Characterization of lactic acid bacteria-based probiotics as potential heavy metal sorbents


Jatindra Nath Bhakta, Research Institute of Molecular Genetics, Faculty of Agriculture, Kochi University, B200, Monobe, Nankoku, Kochi 783-8502, Japan. E-mail:


Aim:  To isolate and characterize lactic acid bacteria (LAB) and determine whether they could potentially be used as heavy metal (cadmium and lead) absorbing probiotics.

Methods and Results:  The study used 53 environmental (mud and sludge) samples to isolate cadmium- and lead-resistant LAB, by following spared plate technique. A total of 255 cadmium- and lead-resistant LAB were isolated from these samples. The survival of 26 of the LAB was found after passing through sequential probiotic characterizations. These 26 probiotic LAB exhibited remarkable variations in their metal-resistant and metal-removal abilities. Of 26, seven (Cd54-2, Cd61-7, Cd69-12, Cd70-13, Pb82-8, Pb96-19 and Cd109-16) and four (Pb71-1, Pb73-2, Pb85-9 and Pb96-19) strains displayed relatively elevated cadmium- and lead-removal efficiencies from water, respectively, compare with that of the remaining strains. Strains Cd70-13 and Pb71-1 showed the highest cadmium (25%) and lead (59%) removal capacity from MRS (De Man, Rogosa and Sharpe) culture medium, respectively, amongst the selected strains and showed a good adhesive ability on fish mucus. A phylogenetic analysis of their 16S rDNA sequences revealed that the strains Cd70-13 and Pb71-1 belong to Lactobacillus reuteri.

Conclusion:  Excellent probiotic, metal sorption and adhesive characteristics of newly identified Lact. reuteri strains Cd70-13 and Pb71-1 were isolated, which indicated their high potential abilities to survive in the intestinal milieu and to uptake the tested metals from the environment.

Significance and Impact of the Study:  To our knowledge, this is the first study that has aimed to isolate, characterize and identify metal-resistant LAB strains that have potential to be a probiotic candidate for food and in vivo challenge studies in the intestinal milieu of fish for the uptake and control of heavy metal bioaccumulation.


Heavy metal contamination is a serious problem not only from the human health perspective but also from broader environmental viewpoint because of its nonbiodegradable, hazardous and toxic properties. The hazardous heavy metals, such as cadmium (Cd) and lead (Pb), are reported amongst the top ten toxic metals in the Priority List of Hazardous Substances (ATSDR 2007) for causing a grossly biological impact by bioconcentration, bioaccumulation and biomagnification phenomena. Cadmium is biologically nonessential but poisonous for plants, animals and humans (Gupta and Gupta 1998). It is recognized as a human carcinogen (IARC 1994). Human exposure to low level of Cd can result in renal diseases, osteomalacia and lung cancer as well as damage the cardiovascular system, liver and reproductive system (USEPA 1992; Hrudey et al. 1995; Belimov et al. 2005). Lead has also been regarded as a long-lasting environmental pollutant, which is responsible for causing various dysfunctions, central and peripheral nervous system damages, memory deterioration and diminishing intellectual capacity for children (Verity 1995; Jarup 2003).

Water and food are the key routes for heavy metal contamination in the biological systems. Aquatic animals, especially fish, are an important source of animal protein, thus is one of the major sources of heavy metal contamination in human body (Friberg et al. 1974; Cheng and Gobas 2007). Various freshwater fishes (Amundsen et al. 1997) and marine fishes (Wong et al. 2001) have been reported to accumulate heavy metals in the body through the consumption of water and food. Fish culturing in contaminated effluents may lead to heavy metal bioaccumulation in tissue system (Hollis et al. 2001; Long and Wang 2005). Therefore, removal of heavy metal from water and food should be the priority measure to control the problem of bioaccumulation.

Heavy metal contamination also significantly affects the microbial community in the environment. Soil contaminated with heavy metals affects the structure (qualitative and quantitative) of microbial communities, resulting in decreased metabolic activity and diversity (Giller et al. 1998). Extensive research has shown that soil bacteria are tolerant to heavy metals and play important roles in mobilization of heavy metals (Gadd 1990; Idris et al. 2004). Therefore, exploration of potentially heavy and toxic metal-resistant bacteria and potential application of them for the treatment of metal contaminated water is one of the low-cost and promising bioremedial technologies. Such bacterial applications in removing heavy metals have been well studied previously (Brown and Lester 1982; Min-sheng et al. 2001; Wei et al. 2009).

