release of bound AFB1.

Peltonen et al. (2001) studied the binding abilities of three strains of Lb. rhamnosus: Lc. ⅓, E-97800 and

SC

CSCC 2420. Lb. rhamnosus Lc. ⅓ was the best binder with 54.6% binding, Lb. rhamnosus CSCC 2420 reached up to 33.1% and Lb. rhamnosus E-97800 to 22.7%. Only the best binder Lb. rhamnosus Lc. ⅓ was

MA NU

further tested. During incubation, the binding seemed to be reversible and decreasing.

Turbic et al. (2002) studied the removal of different components from a solution, including AFB1 in the presence of heterocyclic amine 3-amino-1,4–dimethyl-5H–pyrido[4,3-b]indole (Trp-P-1) by Lb. rhamnosus GG and LC-705 as viable, heat- and acid-treated cells. In line with other studies, heat- and acid treated

ED

bacterial cells bound better than the viable cells. The binding of Lb. rhamnosus GG was slightly better than LC-705, both having still remarkable binding shares from 77% to 92%. The presence of Trp-P-1 improved

PT

the binding of AFB1 thus no competition for the same binding spots was apparent.

CE

Zinedine et al. (2005) studied five Lb. rhamnosus strains isolated from sourdough for their AFB1 removal abilities. Binding varied between 23% and 45%. These strains were further tested at different temperatures

AC

and pH. In general, binding was most efficient at 25 °C, decreased at 37 °C and decreased significantly further at 15 °C. A lower pH 3 and 4.5 reduced the binding below 5% and binding was most efficient at pH 5.5.

Halttunen et al. (2008) studied the AFB1 binding abilities of two Lb. rhamnosus strains, GG and LC705 solution. Neither performed very well in comparison with other studies: Lb. rhamnosus GG bound around 10% and Lb. rhamnosus LC705 slightly less.

Figures 8, 9 and 10 illustrate results from various binding tests with different Lactobacillus rhamnosus strains.

19

ACCEPTED MANUSCRIPT Peltonen et al. (2000) studied the binding of a probiotic Lb. salivarius (LM2-118). The precultured Lb. salivarius bound approximately 17% of the added AFB1 of 7.5 µg during 24 hours incubation.

T

Figure 5 illustrates Lactobacillus plantarum, Lb. reuteri and Lb. salivarius binding results with AFB1 from

RI P

various studies.

SC

4.2 Bifidobacterium

MA NU

Peltonen et al. (2000) studied the binding of AFB1 by Bb. lactis (Bb-12). The binding performance was relatively poor, producing a reduction of approximately 18%.

In a further study Peltonen et al. (2001) studied additional strains of Bifidobacterium. Bb. lactis (CSCC 5094), Bb. longum (CSCC 5304), Bb. animalis (CSCC 1941) and Bb. lactis (CSCC 1906) bound AFB1 relatively

ED

well, namely 34.7%, 37.5%, 45.7% and 48.7%, respectively.

PT

A probiotic bacteria strain, Bb. breve Bbi99/E8 was studied for its binding ability of AFB1 in solution

CE

(Halttunen et al. 2008). Among the studied probiotics, this strain was the best binder with a share of 21.4%.

AC

Bb. bifidum bound 35% to 60% of the added AFB1 during 12 hours incubation (Hernandez-Mendoza et al. 2009b). The complex stability was analyzed after repetitive washes and the binding share was reduced after all three washes, but remained at 25% or greater. Bb. bifidum bound best after four hours incubation but after 12 hours almost half of the bound AFB1 was released.

4.3 Lactococcus

Haskard et al. (2001) studied the abilities of Lc. lactis subsp. cremoris and Lc. lactis subsp. lactis to bind AFB1 initially and retain the aflatoxin after 5 washes as viable, heat- and acid-treated bacteria cells. Lc. lactis subsp. lactis retained almost half of the bound aflatoxins after 5 washes whereas Lc. lactis subsp. cremoris released almost all the bound toxins. Lc. lactis subsp. lactis was significantly better binder (59–69%) than Lc. lactis subsp. cremoris (27–43%).

20

ACCEPTED MANUSCRIPT

Two Lc. lactis ssp. cremoris strains, MK4 and ARH74 and Lc. lactis ssp. lactis E-90414 were studied for their AFB1 binding abilities in the study of Peltonen et al. (2001). The difference in binding shares between Lc.

T

lactis ssp. cremoris strains was substantial; strain MK4 bound only 5.6%, ARH74 bound 41.4% and Lc. lactis

RI P

ssp. lactis E-90414 bound 31.6%.

SC

Shahin (2007) studied the binding of Lactococcus strains. Out of 27 Lactococcus strains isolated from yoghurt, raw milk and cheese, five had detectable aflatoxin binding abilities. The most efficient strain was

MA NU

further studied and identified as Lc. lactis. Heat killed cells (boiled and autoclaved) bound AFB1 significantly better than the viable cells, 86%, 80% and 55%, respectively. Viable cells released approximately half of the bound aflatoxins, thus heat killed cells retained toxin complexes better. AFB1 binding from oils was studied and the heat killed cells bound all the added toxins from edible oils. Washing of viable, boiled and autoclaved cells with buffer, methanol and chloroform released some of the bound aflatoxins. Washing with methanol

ED

and chloroform released relatively little of the bound aflatoxins.

CE

PT

4.4 Leuconostoc

AC

Zinedine et al. (2005) found L. mesenteroides to be a poor binder removing only 2.15% of added AFB1.

4.5 Pediococcus

Zinedine et al. (2005) studied the binding abilities of P. acidilactici (P55). This strain was a poor binder having only a 1.8% reduction at the AFB1 level.

4.6 Propionibacterium

El-Nezami et al. (1998a) studied the binding ability of P. freudenreichii ssp. shermanii JS during 72 hours incubation period. The binding was relatively moderate throughout the incubation period starting from an initial 46.4% binding and declined to 35.7% after 72 hours.

21

ACCEPTED MANUSCRIPT In a binding and complex stability study of Haskard et al. (2001) P. freudenreichii subsp. shermanii JS had a significant difference between the binding performance of viable and killed, heat and acid treated bacteria cells. The acid treated cells bound initially over 80% and retained over 50% of the bound aflatoxin during

RI P

T

washing.

The binding effectiveness of P. freudenreichii ssp. shermanii JS of AFB1 from an aqueous solution was

SC

studied by Halttunen et al. (2008). The strain bound 12.5% during 60 minutes incubation at 37 °C.

MA NU

4.7 Streptococcus

S. thermophilus was tested for its binding ability and stability of the formed complex with AFB1 after repetitive washing (Haskard et al. 2001). The acid treated cells bound best, 63.8% retaining approximately half of the

ED

bound AFB1 after five washes.

Shahin (2007) isolated 15 S. thermophilus strains from raw milk and yogurt and studied the abilities of these

PT

strains to bind AFB1. Seven strains showed potential binding abilities. The highest AFB1 binder was further

CE

tested and the viable cells bound 81%. Heat-killed, boiled and autoclaved cells bound even better, 100% and 83%, respectively. These cells were further tested for the complex stability. According to the results, only

AC

chloroform was able to elute the bound AFB1 from the S. thermophilus bacteria cells.

Figure 6 illustrates binding results of AFB1 from strains of Bifidobacterium, Lactococcus, Leuconostoc, Pediococcus, Propionibacterium and Streptococcus.