One of the emerging fields in this area is the isolation of potentially favourable bacterial species, especially lactic acid bacteria (LAB, lactobacilli and bifidobacteria) from the environmental samples and intestinal contents, probiotic characterization and application in fish as probiotic for nutritional, growth, disease control (Macey and Coyne 2005; Mohideen et al. 2010), and immunological (Panigrahi et al. 2005) purposes. It has also been suggested that LAB strains are able to remove cyanotoxins from water and to uptake mycotoxins from food (Meriluoto et al. 2005; Nybom et al. 2007). A combination of two probiotic strains has also been reported to reduce the gastro-intestinal absorption of aflatoxin B1 in young Chinese men (Pierides et al. 2000; El-Nezami et al. 2006). In addition, application of LAB in removing toxic metals from water has been reported by Halttunen et al. (2007).

However, it is apparent that no attempt has been made so far to isolate Cd- and Pb-removing probiotic strains from metal-resistant LAB community of the environment through assessing the probiotic, metal-resistant and metal-removal characteristics. In this study, we aim to isolate metal-resistant LAB from mud and sludge samples and to determine whether they have the probiotic properties, which could potentially be used as heavy metals (Cd and Pb) sorbent for removing heavy metals from ambience. For this purpose, (i) Cd- and Pb-resistant LAB were isolated and were subsequently employed for sequential probiotic characterization and identification; (ii) the identified probiotic LAB strains were then selected and tested as potential probionts for uptaking metals from water and culture mediums; (iii) the study of in vitro adhesion using selected LAB was conducted to ascertain the colonizing ability in fish intestine and (iv) phylogenetic analysis of the selected probiotic strains based on 16S rDNA was used to determine the genetic relationship with other LAB.

Materials and methods

Collection and processing of sample

Pollutant-loaded effluent transporting channels, wastewater treatment plants and coastal zones are supposed to be heavy metal-contaminated habitats in the environment. Consequently, opportunistic development of metal-resistant property in the community of micro-organisms in the above-mentioned habitats should be common phenomena because of frequent exposure to metal pollutants. On this basis, this study employed a total of 53 mud and sludge samples collected from some costal aqua-farming zones and effluents carrying canals in India (10 samples) and Viet Nam (40 samples), whereas only sludge samples were procured from a wastewater treatment plant in Japan (three samples).

Equal proportions of all samples were blended properly to get a homogenous sample for each station. Thus, three samples were prepared and named as S1, S2 and S3 for the samples of India, Japan and Viet Nam, respectively. These three-parental stock samples were preserved at −20°C for the isolation of metal-resistant LAB.

Metal solution

Cd (2000 mg l−1) and Pb (2000 mg l−1) stock solutions were prepared from CdCl2 and PbNO3 (Cica-Reagent; Kanto Chemical Co., Inc., Tokyo, Japan), respectively. Metal solutions were added to the bacterial culture medium after autoclaving and cooling at c. 40°C.

Isolation and morphological characterization of heavy metal (Cd and Pb)-resistant LAB

In our previous study, only Enterococci sp. was isolated from these samples without using the process of bacterial enrichment culture [i.e. incubation of samples in MRS (De Man, Rogosa and Sharpe, specific media for LAB; Difco Laboratories, Detroit, MI, USA) broth for 3–7 days] (unpublished). Therefore, this study followed the process of bacterial enrichment culture to isolate other potential LAB from the samples using the following methods. The preserved stock samples were brought to normal temperature before use. A 1-g portion of each sample was suspended in 9 ml 0·85% physiological saline (PS) and vortexed for 5 min to get a homogenous bacterial suspension; 1 ml aliquot of each sample suspension was inoculated in 9 ml MRS broth media and incubated for 7 days at 37°C anaerobically using the Anaero-pack rectangular jar with an Anaeropack-Anaero sachet (Mitsubishi Gas Chemical Company, Tokyo, Japan) statically to enrich the population of LAB. Aliquots of cultured broth were serially diluted (10−1 to 10−8) using PS; 100 μl of broth was then inoculated over the 0·017% bromocresol purple impregnated-MRS (BP-MRS) agar media plates supplemented with heavy metals (Cd or Pb) at 50 mg l−1 and incubated at 37°C for 24 h anaerobically.

Distinct yellow colonies were randomly picked up by tooth pick from the higher dilutions of metal supplemented agar plates to represent the metal-resistant bacteria. Isolates were re-streaked two times on MRS agar plates containing respective metal (50 mg l−1) for further purification. The pure culture was then employed for catalase reaction using one drop of 3% hydrogen peroxide solution on each isolate of the re-streaked plates, and immediate formation of bubbles indicated the presence of catalase in the cells. Catalase-negative isolates (CNI) were selected for further morphological study under the phase contrast microscope (Olympus, Tokyo, Japan). CNI was maintained in MRS broth containing 20% glycerol at −85 °C for subsequent studies.

Probiotic characterization of resistant LAB

Potential probiotic LAB was screened using the following method of sequential characterizations.

Acid pH tolerance

The acid-tolerant isolate (ATI) was selected using the modified method of Erkkila and Petaja (2000). Briefly, fresh LAB of 24-h MRS culture were harvested, washed twice with PS and centrifuged at 13000 g for 5 min using centrifuge (Eppendorf AG, Hanburg, Germany) to obtain cell pallet. Washed cell pallets were then suspended (c. 107 CFU ml−1) in sterile phosphate-buffered saline (PBS; NaCl, 9 g l−1, Na2HPO4·2H2O, 9 g l−1and KH2PO4, 1·5 g l−1) adjusted to pH 2·5 ± 0·3 using 5 mol l−1 HCl and incubated at 37 °C for 2 h. Following the acid treatment, the bacterial suspension was used for platting in MRS agar media and tolerant LAB was assessed in terms of colony growth after 48-h incubation at 37°C.

Bile salt tolerance

To select the bile-tolerant isolates (BTI), the successful ATIs were used for bile salt test following the modified method described by Arihara et al. (1998). LAB strains were grown at 37°C for 24 h in MRS broth without bile salt and 1 ml of broth was then used for platting in MRS agar with added bile salt (Sigma-Aldrich, Germany) at concentrations of 1000, 2000 and 4000 mg l−1, respectively. After 48-h incubation at 37 °C, the growth of LAB was evaluated to select the BTI.

Antimicrobial activity assay

The LAB with bile-tolerant ability were cultured in 4 ml MRS broth for 24 h at 37°C anaerobically. Total culture was centrifuged at 13000 g for 5 min, supernatant was collected and filter sterilized by passage through a 0·2-μm Millipore membrane (Millipore, Billerica, MA, USA). Antimicrobial activity was determined by agar disc-diffusion assay using the non-neutralized and neutralized (pH 6·8 adjusted with 5 mol l−1 NaOH) filter-sterilized supernatant (Balcázar et al. 2008). The subcultured (c. 107 CFU ml−1) indicator strains (Escherichia coli and Salmonella sp.) were then flooded over the Mueller–Hinton agar (Difco Laboratories, Detroit, MI, USA) plates at 100 μl plate−1 and air-dried for 30 min. The paper discs (8 mm; Advantec, Roshi Kaisha Ltd, Tokyo, Japan) were overlaid onto the bacterium-seeded agar plates and impregnated with sterilized supernatant of cultured LAB strains at 50 μl disc−1. The antimicrobial activity was determined by measuring the diameter (mm) of clear zone around the disc after incubation at 37°C for 24 h.

Measurement of lactic acid yield and pH

An aliquot of the supernatant of 24-h LAB cultured MRS broth was collected by centrifugation, filtered and used for the determination of lactic acid concentration using high-performance liquid chromatography (HPLC). Another aliquot was employed for measuring the pH changes of the broth occurred because of LAB growth. The HPLC apparatus consists of Jasco liquid chromatography solvent delivery system and UV/Vis detector (PU-2080 and UV-2075; Tokyo, Japan,) with Cosmosil packed column 5C18-PAQ (4·6 × 250 mm). Used mobile phase was 20 mmol l−1 H3PO4. Filtered 10 μl aliquot of MRS broth was injected into the column, and HPLC was run with a flow rate, 1·0 ml min−1; column temperature, 30°C; and detection UV, 210 nm. Lactic acid content in the broth was expressed in mmol l−1 comparing with known standard lactic acid (Nacalai Tesque Inc., Kyoto, Japan) concentrations.

Antibiotic resistant assay

The MRS plates were prepared by spreading the fresh overnight cultured LAB (100 μl per plate broth of c. 10CFU ml−1 LAB) for antibiotic susceptibility assay using disc-diffusion method described elsewhere. The paper discs were impregnated by four different concentrations of antibiotics (Sigma-Aldrich Chemical GmbH) solution (sulphamethoxazole and streptomycin at 50, 100, 300 and 500 mg l−1; chloramphenicol and oxytetracycline at 10, 50, 100 and 300 mg l−1) at 50 μl per disc. After incubation for 24 h at 37°C, antibiotic sensitivity was assessed by measuring the diameter (mm) of clear zone around the disc.

Identification of metal-resistant LAB

DNA of LAB was extracted using the standard method (Ruiz-Barba et al. 2005). In brief, LAB suspension was made with 100 μl of sterilized MQ (Milli-Q) water in 1·5-ml microcentrifuge tubes and then 100 μl of chloroform/isoamyl alcohol (24 : 1) was added to the suspension. After vortex for 5 s, the mixture was centrifuged at 16 000 g for 5 min at 4 °C. Supernatant aqueous solution of the microcentrifuge tube was used as a source of DNA template for polymerase chain reaction (PCR).

Isolated 16S rDNA fragments were amplified by PCR using the universal primers FProR (5′-AGAGTTTGATCCTGGCTCAG-3′) and R534 (3′-GGTCGTCGGCGCCATTA-5′) (invitrogen) with the thermocycler PC818 (ASTEC programme temperature control system). The mixture (20 μl) of PCR was prepared by adding 10 μl of AmpliTaq Gold® 360 Master Mix with 0·5 μl of 360 GC Enhancer (Applied Biosystems, Foster City, CA), 1 μl of each primer, 2·5 μl of nuclease free water and 5 μl of template DNA. The PCR programme was as follows: 95°C for 10 min; 30 cycles of 95°C for 30 s, 60°C for 30 s and 72°C for 1 min; and a final extension step at 72°C for 7 min. The PCR products were detected on agarose gel (1·2%) stained with ethidium bromide.