4.8 Mixtures of LAB strains

Oluwafemi et al. (2010) studied a mixture of Lactobacillus, Lb. acidophilus, Lb. brevis and Lb. plantarum for aflatoxin binding at different aflatoxin levels in artificially contaminated maize grains. At low aflatoxin levels (54 ng/g and 140 ng/g) the mixture bound less than the individual Lb. acidophilus and Lb. brevis. Only Lb. plantarum was alone weaker than the mixture. At higher levels of AFB1 (245 ng/g and 588 ng/g), binding was approximately similar among the individual strains.

22

ACCEPTED MANUSCRIPT

Halttunen et al. (2008) studied a mixture of four different probiotic bacteria strains for their ability to bind AFB1 from a solution. The probiotics tested and their individual binding shares were: Lb. rhamnosus GG

T

15.5% and Lb. rhamnosus LC705 51.5%, P. freudenreichii shermanii JS 23.9% and Bb. breve Bbi99/E8

RI P

8.9%. The mixture bound around 15% whereas theoretical share was higher.

5.1 Lactobacillus

MA NU

SC

5. Binding of aflatoxin M1

In binding of AFM1 from PBS Lb. acidophilus LA1 showed a binding ability of 18.3% in viable (and 25.5% in heat-killed cells (Pierides et al. 2000).

ED

El Khoury et al. (2011) studied the ability of Lb. bulgaricus to reduce AFM1 from PBS and yoghurt. Binding was 40% after two hours PBS incubation and increased up to 87.6% after 14 hours. In yoghurt the AFM1

PT

binding reached up to 60% after 6 hours yoghurt incubation.

CE

Sarimehmetoğlu and Küplülü (2004) analyzed commonly used yoghurt bacteria, Lb. delbrueckii subsp. bulgaricus for its binding ability of AFM1 in PBS and in milk. Binding was better in milk (27.6%) than in PBS

AC

(18.7%) after 4 hours incubation at 37 °C.

Pierides et al. (2000) tested Lb. gasseri for its ability to remove AFM1 from liquid PBS during 15 to 16 hours incubation at 37 °C. Heat killed bacteria had a better AFM1 binding ability than the viable bacteria, 61.5% and 30.8%, respectively.

Pierides et al. (2000) studied the abilities of Lb. rhamnosus GG (ATCC 53013), Lb. rhamnosus LC-705 and Lb. rhamnosus 1/3 to bind AFM1 from PBS. Lb. rhamnosus GG bound over 50% of the AFM1 in PBS in all tested forms (precultured, freeze dried, viable and heat killed). Viable Lb. rhamnosus LC705 bound around 45 -46% and the heat-killed more than 50%. The heat killed Lb. rhamnosus 1/3 strain bound 40% and the viable 18% of the added AFM1. Lb. rhamnosus GG and LC-705 were further tested in skim milk and in full cream milk. Lb. rhamnosus GG bound with limitations: viable cells bound 19% of AFM1 in skim milk and 26% 23

ACCEPTED MANUSCRIPT in full cream milk. The heat killed Lb. rhamnosus GG bound 27% of AFM1 in skim milk and 37% in full cream milk. The viable Lb. rhamnosus LC-705 bound over 60% of the AFM1 in skim and full cream milk when the

T

binding share of heat-treated cells remained at around 30%.

RI P

5.2 Lactococcus

SC

Viable and heat killed Lc. lactis ssp. cremoris (ARH74) strain removed 40.4% and 38.9% of AFM1,

MA NU

respectively, from PBS (Pierides et al. 2000).

5.3 Streptococcus

ED

Sarimehmetoğlu and Küplülü (2004) studied the aflatoxin binding by a S. thermophilus strain in reconstituted

PT

milk and PBS. The binding was better in milk than in PBS, 39.16% and 29.42%, respectively.

El Khoury et al. (2011) studied the AFM1 binding by a S. thermophilus strain in PBS and yoghurt. In PBS, the

CE

binding share increased from an instantaneous 20% up to 60% in 14 hours. During 6 hours incubation in

AC

yoghurt, the binding share of 37.7% was reached.

5.4 Mixtures of LAB

Sarimehmetoğlu and Küplülü (2004) used a yoghurt mixture (S. thermophilus and Lb. delbrueckii subsp. bulgaricus) to study the AFM1 binding during yoghurt fermentation. The mixture bound only 15% of the AFM1 added to the yoghurt.

El Khoury et al. (2011) studied the ability of yoghurt culture mixture (S. thermophilus and Lb. delbrueckii subsp. bulgaricus) to remove AFM1 from PBS and yoghurt. In both matrices binding increased during 6 hours incubation and reached approximately 45% of AFM1 removal level. In PBS the incubation was continued up to 14 hours and the binding share of the mixture reached almost 65%.

24

ACCEPTED MANUSCRIPT Figure 7 illustrates results from binding tests of AFM1 by various LAB strains.

T

6. In vivo and Caco-2 cell model studies of AFB1 and AFM1

RI P

El-Nezami et al. (2000) studied the AFB1 removal in vivo by Lb. rhamnosus GG and LC-705 from the liquid medium using a chicken intestinal loop technique. The removal effect was instant within 1 minute and the

SC

share of bound AFB1 increased during 60 minutes incubation. Chickens received 1.5 µg of AFB1 injected into the duodenum with or without additional Lb. rhamnosus GG or LC-705. The chicken loops were injected

MA NU

bacteria-aflatoxin complex pellets to test the AFB1 bacterial complex stability. In the presence of Lb. rhamnosus GG the uptake of AFB1 was reduced over 70%. The reduction of uptake in the presence of Lb. rhamnosus LC-705 was less, being 37%. Thus, Lb. rhamnosus GG was a significantly more efficient AFB1 reducer than LC-705.

ED

Using the same technique as in the earlier study, El-Nezami et al. (2000) analyzed the AFB1 uptake in the duodenal tissue of chicken in the presence of P. freudenreichii ssp. shermanii JS. The reduction in uptake

CE

PT

was found to be 63%.

Kankaanpää et al. (2000) studied the adhesion properties of Lb. rhamnosus GG in a Caco-2 cell model. The

strain.

AC

adhesion of probiotic Lb. rhamnosus GG was reduced from 30% to 5% when AFB1 was removed by the

Gratz et al. (2004) assessed Lb. rhamnosus GG adherence of the bacteria-aflatoxin complex to the intestinal mucus and the AFB1 binding after the cell adherence to the mucus. AFB1 and the bacteria complex had a reduced binding share to the mucus and the bacteria that had adhered to the mucus bound significantly less AFB1.

In an in vitro study Gratz et al. (2004) used a probiotic bacteria mixture of P. freudenreichii subsp. shermanii JS and Lb. rhamnosus LC-705 to assess the formed AFB1 bacteria complex interference in mucus and the binding of AFB1 to bacteria already adhered to mucus. The results suggested that the AFB1 binding of P.

25

ACCEPTED MANUSCRIPT freudenreichii subsp. shermanii JS and Lb. rhamnosus LC-705 occurs simultaneously in the presence of mucus and is more susceptible to the binding interfering factors and thus perform poorer binding results.