The PCR amplicons were purified using DNA gel extraction kit (Wizard® SV Gel and PCR Clean-Up System; Promega, Madison, WI) following the manufacturer’s instructions. Sequencing PCR of purified DNA was performed by BigDye and R534 primer following the programme: 98°C for 1 min; 40 cycles of 98°C for 10 s, 50°C for 5 s; and 60°C for 2·5 min. DNA sequencing was performed with an automated DNA sequencer (3100-Avant Genetic Analyzer; Applied Biosystems). Identification was carried out by searching the Genbank DNA database using blast (Basic logical alignment search tool) at NCBI and DDBJ.

Metal-resistant and removal characterizations of LAB

The best metal uptaking probiotic LAB strain was identified by the following metal-resistant and removal characterizations.

Metal-resistant patterns.

Metal-resistant profiles of the identified LAB against Cd and Pb were determined by measuring minimum inhibitory concentrations (MICs) following the previously described in vitro disc-diffusion assay. MRS agar media was used to prepare the LAB-seeded plates, and different concentrations of Cd (50–1000 mg l−1) and Pb (50–2000 mg l−1) solutions were used at 50 μl per disc for the impregnation of discs overlaid on plates. MIC was assessed by determining the lowest concentration of bacterial growth inhibition.

Metal removal profiles

To select the best strains with excellent metal (Cd or Pb)-removal efficiency from water, all identified LAB were employed for metal removing study following the method described by Pazirandeh et al. (1998) with some modifications. Freshly cultured LAB were harvested in 2-ml Eppendorf tubes, centrifuged at high speed to pallet the cells and washed twice using sterilized MQ water. Pellet of the cell [3 mg ml−1 (wet weight)] was resuspended in sterilized Cd (1000 μg l−1) or Pb (6000 μg l−1) solutions prepared from previously mentioned stocks and incubated at 37 °C. After 2 h, samples were collected in 2 ml Eppendorf tubes and centrifuged at 8000 g for 10 min, and metal contents in supernatant were measured using the AAS (AA-6800; Shimadzu, Kyoto, Japan). Metal-removal efficiency (MRE) was calculated by using the following equation:


where C is designated for metal concentration in water, and initial and final concentrations were represented as Ci and Cf, respectively. t was depicted as time, then final and initial time were represented as tf and ti, respectively, and M is mass of the LAB cell. The MRE data were expressed as μg h−1 mg−1 in these experimental conditions.

Further, the selected metal (Cd or Pb)-specific LAB mentioned above with higher MRE classes were used for removing respective metal from MRS culture medium to distinguish the strain with excellent metal removal capacity from culture medium. Likewise, freshly cultured LAB cell was harvested, washed, inoculated [10 mg ml−1 (wet weight)] into MRS medium containing 10 mg l−1 Cd or Pb and incubated at 37 °C anaerobically with agitation at 200 excursions per min. A control for each metal was maintained without receiving LAB. Samples were collected at 0, 6, 12 and 24 h and centrifuged, and supernatants were used to analyse the metal concentration by AAS. Metal removal rate was expressed in percentage.

Characterization of heavy metal (Cd and Pb) binding to LAB cells

The strains showed the greatest Cd or Pb uptaking ability in MRS medium was further considered to elucidate the metal binding/bioaccumulation properties. The binding experiment was performed as described in the above-mentioned study of metal removal from MRS medium. Samples were collected at 12, 24 and 48 h, centrifuged to separate the cell pellets and washed thrice with sterilized MQ to remove free heavy metal ions. The cell pellets were resuspended in 10 mmol l−1 sterilized EDTA and agitated for 30 min at room temperature for desorbing the metals bound on outer surface of the cell membrane (OSM). The supernatant and cell pellets were separated by centrifugation and analysed by AAS to determine the heavy metal content in the OSM and intracellular space (ICS).