T

Gratz et al. (2005) assessed a probiotic mixture of P. freudenreichii subsp. shermanii JS and Lb. rhamnosus

RI P

LC-705 in vitro and in chicks’ duodenal content. Binding was highest (66%) after 1 minute decreasing to 56.8% after 30 minutes incubation. The bound AFB1 was recovered from the pellets with chloroform after 1

SC

minute and 30 minute incubation, but only 45.5% and 38.6% was recovered, respectively. The bacteria mixture reduced the uptake of AFB1 both in the lumen and duodenal tissue. Recoveries from the lumen and

MA NU

duodenal tissue were higher with bacteria mixture than with separate strains. A longer incubation time reduced the recovery suggesting that AFB1 absorption continued throughout the incubation period.

One visible effect of the aflatoxin exposure is a loss of weight of the host (Gratz et al. 2006; Hathout et al. 2011). Gratz et al. (2006) investigated the effect of Lb. rhamnosus GG to reduce the weight loss of test

ED

animals upon exposure to aflatoxins. A weight loss was observed in a group treated alone with AFB1. Since Lb. rhamnosus GG alone did not have any effect on the weight, it was concluded that this probiotic strain

PT

may reduce the availability of the AFB1 in the gastrointestinal tract and reduce its toxic effects.

CE

Using the Caco-2 model Gratz et al. (2007) investigated the binding of AFB1 during 1 hour and 24 hours incubation with two different Lactobacillus rhamnosus GG levels. The binding was a rapid process, 10

AC

measuring 61% for bacteria concentration of 5x10

10

CFU/ml and 40% for 1x10

CFU/ml after 1 hour

incubation. After 24 hours, no significant changes occurred at the binding levels.

Hernandez-Mendoza et al. (2010) investigated the potential protective effect of Lb. casei Shirota against AFB1 in a murine model. The ability of Lb. casei Shirota to bind AFB1 was determined by fluorescent monoclonal antibody staining and the bacteria – AFB1 interactions were assessed by atomic force microscopy. The images showed that AFB1 binds into the bacterial cell envelope. The binding was found to cause structural changes that modify the cell surface of bacteria. Lb. casei Shirota supplementation effect on intestinal absorption of the toxin was evaluated by detecting biological markers AFB1-Lys adducts from blood during a 3-week aflatoxin exposure in a murine model. Rats receiving AFB 1 plus bacteria showed

26

ACCEPTED MANUSCRIPT significantly lower level of AFB1-Lys adducts in serum compared to rats treated only with AFB1.The results suggested that Lb. casei Shirota is able to bind aflatoxins at intestinal level, thus preventing their absorption.

T

Hernandez-Mendoza et al. (2010) reported protective properties of Lb. casei Shirota against aflatoxin

RI P

induced damage in rats. Aflatoxin B1-lysine adducts levels decreased in serum after the Lb. casei Shirota treatment. Usage of LAB as a biocontrol method is supported also by the rat study of Hathout et al. (2011)

SC

who observed that Lb. casei and Lb. reuteri can provide protection against AFB1 induced oxidative stress. A dietary treatment with live Lb. casei and Lb. reuteri after aflatoxin contaminated diet succeeded to prevent

MA NU

the liver and kidney injuries in test animals. Lb. reuteri was more effective in the aflatoxin toxicity reduction indicating that it was also a better binder and/or has more efficient antioxidant properties.

Kabak and Ozbey (2012a) studied the bioavailability of AFM1 in a digestion model with six probiotic bacteria. The bioaccessability of AFM1 from spiked and naturally contaminated milk was from 80% to 86%.The

ED

presence of probiotic bacteria reduced the AFM1 bioaccessability by 15.5% to 31.6%. Lb. acidophilus NCFM 150B was the best binder and Lb. casei Shirota poorest in all tested AFM1 concentrations (0.042 µg/L to

PT

0.882 µg/Liter).

CE

The results from Deabes et al. (2012) showed that aflatoxins were genotoxic in the bone marrow and spermatocyte cells and had cytotoxic effects on both cell types. Aflatoxin exposure impaired the DNA

AC

synthesis and chromosome segregation and progression through mitosis. Lb. rhamnosus GG consumption before the aflatoxin exposure reduced aflatoxin induced genotoxicity and cytotoxicity in kidney and liver cells.

7. Nature of aflatoxin binding by LAB

In addition to different binding compounds, the binding of aflatoxin by LAB appears extremely strain specific. LAB have different properties and different cell wall compositions, even within the same order. The Lactobacillus genus has been found to have an S-layer made of protein subunits and detailed protein studies have been carried out on Lb. helveticus, Lb. brevis, Lb. acidophilus and Lb. crispatus (Delcour et al. 1999).

27

ACCEPTED MANUSCRIPT The cell wall of one of the most studied LAB, Lb. rhamnosus, has polysaccharides where it binds aflatoxins. These polysaccharides occur in three main forms: cell wall polysaccharide, peptidoglycan and teichoic or

T

lipoteichoic acids (Delcour et al. 1999; Haskard et al. 2000).

RI P

The results from Haskard et al. (2000) showed that the binding of aflatoxin to carbohydrate components may be predominant as largest decrease in the binding ability of Lb. rhamnosus GG cells appeared when the

SC

cells were treated with m-periodate. M-periodate cause oxidation of carbohydrates to aldehydes and carbon acid groups. The results suggested that a carbohydrate binding may have a greater involvement in the viable

MA NU

than acid or heat-treated cells. Hydrophobic interactions are expected at least in heat- and acid treated bacteria where the protein denaturation treatment may expose more hydrophobic binding areas. Salts are expected to influence the bacterial surface charge and the results show that electrostatic interactions do have minor effects on binding. A hydrochloric acid treatment enhanced the bacteria cell binding nearly up to

ED

100%. Acids react with AFB1 and form aflatoxin B2.

An involvement of lipids in the binding process is unlikely (Haskard et al. 2000) and exopolysaccharides

PT

have only remote binding abilities (Hernandez-Mendoza et al. 2009a, 2009b). According to Haskard et al. (2000) proteins may have some involvement in the binding. Lahtinen et al. (2004) ruled exopolysaccharides, 2+

and Mg

2+

CE

cell wall proteins and Ca

out in the AFB1 binding of Lb. rhamnosus GG. Neither chelators nor

AC

divalent cations affected the binding of AFB1 and these binding points do not compete with aflatoxins.

Despite decreasing binding effects of Lb. rhamnosus GG after some treatments, the binding still occurred in substantial amounts suggesting an involvement of multiple binding components in cells (Haskard et al. 2000). The study of Hernandez-Mendoza et al. (2009a) supports the fact that the binding phenomenon is not caused by one factor but is a miscellaneous system of supportive bacteria cell wall components.

Acid treated bacteria cells showed higher capacities to bind existing AFB1 initially and also to keep the toxins bound even after five aqueous washes (Haskard et al. 2001). Pierides et al. (2000) showed that the heat killed bacteria cells can bind AFM1 better than the viable cells, also in the same strain. AFB1 binding appears to occur to a larger extent compared to AFM1. This reduced binding may be due to the fact that the AFB1

28

ACCEPTED MANUSCRIPT derivate AFM1 has an additional hydroxyl group and a lower hydrophobicity and thus is not removed as efficiently as AFB1 (Turbic et al. 2002).