In vitro adhesion assay

To use as probiotic in intestinal milieu, the adhesion ability of LAB onto mucosal epithelium cell of intestine is also an important criterion to assess the colonization properties. As our selected LAB isolated from sludge and sediment samples will be applied in fish to control the heavy metal bioaccumulation, fish intestinal mucus was considered for assessing the adhesion ability of LAB. Preparation of mucus, radioactive thymidine (methyl-3H; 10 μl ml−1, 84 Ci mmol l−1; American Radio labeled Chemicals Inc., St Louis, MO, USA) labelling of LAB and adhesion study were performed following the standard methods (Balcázar et al. 2008) with slight modifications. Briefly, 100 μl of intestinal mucus were immobilized on polystyrene microtiter plate wells. After overnight incubation at 4°C, the wells were washed twice with 200 μl of PBS to remove excess mucus. Radioactively labelled bacteria (107 CFU ml−1) were added at 100 μl per well and incubated for 1 h at room temperature. The wells were washed twice with 250 μl PBS to remove unbound bacteria, and bound bacteria were treated by 1% sodium dodecyl sulphate in 0·1 mol l−1 NaOH and incubated at 60°C for 1 h. A control well received no mucus was maintained for each LAB. Adhesion ability of LAB was assessed as percentage of radioactivity recovered compare with the radioactivity of bacterial suspension added to the immobilized mucus. The radioactivity was quantified by using the liquid scintillation counter.

Accession number and phylogenetic analysis of best metal uptaking LAB strains

The 16S rDNA sequences of efficient Cd- and Pb-uptaking LAB strains were deposited in DDBJ/EMBL/GenBank for serial accession numbers. A phylogenetic tree of selected best LAB strains was constructed according to the neighbour-joining method using the partial sequence (490 bp) of 16S rDNA.

Statistical analysis

Mean data of at least two independent experiments with three replicates of different characterization studies were used for the evaluation of results. Correlation studies were performed using spss 10 (IBM, New York, NY), and the accepted significant was at 0·01 levels.


Isolation and morphological characterization of heavy metal (Cd and Pb)- resistant LAB

A total of 255 heavy metal-resistant yellow coloured colonies were picked up from Cd (109 colonies) and Pb (146 colonies) supplemented plates of three sampling stations. Amongst 255 colonies, 204 (Cd-resistant 94 and Pb-resistant 110) were selected as CNI (Table 1).

Table 1.   Catalase negative, acid pH, and bile salt-tolerant lactic acid bacteria isolates in three sampling stations
Sampling stationsCd-resistant isolatesPb-resistant isolates
1000 mg l−12000 mg l−14000 mg l−11000 mg l−12000 mg l−14000 mg l−1
  1. ATI, acid-tolerant isolate; BTI, bile-tolerant isolates; CNI, catalase-negative isolates.


Mixed morphologies of rod, cocci and oval-shaped isolates were observed under the phase contrast microscope. Twenty-five Cd-resistant and 30 Pb-resistant LAB isolates were rod shaped, whereas the remaining all were cocci and oval shaped.

Probiotic characterization of resistant LAB

Acid pH and bile salt tolerance

All catalase negative LAB isolates were employed for testing their abilities to grow at pH 2·5 ± 0·3 to select the potential acid-tolerant probiotic strains. Forty-six and 65 isolates were determined as ATI from 94 Cd-resistant and 110 Pb-resistant isolates, respectively, in this study (Table 1). S1 showed higher number of ATI than the rest sampling stations (S2 and S3) in both Cd- and Pb-resistant CNI.

After acid tolerance test, 80 (33 Cd-resistant and 47 Pb-resistant), 46 (16 Cd-resistant and 30 Pb-resistant) and 35 (12 Cd-resistant and 23 Pb-resistant) LAB were successfully passed in the bile salt tolerance test at the concentration levels of 1000, 2000 and 4000 mg l−1, respectively (Table 1). Like ATI, the abundance of BTI was greater in S1 than the remaining stations (S2 and S3). The number of BTI decreased with the increase in the bile salt concentration, and 35 highest BTI were selected from the plate containing 4000 mg l−1 bile salt to carry out next step of probiotic characterization.

Antimicrobial activity, lactic acid yield and pH measurement.

Non-neutralized and neutralized supernatants of all 35 BTI cultured MRS broth were employed to study the antimicrobial activity. Amongst them, the cultured supernatant of 23 (10 Cd-resistant and 13 Pb-resistant) LAB under non-neutralized condition exhibited the antimicrobial activity against both E. coli and Salmonella sp., whereas 2 Cd-resistant (Cd107-14 and Cd108-15) and 1 Pb-resistant (Pb74-3) LAB strains showed the growth inhibition effects only on E. coli and Salmonella sp., respectively (Fig. 1a,b). In both E. coli and Salmonella sp., the diameter of clear zone ranged from 0 to 13·7 mm and 0 to 11·5 mm in Cd- and Pb-resistant LAB, respectively. Nine Pb-resistant LAB (Pb79-6, Pb81-7, Pb90-14, Pb92-16, Pb94-17, Pb95-18, Pb98-21, Pb110-22 and Pb111-23) registered no growth inhibition zones against both E. coli and Salmonella sp. amongst 35 Cd- and Pb-resistant LAB (Fig. 1a,b). No antimicrobial effect was observed by neutralized supernatant of all examined LAB.

Figure 1.