T

Aflatoxin binding by LAB is considered as an extracellular and weak, noncovalent bond (Haskard et al. 2001)

RI P

between the toxin and cell wall peptidoglycan, the components tightly associated to the peptidoglycan (Lahtinen et al. 2004) and carbohydrate compounds (Haskard et al. 2000). Compounds in the bacteria cell

SC

wall have a significant impact in the bond formation and the bacteria metabolism might reduce the binding effect between bacteria cell and toxin as the non-viable strains have showed in many studies to exert a

MA NU

better binding than live bacteria.

Binding of AFM1 in milk by viable Lb. rhamnosus LC705 was better than by heat killed and even better than in PBS (Pierides et al. 2000).The effect of the matrix remains unknown and viable strains could perform

ED

better in some matrices.

Methanol, acetonitrile, chloroform, benzene and boiling did not appear to have any structural integrity, shape

PT

or size effect on Lb. rhamnosus GG and LC-705 as detected with light microscopy and Gram staining. On the other hand, sodium hydroxide and hydrochloric acid had an effect on the structural integrity of these

CE

strains (Haskard et al. 2001).

AC

In a study of Lb. casei (Zinedine et al. 2005) incubation at 37 °C decreased the binding compared with incubation at 25-30 °C. In the study of Peltonen et al. (2001) the incubation was done at 37 °C and some strains had relatively low binding levels, for example Lb. delbrueckii, only 17.3%. Temperature did not affect releasing of the bound aflatoxins (Haskard et al. 2001).

Haskard et al. (2001) tested the release of bound aflatoxin from bacteria cell using autoclaving and sonication. Autoclaving caused denaturation of bacterial proteins and enzymes but did not cause the most strongly bound AFB1 to be released; suggesting that AFB1 is not bound to loosely attached bacterial cell wall components.

29

ACCEPTED MANUSCRIPT It is likely that AFB1 and intestinal mucus do not have the same binding sites, as heat treated bacteria lose their ability to bind intestinal mucus whereas a heat-treatment of the bacteria cells enhanced the AFB1 binding (Gratz et al. 2004). Proteins must be involved in the mucus binding and carbohydrates in AFB1

T

binding, at least in the more comprehensive studied Lb. rhamnosus GG and LC-705 strains. Especially the

RI P

potential feature for non-viable cells not to bind mucus could decrease the AFB1 bioavailability throughout

SC

the gastrointestinal tract.

Haskard et al. (2001) used ELISA to prove bacteria surface binding of aflatoxins for the first time. They

MA NU

suggested that AFB1 is bound to bacteria by weak non-covalent interactions, such as associating with hydrophobic pockets on the bacterial surface. Haskard et al. (2000) showed binding of AFB1 is higher with non-viable cells. As metabolic activity cannot occur, covalent interactions between cell and toxins cannot be responsible for the binding. pH range of 2.5 – 8.5 had no effect on binding (Haskard et al. 2000). A further study proved that a cation-exchange mechanism is not operating as the binding is unaffected by the pH

ED

variation between 2 to 10 (Haskard et al. 2001).

PT

Bacterial cell wall alone can bind almost the same amount as the complete bacteria, and as viability of the bacterial cell is not essential, the cell wall appears to provide the binding locus to the aflatoxins. Non-viability

CE

of cells implies that the nutrient intake is not present and the cell metabolism does not occur which could

AC

indicate that cell wall is more prone and susceptible to binding.

8. Discussion

Aflatoxins often contaminate maize and peanuts which are commonly grown in tropical developing countries. Aflatoxins can also be found in milk as a result of feeding aflatoxin-contaminated feeds to dairy animals. Especially in the context of climate change, aflatoxins constitute a real risk to human health. An unpredictable weather event can lead to inadequate drying of maize on farms and in households. Such a situation can promote fungal growth leading eventually to formation of aflatoxin. LAB strains are traditionally used in many staple foods in form of fermented maize and dairy products in many parts of the world.

30

ACCEPTED MANUSCRIPT Several LAB strains have shown potential for used as a novel biological method for inhibition of fungal growth and binding of existing aflatoxins in feed and food. In these studies, several potential and commonly used LAB strains have been analyzed. The inhibition of fungal growth by LAB seems highly strain specific

RI P

T

but the stage of fungal growth can also determine the effectiveness of the LAB added.

The inhibition of fungal growth is most likely determined by a combination of different factors. Organic acid

SC

production by LAB seems excluded as a primary and exclusive inhibition mechanism against fungal growth. One possible factor could be the changing living conditions and availability of nutrients. If competition occurs

MA NU

over nutrients in the substrate, this may be due to the fact that bacteria are simpler organisms with a faster metabolism compared with fungi. At least in the culture medium conditions bacteria adapt faster, produce more cell biomass resulting in lower nutrient levels for fungi to grow (Gerbaldo et al. 2012).

Heat treated and killed bacteria show better binding abilities of AFB1 than viable bacteria cells. This binding

ED

ability has benefits as a technological point of view, as bacteria strains which are not traditionally or indigenously involved in the fermentation, could be potentially added to the process without or with only

PT

slight effects to the properties of the final product.

CE

Similar to fungal growth inhibition, the abilities of LAB to bind AFB1 and AFM1 by appear to be highly strain specific. Some of the strains possess greater potential and have presented other beneficial aspects beside

AC

the technological product properties. These properties include health improving properties such as the ability to adhere to the intestinal mucus and colonize the tract. So far, the strains that have demonstrated best potential for AFB1 binding are Lb. acidophilus, Lb. casei, Ld. delbrueckii and Lb. rhamnosus. In the case of AFM1 the binding is less prominent than with AFB1 but still strains of Lb. rhamnosus, Lc. lactis ssp. cremoris, and S. thermophilus have a significant binding potential. All the tested bacterial mixtures have shown relatively low binding shares in comparison to individual strains.

As the aflatoxin binding seems to be highly related to the strain, matrix, temperature, pH, incubation time and related conditions, most likely a specific LAB strain is required against fungal contamination. The binding is most likely a surface phenomenon, with a significant involvement of carbohydrates, peptidoglycan and some level of protein involvement occurring instantly. Cellular lipids do not seem to be involved in the binding

31

ACCEPTED MANUSCRIPT process. Non-viable LAB cells have proven more efficient in binding of AFB1 and thus their binding mechanisms or sites might be different from those in viable cells. The binding seems to be at least a slightly reversible phenomenon. Virtually all studied LAB strains have released bound aflatoxins to some extent. The

T

binding stability is, therefore, one important aspect in evaluation of LAB strains for the aflatoxin binding

RI P

potential. Salts may improve the binding by influencing the cell surface charge. The binding of aflatoxins by LAB is likely to be a sum of multiple individual factors. Further research is needed to elucidate the binding

SC

mechanisms.

MA NU

Lb. plantarum and Lb. fermentum are widely used in silage production and according to several studies these strains have good potential to be used in a novel way for prevention of fungal growth in maize after harvest and during storage. In addition, Lb. rhamnosus is one of the most studied probiotic LAB with a potential exploitation in aflatoxin binding. Both in vivo and in vitro studies have presented encouraging

ED

results to develop this approach further.

Especially Lb. rhamnosus strains GG and LC-705 have the potential to bind aflatoxins and reduce aflatoxin

PT

uptake after consumption of contaminated food. These bacteria cells can form aflatoxin-bacterial cell complexes and could reduce aflatoxin accumulation in the human gastrointestinal tract. The best binders

CE

seem to deliver also health effects as probiotics. The real concern is the adherence of the aflatoxin-bacterial cell complex in the intestinal tract. Evidence shows that after the complex is formed, its adherence to the

is reduced.

AC

mucus is impaired but if the bacteria have already adhered to the mucus, bacteria’s ability to bind aflatoxins

Since mycotoxin absorption takes place in the small intestine (Kabak and Ozbey, 2012b) the formation of a mycotoxin-bacterial cell complex should occur before the contaminated food reaches the small intestine. The results from a limited number of animal studies are encouraging and suggest that LAB consumption with a meal can reduce the aflatoxin absorption.

9. Conclusions

32

ACCEPTED MANUSCRIPT A LAB based biocontrol method seems to present a promising approach for controlling fungal growth in the feed-food chains, hence mitigating the health risks caused by fungal toxins. Aflatoxin binding is a reversible reaction, which occurs on the bacterial surface and involves interaction with carbohydrates, peptidoglycan

T

and to some extent protein structures. Aflatoxin binding seems to be highly related to strain, matrix,

RI P

temperature, pH, incubation time and related conditions. For these reasons, most likely specific LAB strains are required against fungal caused contamination in various applications. Especially, strains of the genus

SC

Lactobacillus are relatively well studied and according to published research reports, provide significant potential to mitigate aflatoxin risks associated with contaminated food and feed matrices. Further research is

MA NU

still needed to screen potential LAB strains from traditional fermented products manufactured in different parts of the world and test their feasibility in different matrices.

Acknowledgment

ED

This article was prepared as a contribution to the FoodAfrica Programme which is mainly financed by the Ministry for Foreign Affairs of Finland. The kind assistance of International Livestock Research Institute ILRI

CE

Author contributions

PT

is acknowledged.

AC

Sara Ahlberg is the principal author and responsible for compiling the article and the other authors have assisted in article structuring and editing.

9. References Bhat, R., Rai, R., Karim, A., 2010. Mycotoxins in food and feed : present status and future concerns. Comprehensive Reviews in Food Science and Food Safety 9, 57–81.

Broberg, A., Jacobsson, K., Ström, K., Schnürer, J., 2007. Metabolite profiles of lactic acid bacteria in grass silage. Applied and Environmental Microbiology 73, 5547–52. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2042065/pdf/2939-06.pdf. Accessed 05 May 2014.

33

ACCEPTED MANUSCRIPT Dalié, D., Deschamps, A., Richard-Forget, F., 2010. Lactic acid bacteria – Potential for control of mold growth and mycotoxins: A review. Food Control 21, 370–380.

T

Deabes, M., Darwish, H., Abdel-Aziz, K., Farag, I., Nada, S., Tawfek, N., 2012. Protective effects of

RI P

Lactobacillus rhamnosus GG on aflatoxins-induced toxicities in male albino mice. Journal of Environmental Analytical & Toxicology 2, 132. DOI http://dx.doi.org/10.4172/2161-0525.1000132.

SC

Delcour. J., Ferain, T., Deghorain, M., Palumbo, E., Hols, P., 1999. The biosynthesis and functionality of the

MA NU

cell-wall of lactic acid bacteria. Antonie van Leeuwenhoek 76, 159–184.

El-Nezami, H., Kankaanpää, P., Salminen, S., Ahokas, J., 1998a. Ability of dairy strains of lactic acid bacteria to bind a common food carcinogen, aflatoxin B1. Food and Chemical Toxicology 36, 321–326.

El-Nezami, Hani., Kankaanpää, P., Salminen, S., Ahokas, J., 1998b. Physicochemical alterations enhance

ED

the ability of dairy strains of lactic acid bacteria to remove aflatoxin from contaminated media. Journal of Food Protection 61, 466–468.

PT

El-Nezami, H., Mykkänen, H., Kankaanpää, P., Salminen, S., Ahokas, J., 2000. Ability of Lactobacillus and

549–552.

CE

Propionibacterium strains to remove aflatoxin B1 from the chicken duodenum. Journal of Food Protection 63,

AC

FAO, 2003. Worldwide regulations for mycotoxins in food and feed in 2003: 3. Mycotoxin regulations in 2003 and current developments. Available from: http://www.fao.org/docrep/007/y5499e/y5499e07.htm. Accessed 26 May 2014.

Gerbaldo, G., Barberis, C., Pascual, L., Dalcero, A., Barberis, L., 2012. Antifungal activity of two Lactobacillus strains with potential probiotic properties. FEMS Microbiology Letters 332, 27–33.

Gourama, H., Bullerman, LB., 1995. Inhibition of growth and aflatoxin production of Aspergillus flavus by Lactobacillus species. Journal of Food Protection 58, 1249–1256.

Gratz, S., Mykkänen, H., El-Nezami, H., 2005. Aflatoxin B1 binding by a mixture of Lactobacillus and Propionibacterium: in vitro versus ex vivo. Journal of Food Protection 68, 2470–2474.

34

ACCEPTED MANUSCRIPT Gratz, S., Mykkänen, H., Ouwehand, A.C., Juvonen, R., Salminen, S., El-Nezami, H., 2004. Intestinal mucus alters the ability of probiotic bacteria to bind aflatoxin B1 in vitro. Applied and Environmental Microbiology 70,

T

1–4.

RI P

Gratz, S., Täubel, M., Juvonen, R.O., Viluksela, M., Turner, P.C., Mykkänen, H., El-Nezami, H., 2006. Lactobacillus rhamnosus strain GG modulates intestinal absorption, fecal excretion, and toxicity of aflatoxin

SC

B1 in rats. Applied and Environmental Microbiology 72, 7398–400.

Gratz, S., Wu, Q., El-Nezami, H., Juvonen, R., Mykkänen, H., Turner, P.C., 2007. Lactobacillus rhamnosus

Environmental Microbiology 73, 3958–64.

MA NU

strain GG reduces aflatoxin B1 transport, metabolism, and toxicity in Caco-2 cells. Applied and

Guengerich, F., 2001. Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chemical Research in Toxicology 14, 611-650. Available from:

ED

http://pubs.acs.org/doi/abs/10.1021/tx0002583 Accessed 26 May 2014.

PT

Halttunen, T., Collado, M., El-Nezami, H., Meriluoto, J., Salminen, S., 2008. Combining strains of lactic acid bacteria may reduce their toxin and heavy metal removal efficiency from aqueous solution. Letters in Applied

CE

Microbiology 46, 160–165.

Haskard, C., Binnion, C., Ahokas, J., 2000. Factors affecting the sequestration of aflatoxin by Lactobacillus

AC

rhamnosus strain GG. Chemico-Biological Interactions 128, 39–49.

Haskard, C., El-Nezami, H., Kankaanpää, P., Salminen, S., Ahokas, J., 2001. Surface binding of aflatoxin B1 by lactic acid bacteria. Applied and Environmental Microbiology 67, 3086–3091.

Hathout, A., Mohamed, S., El-Nekeety, A., Hassan, N., Aly, S., Abdel-Wahhab, M., 2011. Ability of Lactobacillus casei and Lactobacillus reuteri to protect against oxidative stress in rats fed aflatoxinscontaminated diet. Toxicon 58, 179–186.

Hernandez-Mendoza, A., Garcia, H., Steele, J., 2009a. Screening of Lactobacillus casei strains for their ability to bind aflatoxin B1. Food and Chemical Toxicology 47, 1064–1068.