 The antimicrobial activity (upper panel; inline imageEscherichia coli and inline imageSalmonella sp.) and the relationship between lactic acid concentration (inline image) and pH (inline image) (lower panel) of 24-h Cd-resistant (a) and Pb-resistant (b) lactic acid bacteria strains cultured MRS broth media.

Thirty-five bile salt-tolerant LAB strains were used to determine the amount of lactic acid yield and state of pH developed in 24-h cell-free cultured MRS broth. The lactic acid concentrations of MRS broth varied from 221·6 to 364·7 mmol l−1 and 57 to 414·7 mmol l−1 in the Cd- and Pb-resistant LAB, respectively (Fig. 1a,b). Likewise, the pH of MRS broth varied from 3·98 to 4·72 and 3·93 to 6·38 in Cd- and Pb-resistant LAB, respectively (Fig. 1a,b).

Antibiotic resistant profiles.

Antibiotic-dependant responses were observed in the resistant profiles of 26 (12 Cd- and 14 Pb-resistant) antibiotic activity-pronouncing LAB strains tested. The diameters of the resultant growth inhibition zones varied from 0 to 22 mm in four antibiotics (trimethoprim, chloramphenicol, streptomycin and oxytetracycline) against all LAB strains (Table 2). An elevated percentage (Cd-resistant, 66·6–100% and Pb-resistant, 50–100%) of all LAB strains showed the clear zones at higher concentrations (≥500 mg l−1) in trimethoprim, streptomycin and oxytetracycline than those of the remaining LAB. All and 30% of tested LAB isolates showed clear growth inhibition at lower concentration against chloramphenicol (50 mg l−1) and oxytetracycline (10 mg l−1), respectively, than against other tested antibiotics (Table 2).

Table 2.   Antibiotic resistant activities (±SE) and activity pronouncing minimum concentrations in isolated lactic acid bacteria strains
Types of resistantStrainsTrimethoprimChloramphenicolStreptomycinOxytetracycline
Activity (mm)Concentration (mg l−1)Activity (mm)Concentration (mg l−1)Activity (mm)Concentration (mg l−1)Activity (mm)Concentration (mg l−1)
Cd-resistantCd54-218 ± 0·630018 ± 1·25010 ± 0·65013 ± 0·5300
Cd56-318 ± 1·230015 ± 0·35010 ± 0·350019 ± 1·710
Cd57-414 ± 0·330020 ± 0·65015 ± 0·95015 ± 0·6300
Cd59-50>50018 ± 1·25011 ± 0·550012 ± 0·910
Cd60-620 ± 1·230013 ± 0·8509 ± 0·350017 ± 0·610
Cd61-710 ± 0·950020 ± 0·6500>5000>500
Cd63-911 ± 1·250022 ± 1·2500>50012 ± 1·2500
Cd69-120>50018 ± 0·75011 ± 0·510013·5 ± 0·3300
Cd70-1324 ± 0·950020 ± 0·95010 ± 0·710012 ± 1·1300
Cd107-140>50015 ± 0·6500>5000>500
Cd108-1514 ± 1·750017 ± 0·3500>5000>500
Cd109-1612·5 ± 0·650013 ± 0·7500>5000>500
Pb-resistantPb71-115 ± 0·350021 ± 0·65011·5 ± 1·25015 ± 1·2300
Pb73-210 ± 0·630015 ± 1·25011 ± 0·630015 ± 0·9300
Pb74-320 ± 1·430013 ± 0·45011·5 ± 0·950017 ± 1·710
Pb76-420 ± 0·930018 ± 0·5500>50022 ± 1·210
Pb78-516 ± 0·630017 ± 0·3500>50020 ± 1·810
Pb82-812 ± 1·250017 ± 0·6500>5000>500
Pb85-922 ± 0·650018 ± 0·4500>50012 ± 0·710
Pb86-1020 ± 1·550016 ± 1·3500>50020 ± 1·410
Pb87-110>50018·5 ± 0·35010 ± 0·330016 ± 1·3300
Pb88-120>50012 ± 0·65012 ± 1·250016 ± 0·610
Pb89-130>50013 ± 0·9500>50012 ± 1·710
Pb91-1522 ± 0·650013·5 ± 0·5500>50010 ± 0·3300
Pb96-1913 ± 0·850016 ± 1·25010 ± 0·85010 ± 0·7300
Pb97-2012 ± 0·950016 ± 0·3500>50015 ± 0·6300

Identification of metal-resistant LAB

PCR-amplified 16S rDNA of 26 Cd- and Pb-resistant LAB isolates resulted in the synthesis of characteristic single band of about 500 bp. The partial sequences of purified 16S rDNA amplicon of isolated LAB were subjected to bacterial identification. 16S rDNA sequencing data of 9 (Cd54-2, Cd57-4, Cd60-6, Pb76-4, Pb78-5, Pb86-10, Pb87-11, Pb89-13 and Pb91-15), 2 (Cd56-3 and Cd59-5), 6 (Cd69-12, Cd70-13, Cd107-14, Cd108-15, Pb71-1 and Pb73-2), 1 (Pb96-19), 4 (Cd61,-7, Cd63-9, Pb82-8 and Pb88-12), 1 (Pb97-20) and 3 (Cd109-16, Pb74-3 and Pb85-9) LAB strains clearly showed 97 to 99% homology to Lactobacillus amylovorus, Lactobacillus johnsonii, Lactobacillus reuteri, Lactobacillus dextrinicus, Pediococcus acidilactici, Enterococcus hirae and Enterococcus faecium, respectively.

Metal-resistant and removal characterizations of LAB

Metal-resistant pattern.

To ascertain the metal-resistant pattern, identified LAB were employed to determine the MIC. All 26 LAB strains registered significantly wide variations with high magnitude of Cd- and Pb-resistant patterns (Table 3). The MIC values for Cd varied from 60 to >1000 mg l−1 and 50 to >500 mg l−1 in Cd- and Pb-resistant strains, respectively, whereas all LAB strains showed a remarkably elevated MIC values (>2000 mg l−1) against Pb (Table 3).

Table 3.   Minimum inhibitory concentration (MIC) patterns of isolated Cd- and Pb-resistant lactic acid bacteria strains against Cd and Pb
Types of resistantStrainsMIC (mg l−1)

Metal removal profile.

All 26 identified LAB strains were employed in Cd- and Pb-removal study to select the LAB strains with high potential of Cd- and Pb-removal efficiencies, respectively, because most of the LAB strains have high Cd- and Pb-resistant abilities. The removal of Cd (99·78 to 870·4 μg l−1) and Pb (139 to 3425 μg l−1) varied remarkably depending on the LAB strains tested (Fig. 2a,b). Metal-removal efficiencies of 26 LAB ranged from 0·016 to 0·145 μg h−1 mg−1 (wet weight of cell) and 0·023 to 0·602 μg h−1 mg−1 (wet weight of cell) in Cd and Pb, respectively.

Figure 2.

 Cd (a) and Pb (b) removal (inline image) and removal efficiency (♦) characteristics of isolated Cd- and Pb-resistant lactic acid bacteria (LAB) strains (Inset reveals the percentage occurrence of LAB strains in different metal-removal efficiency classes).

Frequency distribution of examined strains exhibited that LAB of higher (>0·13 to ≤0·145 μg h−1 mg−1) and lower (>0·016 to ≤0·1 μg h−1 mg−1) Cd-removal efficiency classes were 26 and 30%, respectively, whereas moderate (>0·1 to ≤0·13 μg h−1 mg−1) Cd-removal efficiency classes occurred at 44%. Likewise, frequency distribution showed that 15·4% of all identified strains appeared as higher Pb-removal efficiency classes (>0·5 to ≤0·602 μg h−1 mg−1), whereas moderate (>0·2 to ≤0·5 μg h−1 mg−1) and lower (>0·023 to ≤0·2 μg h−1 mg−1) Pb-removal efficiency classes occurred at 50 and 34·6%, respectively.

Amongst 26 LAB tested, 7 (Lact. amylovorus Cd54-2, Ped. acidilactici Cd61-7, Lact. reuteri Cd69-12 and Cd70-13, Ped. acidilactici Pb82-8, Lact. dextrinicus Pb96-19 and Ent. faecium Cd109-16) and 4 (Lact. reuteri Pb71-1 and Pb73-2, Ent. faecium Pb85-9 and Lact. dextrinicus Pb96-19) belonged to higher Cd- (>0·13 to ≤0·145 μg h−1 mg−1) and Pb- (>0·5 to ≤0·602 μg h−1 mg−1) removal efficiency classes, respectively, because of their comparatively higher metal removal properties. Therefore, these LAB were tested for further respective metal removal from MRS broth. In MRS medium, the total metal removal (Cd 0·34–2·5 mg l−1 and Pb 0·3–5·9 mg l−1) pattern of selected LAB strains was similar to that of the aqueous phase (Fig. 3a,b). Percentage of metal removal from MRS varied between 3·42–25% for Cd and 3–59% for Pb. Maximum percentage removals were found in the Cd70-13 strain (25%) for Cd and Pb71-1 (59%) for Pb.

Figure 3.

 Cd (a) and Pb (b) removal of the selected higher seven Cd-(Control inline image, Cd54-2 inline image, Cd61-7 inline image, Cd69-12 inline image, Cd70-13 inline image, Cd109-16 inline image, Pb82-8 inline image and Pb96-19 inline image ) and four Pb- (Control inline image, Pb71-1 inline image, Pb73-2 inline image, Pb85-9 inline image and Pb96-19 inline image ) removal efficiency exhibiting lactic acid bacteria strains, respectively, from MRS broth.