35

ACCEPTED MANUSCRIPT Hernandez-Mendoza, A., Guzman-de-Peña, D., Garcia, H.S., 2009b. Key role of teichoic acids on aflatoxin B1 binding by probiotic bacteria. Journal of Applied Microbiology 107, 395–403.

T

Hernandez-Mendoza, A., Guzman-De-Peña, D., González-Córdova, AF., Vallejo-Córdoba, B., Garcia, H.S.,

RI P

2010. In vivo assessment of the potential protective effect of Lactobacillus casei Shirota against aflatoxin B1. Dairy Science & Technology 90, 729–740.

SC

IARC, 2002. Volume 82 Some Traditional herbal medicines, some mycotoxins, naphthalene and styrene. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Available from:

MA NU

http://monographs.iarc.fr/ENG/Monographs/vol82/mono82.pdf. Accessed 23 October 2013.

Ismaiel, A., Ghaly, MF., El-Naggar, A.K., 2011. Milk kefir: ultrastructure, antimicrobial activity and efficacy on aflatoxin B1 production by Aspergillus flavus. Current Microbiology 62, 1602–9.

ED

FAO, WHO, 1993. Report of the twenty-fourth session of the Codex committee in food additives and contaminants. Joint FAO/WHO Food Standards Programme. The Hague. Available from:

PT

http://www.google.co.ke/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CCoQFjAA&url=http%3A%2F %2Fwww.codexalimentarius.org%2Finput%2Fdownload%2Freport%2F18%2Fal93_12e.pdf&ei=qrdnUuynLs

CE

SUtAbEnYCADw&usg=AFQjCNHn8wXZ0MMZQbFD1rsyTM-M5pDoBw&bvm=bv.55123115,d.Yms.

AC

Accessed 23 October 2013.

FAO, WHO, 1999. Report of the thirtieth session of the Codex committee on food additives and contaminants. Joint FAO/WHO Food Standards Programme. Rome. Available from: http://www.google.co.ke/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CC0QFjAA&url=http%3A%2F %2Fwww.codexalimentarius.org%2Finput%2Fdownload%2Freport%2F24%2FAl99_12e.pdf&ei=DbhnUo3ZK YnEswbN04GQCw&usg=AFQjCNHAqVttR_P-BpRU0UfTqn5VdlPTdw&bvm=bv.55123115,d.Yms. Accessed 15 August 2014.

Kabak, B., Ozbey, F., 2012a. Assessment of the bioaccessibility of aflatoxins from various food matrices using an in vitro digestion model, and the efficacy of probiotic bacteria in reducing bioaccessibility. Journal of Food Composition and Analysis 27, 21–31.

36

ACCEPTED MANUSCRIPT Kabak, B., Ozbey, F., 2012b. Aflatoxin M1 in UHT milk consumed in Turkey and first assessment of its bioaccessibility using an in vitro digestion model. Food Control 28, 338–344.

T

Kankaanpää, P., Tuomola, E., El-Nezami, H., Ahokas, J., Salminen, S.J., 2000. Binding of aflatoxin B1 alters

RI P

the adhesion properties of Lactobacillus rhamnosus strain GG in a Caco-2 model. Journal of Food Protection 63, 412–414.

SC

El Khoury, A., Atoui, A., Yaghi, J., 2011. Analysis of aflatoxin M1 in milk and yogurt and AFM1 reduction by

MA NU

lactic acid bacteria used in Lebanese industry. Food Control 22, 1695–1699.

EU, 2002. Directive 2002/32/EC. Available from: http://eur-

lex.europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:2002L0032:20061020:EN:PDF. Accessed 14 August 2014.

foodstuffs. Available from: http://eur-

ED

EU, 2006. Commission regulation (EC) No 1881/2006, setting maximum levels for certain contaminants in

PT

lex.europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:2006R1881:20100701:EN:PDF. Accessed 14 August 2014.

CE

Kpodo, K., Thrane, U., Hald, B., 2000. Fusaria and fumonisins in maize from Ghana and their co-occurrence

AC

with aflatoxins. International Journal of Food Microbiology 61, 147–157.

Lahtinen, S.J., Haskard, C.A., Ouwehand, A.C., Salminen, S.J., Ahokas, J.T., 2004. Binding of aflatoxin B1 to cell wall components of Lactobacillus rhamnosus strain GG. Food Additives & Contaminants 21, 158–164.

Laitila, A., Alakomi, H-L., Raaska, L., Mattila-Sandholm, T., Haikara, A., 2002. Antifungal activities of two Lactobacillus plantarum strains against Fusarium molds in vitro and in malting of barley. Journal of Applied Microbiology 93, 566–576.

Di Natale, F., Gallo, M., Nigro, M., 2009. Absorbents selection for aflatoxins removal in bovine milks. Journal of Food Engineering 95, 186–191.

Ndagano, D., Lamoureux, T., Dortu, C., Vandermoten, S., Thonart, P., 2011. Antifungal activity of 2 lactic acid bacteria of the Weissella genus isolated from food. Journal of Food Science 76, 305–311. 37

ACCEPTED MANUSCRIPT Niderkorn, V., Boudra, H., Morgavi, D., 2006. Binding of Fusarium mycotoxins by fermentative bacteria in vitro. Journal of Applied Microbiology 101, 849–856.

T

Oluwafemi, F., Da-Silva, F.A., 2009. Removal of aflatoxins by viable and heat-killed Lactobacillus species

RI P

isolated from fermented maize. Journal of Applied Biosciences 16, 871–876.

Oluwafemi, F., Kumar, M., Bandyopadhyay, R., Ogunbanwo, T., Ayanwande, K.B., 2010. Bio-detoxification

SC

of aflatoxin B1 in artificially contaminated maize grains using lactic acid bacteria. Toxin Reviews 29, 115–

MA NU

122.

Onilude, A.A., Fagade, O.E., Bello, M.M., Fadahunsi, I.F., 2005. Inhibition of aflatoxin-producing aspergilli by lactic acid bacteria isolates from indigenously fermented cereal gruels. African Journal of Biotechnology 4, 1404–1408.

ED

Peltonen, K., El-Nezami, H., Haskard, C., Ahokas, J., Salminen, S., 2001. Aflatoxin B1 binding by dairy strains of lactic acid bacteria and bifidobacteria. Journal of Dairy Science 84, 2152–6.

PT

Peltonen, K.D., El-Nezami, H.S., Salminen, S.J., Ahokas, J.T., 2000. Binding of aflatoxin B1 by probiotic

CE

bacteria. Journal of the Science of Food and Agriculture 80, 1942–1945.

Pierides, M., El-Nezami, H., Peltonen, K., Salminen, S., Ahokas, J., 2000. Ability of dairy strains of lactic acid

AC

bacteria to bind aflatoxin M1 in a food model. Journal of Food Protection 63, 645–650.

Pizzolitto, R.P., Salvano, M.A., Dalcero, A.M., 2012. Analysis of fumonisin B1 removal by microorganisms in co-occurrence with aflatoxin B1 and the nature of the binding process. International Journal of Food Microbiology 156, 214–221.

Sarimehmetoğlu, B., Küplülü, Ö., 2004. Binding ability of aflatoxin M1 to yoghurt bacteria. Ankara Üniversitesi Veteriner Fakültesi Dergisi 51, 195–198. Available from: http://dergiler.ankara.edu.tr/dergiler/11/214/1752.pdf. Accessed 14 August 2014.

Shahin, A., 2007. Removal of aflatoxin B1 from contaminated liquid media by dairy lactic acid bacteria. International Journal of Agriculture & Biology 9, 71–75.

38

ACCEPTED MANUSCRIPT Shetty, H.P., Hald, B., Jespersen, L., 2007. Surface binding of aflatoxin B1 by Saccharomyces cerevisiae strains with potential decontaminating abilities in indigenous fermented foods. International Journal of Food

T

Microbiology 113, 41–46.

RI P

Strosnider, H., Azziz-Baumgartner, E., Banziger, M., Bhat, R.V., Breiman, R., Brune, M-N., DeCock, K., Dilley, A., Groopman, J., Hell, K., Henry, S.H., Jeffers, D., Jolly, C., Jolly, P., Kibata, G.N., Lewis, L., Liu, X., Luber, G., McCoy, L., Mensah, P., Miraglia, M., Misore, A., Njapau, H., Ong, C-N., Onsongo, M.T.K., Page,

SC

W.S., Park, D., Patel, M., Phillips, T., Pineiro, M., Pronczuk, J., Schurz Rogers, H., Rubin, C., Sabino, M., Schaafsma, A., Shephard, G., Stroka, J., Wild, C., Williams, J.T., Wilson, D., 2006. Workgroup report : Public

MA NU

health strategies for reducing aflatoxin exposure in developing countries. Environmental Health Perspectives 114, 1898–1903.

Turbic, A., Ahokas, J., Haskard, C., 2002. Selective in vitro binding of dietary mutagens, individually or in

ED

combination, by lactic acid bacteria. Food Additives & Contaminants 19, 144–153.

Turner, P.C., Sylla, A., Gong, Y.Y., Diallo, M.S., Sutcliffe, A.E., Hall, A.J., Wild, C.P., 2005. Reduction in

PT

exposure to carcinogenic aflatoxins by postharvest intervention measures in west Africa : a community-

CE

based intervention study. Lancet 365, 1950-1956.

Valerio, F., Favilla, M., de Bellis, P., Sisto, A., de Candia, S., Lavermicocca, P., 2009. Antifungal activity of

AC

strains of lactic acid bacteria isolated from a semolina ecosystem against Penicillium roqueforti, Aspergillus niger and Endomyces fibuliger contaminating bakery products. Systematic and Applied Microbiology 32, 438–448.

Valerio, F., Lavermicocca, P., Pascale, M., Visconti, A., 2004. Production of phenyllactic acid by lactic acid bacteria: an approach to the selection of strains contributing to food quality and preservation. FEMS Microbiology Letters 233, 289–295.

Vanne, L., Kleemola, T., Haikara, A., 2001. Screening of the antifungal effect of lactic acid bacteria against toxigenic Penicillium and Aspergillus strains. BMS International Symposium on Bioactive Fungal Metabolites - Impact and Exploitation. University of Wales Swansea, British Mycological Society, (p. 100).

39

ACCEPTED MANUSCRIPT Wu, F., Narrod, C., Tiongco, M., Liu, Y., 2011. The health economics of aflatoxin: global burden of disease. Available from: http://www.ifpri.org/sites/default/files/publications/aflacontrol_wp04.pdf. Accessed 14 August

T

2014.

RI P

Wu, F., Tritscher, A., 2011. Aflatoxins: a global public health problem. Available from: http://www.agriskmanagementforum.org/sites/agriskmanagementforum.org/files/WHO%20-%20Aflatoxin-

SC

public%20health%20issue.pdf. Accessed 14 August 2014.

Yang, E., Chang, H., 2010. Purification of a new antifungal compound produced by Lactobacillus plantarum

MA NU

AF1 isolated from kimchi. International Journal of Food Microbiology 139, 56–63.

Zinedine, A., Faid, M., Benlemlih, M., 2005. In vitro reduction of aflatoxin B1 by strains of lactic acid bacteria

ED

isolated from Moroccan sourdough bread. International Journal of Agriculture and Biology 7, 67–70.

PT

List of Figures

Combined from FAO, 2003.

CE

Figure 1. Aflatoxin limits for feed, food and milk distributed between countries.

AC

Figure 2. Aflatoxin binding by selected Lb. acidophilus, Lb. amylovorus and Lb. brevis strains. [2] (El-Nezami et al. 1998a) [1] (Haskard et al. 2001), [6] (Peltonen et al. 2001), [3] (Oluwafemi and Da-Silva 2009), [4] (Oluwafemi et al. 2010) [5] (Pizzolitto et al. 2012), [7] (Zinedine et al. 2005)

Figure 3. Aflatoxin binding by selected Lb. casei strains. [2] (El-Nezami et al. 1998a), [1] (Haskard et al. 2001), [7] (Zinedine et al. 2005), [8] (Hernandez-Mendoza et al. 2009a), [9] (HernandezMendoza et al. 2009b), [3] (Oluwafemi and Da-Silva 2009), [4] (Oluwafemi et al. 2010)

Figure 4. Aflatoxin binding by selected Lb. crispatus, Lb. delbrueckii, Lb. fermentum, Lb. gasseri, Lb. helveticus, Lb. johnsonii, Lb. lactis and Lb. paracasei strains. [10] (Peltonen et al. 2000), [1] (Haskard et al. 2001), [6] (Peltonen et al. 2001), [3] (Oluwafemi and Da-Silva 2009), [4] (Oluwafemi et al. 2010) [2] (El-Nezami et al. 1998a), [9] (Hernandez-Mendoza et al. 2009b), [7] (Zinedine et al. 2005)

40

ACCEPTED MANUSCRIPT Figure 5. Aflatoxin binding by selected Lb. plantarum, Lb. reuteri and Lb. salivarius strains. [1] (Haskard et al. 2001), [6] (Peltonen et al. 2001), [7] (Zinedine et al. 2005), [3] (Oluwafemi and Da-Silva 2009), [4] (Oluwafemi et al.

T

2010), [9] (Hernandez-Mendoza et al. 2009b), [9] (Peltonen et al. 2000)

RI P

Figure 6. Aflatoxin binding by selected Bifidobacterium, Lactococcus, Leuconostoc, Pediococcus, Propionibacterium and Streptococcus strains.

[9] (Peltonen et al. 2000), [6] (Peltonen et al. 2001), [15] (Halttunen et al. 2008), [9] (Hernandez-Mendoza et al. 2009b), [1] (Haskard et

SC

al. 2001), [17] (Shahin, 2007), [7] (Zinedine et al. 2005), [2] (El-Nezami et al. 1998a), [1] (Haskard et al. 2001)

MA NU

Figure 7. Aflatoxin M1 binding by selected LAB strains.

[18] (Pierides et al. 2000), [19] (El Khoury et al. 2011), [20] (Sarimehmetoğlu and Küplülü 2004)

Figure 8. Aflatoxin binding by Lb. rhamnosus GG ATCC 53103 in vitro and in vivo. [2] (El-Nezami et al. 1998a), [11] (El-Nezami et al. 1998b), [1] (Haskard et al. 2001) [14] (Turbic et al. 2002) [16] (El-Nezami et al. 2000)

ED

[12] (Lahtinen et al. 2004)

PT

Figure 9. Aflatoxin binding by Lb. rhamnosus LC-705. [2] (El-Nezami et al. 1998a), [11] (El-Nezami et al. 1998b), [16] (El-Nezami et al. 2000), [1] (Haskard et al. 2001), [14] (Turbic et al.

CE

2002), [15] (Halttunen et al. 2008)

Figure 10. Aflatoxin binding by selected Lb. rhamnosus strains.

AC

[13] (Haskard et al. 2000), [6] (Peltonen et al. 2001), [7] (Zinedine et al. 2005), [15] (Halttunen et al. 2008)

41

ACCEPTED MANUSCRIPT

MA

NU

SC RI PT

Figure 1. Aflatoxin limits for feed, food and milk distributed between countries. Combined from FAO, 2003.

AC C

EP

TE

D

Figure 1.

42

ACCEPTED MANUSCRIPT

MA

NU

SC RI PT

Figure 2. Aflatoxin binding by selected Lb. acidophilus, Lb. amylovorus and Lb. brevis strains. [2] (El-Nezami et al. 1998a) [1] (Haskard et al. 2001), [6] (Peltonen et al. 2001), [3] (Oluwafemi and DaSilva 2009), [4] (Oluwafemi et al. 2010) [5] (Pizzolitto et al. 2012), [7] (Zinedine et al. 2005)

AC C

EP

TE

D

Figure 2. Clear column = unmodified, viable bacteria cells Red = killed bacteria cells Violet-red = killed bacteria cells in food matrix Violet = food matrix * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text)

43

ACCEPTED MANUSCRIPT

MA

NU

SC RI PT

Figure 3. Aflatoxin binding by selected Lb. casei strains. [2] (El-Nezami et al. 1998a), [1] (Haskard et al. 2001), [7] (Zinedine et al. 2005), [8] (HernandezMendoza et al. 2009a), [9] (Hernandez-Mendoza et al. 2009b), [3] (Oluwafemi and Da-Silva 2009), [4] (Oluwafemi et al. 2010)

AC C

EP

TE

D

Figure 3. Clear column = unmodified, viable bacteria cells Red = killed bacteria cells Violet = food matrix Violet-red = killed bacteria cells in food matrix * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text)

44

ACCEPTED MANUSCRIPT

TE

D

MA

NU

SC RI PT

Figure 4. Aflatoxin binding by selected Lb. crispatus, Lb. delbrueckii, Lb. fermentum, Lb. gasseri, Lb. helveticus, Lb. johnsonii, Lb. lactis and Lb. paracasei strains. [10] (Peltonen et al. 2000), [1] (Haskard et al. 2001), [6] (Peltonen et al. 2001), [3] (Oluwafemi and DaSilva 2009), [4] (Oluwafemi et al. 2010) [2] (El-Nezami et al. 1998a), [9] (Hernandez-Mendoza et al. 2009b), [7] (Zinedine et al. 2005)

AC C

EP

Figure 4. Clear column = unmodified, viable bacteria cells Red = killed bacteria cells Violet = food matrix Violet-red = killed bacteria cells in food matrix * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text) Orange = cell free supernatant

45

ACCEPTED MANUSCRIPT

D

MA

NU

SC RI PT

Figure 5. Aflatoxin binding by selected Lb. plantarum, Lb. reuteri and Lb. salivarius strains. [1] (Haskard et al. 2001), [6] (Peltonen et al. 2001), [7] (Zinedine et al. 2005), [3] (Oluwafemi and DaSilva 2009), [4] (Oluwafemi et al. 2010), [9] (Hernandez-Mendoza et al. 2009b), [9] (Peltonen et al. 2000)

AC C

EP

TE

Figure 5. Clear column = unmodified, viable bacteria cells Red = killed bacteria cells Violet = food matrix Violet-red = killed bacteria cells in food matrix * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text)

46

ACCEPTED MANUSCRIPT

TE

D

MA

NU

SC RI PT

Figure 6. Aflatoxin binding by selected Bifidobacterium, Lactococcus, Leuconostoc, Pediococcus, Propionibacterium and Streptococcus strains. [9] (Peltonen et al. 2000), [6] (Peltonen et al. 2001), [15] (Halttunen et al. 2008), [9] (HernandezMendoza et al. 2009b), [1] (Haskard et al. 2001), [17] (Shahin 2007), [7] (Zinedine et al. 2005), [2] (ElNezami et al. 1998a), [1] (Haskard et al. 2001)

AC C

EP

Figure 6. Clear column = unmodified, viable bacteria cells Red = killed bacteria cells Violet = food matrix Violet-red = killed bacteria cells in food matrix * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text)

47

ACCEPTED MANUSCRIPT

TE

D

MA

NU

SC RI PT

Figure 7. Aflatoxin M1 binding by selected LAB strains. [18] (Pierides et al. 2000), [19] (El Khoury et al. 2011), [20] (Sarimehmetoğlu and Küplülü 2004)

AC C

EP

Figure 7. Red = killed bacteria cells Blue = dairy matrix Blue-red = killed bacteria cells in dairy matrix * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text)

48

ACCEPTED MANUSCRIPT

TE

D

MA

NU

SC RI PT

Figure 8. Aflatoxin binding by Lb. rhamnosus GG ATCC 53103 in vitro and in vivo. [2] (El-Nezami et al. 1998a), [11] (El-Nezami et al. 1998b), [1] (Haskard et al. 2001) [14] (Turbic et al. 2002) [16] (El-Nezami et al. 2000) [12] (Lahtinen et al. 2004)

AC C

EP

Figure 8. Red = killed bacteria * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text) Orange = cell free

49

ACCEPTED MANUSCRIPT

ED

MA NU

SC

RI P

T

Figure 9. Aflatoxin binding by Lb. rhamnosus LC-705. [2] (El-Nezami et al. 1998a), [11] (El-Nezami et al. 1998b), [16] (El-Nezami et al. 2000), [1] (Haskard et al. 2001), [14] (Turbic et al. 2002), [15] (Halttunen et al. 2008)

AC

CE

PT

Figure 9. Red = killed bacteria cells * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text)

50

ACCEPTED MANUSCRIPT

TE

D

MA

NU

SC RI PT

Figure 10. Aflatoxin binding by selected Lb. rhamnosus strains. [13] (Haskard et al. 2000), [6] (Peltonen et al. 2001), [7] (Zinedine et al. 2005), [15] (Halttunen et al. 2008)

AC C

EP

Figure 10. Red = killed bacteria cells * = variation in conditions (e.g. pH, temperature, concentration, verify from original source or text) X = binding stability after distinct treatments (verify from original source or text)

51

ACCEPTED MANUSCRIPT Sara Ahlberg 9 April, 2015 A review: Potential of lactic acid bacteria in aflatoxin risk mitigation

AC

CE

PT

ED

MA NU

SC

RI P

T

Highlights • Lactic acid bacteria show potential to bind aflatoxins • Aflatoxin binding by lactic acid bacteria is strain specific • Aflatoxin absorption could be reduced due to the bacteria interaction with mucus

52

Potential of lactic acid bacteria in aflatoxin risk mitigation.

Aflatoxins (AF) are ubiquitous mycotoxins contaminating food and feed. Consumption of contaminated food and feed can cause a severe health risk to hum...
1MB Sizes 0 Downloads 17 Views