Characterization of heavy metal (Cd and Pb) binding in LAB cell

The amount of respective heavy metals bound to the cells of respective strains was shown in Fig. 4a,b. The maximum metal content was observed at 48 and 24 h in Cd70-13 and Pb71-1 strains, respectively. In Cd70-13 strain, OSM showed increasing trend with time and ICS exhibited a lower value at 48 h in Cd concentration. Although there was no remarkable temporal change in Pb content on OSM of Pb71-1 strain, the ICS showed higher amount of Pb at 12 and 24 h than that of 48 h.

Figure 4.

 Cd (a) and Pb (b) binding characteristics of Cd70-13 and Pb71-1 strains, respectively, on outer surface of the cell membrane (□ OSM) and in intracellular space (inline image ICS).

In vitro adhesion assay.

In intestinal mucus, the percentages of adhesion were 16·5 and 18·8% in Cd70-13 and Pb71-1 strains, respectively. The adhesion abilities were greater (Cd70-13, 46% and Pb71-1, 27%) in intestinal mucus than that of the polystyrene in both LAB strains.

Accession number and phylogenetic analysis of efficient metal uptaking LAB strains.

The received serial accession numbers of Cd54-2, Cd61-7, Cd69-12, Cd70-13, Cd109-16, Pb71-1, Pb73-2, Pb82-8, Pb85-9 and Pb96-19 LAB strains with higher MRE classes were from AB627836 to AB627845, respectively. Phylogenetic analysis of the selected strains based on the neighbour-joining method resulted into five major clusters (Fig. 5), which are Cluster – I formed with Lact. amylovorus, Cluster – II formed with Lact. reuteri, Cluster – III formed with Ped. acidilactici, Cluster – IV formed with Lact. dextrinicus and cluster – V formed with Ent. faecium, respectively. The selected potential Cd- and Pb-removal strains, Cd70-13 and Pb71-1, belonged to the cluster – II, Lact. reuteri in the constructed phylogenetic tree.

Figure 5.

 Neighbour-joining phylogenetic tree based on 16S rDNA gene sequences showing affiliation of selected lactic acid bacteria strains with closely related members in GenBank and ribosomal database project (RDP). Values on lines indicate the genetic distance.


As the impregnated BP in MRS agar media turns the colony colour into yellow with the action of LAB-synthesized lactic acid, the preliminary screening of Cd- and Pb-resistant LAB was executed by isolating the yellow colonies from the respective metal (Cd or Pb) supplemented BP-MRS agar plates. Consequently, 255 heavy metal-resistant isolates were identified as Cd-resistant (109 isolates) and Pb-resistant (146 isolates) LAB from present sampling stations. Catalase is a common enzyme found in nearly all living organisms exposed to oxygen, where it functions to catalyse the decomposition of hydrogen peroxide to water and oxygen (Chelikani et al. 2004). It has one of the highest turnover numbers of all enzymes; one molecule of catalase can convert millions of molecules of hydrogen peroxide to water and oxygen per second (Goodsell 2004). Generally, anaerobic microbes do not possess the catalase enzyme in the cell; therefore, no bubbles were produced when the cells were exposed to hydrogen peroxide. Taking advantage of this phenomenon, this study employed the catalase test as the second stage of preliminary screening process. As a result, 204 (Cd-resistant 94 and Pb-resistant 110) isolates were selected as catalase negative LAB. These data suggested that a considerable number of LAB acquired resistant properties probably due to frequent heavy metal exposure in the investigated regions. Therefore, mud and sludge of effluent transporting channels, wastewater treatment plants and coastal habitats could be the potential sources of metal-resistant LAB for the selection and identification of beneficial probiotics.

As probiotic bacteria face the challenging acidic and bile salty environment in stomach and intestine during their establishment process, acid- and bile-tolerant abilities are the two primary important criteria of LAB for establishing within the gastrointestinal system of any animals as probiotic. High acidity in stomach and high concentration of bile components in proximal intestine of the host influence probiotic strain selection (Hyronimus et al. 2000). Therefore, this study selected 46 Cd-resistant and 65 Pb-resistant LAB as potential ATI because of their ability to survive at pH 2·5 ± 0·3 for 2 h periods. The pH levels of gastric juice may vary from 2·0 to 3·5 depending on the feeding time, the growing stage or the kind of animal (Yu and Tsen 1993). Probiotic bacteria have been selected using PBS buffer with pH 2·5 for 3 h at 37 °C (Pennacchia et al. 2004). Bile-tolerant test also showed that 35 (12 Cd-resistant and 23 Pb-resistant) LAB had the highest bile salt (4000 mg l−1) tolerance. The Lactobacillus strains can grow in MRS agar supplemented with 3000 mg l−1 bile salt (Pennacchia et al. 2004). The strains, Ped. acidilactici (P2), Lactobacillus curvatus (RM 10) and Lactobacillus sake (L2), were resistant to 3000 mg l−1 bile salt at pH 6 (Erkkila and Petaja 2000). Considering these high tolerant abilities, it may be concluded that the selected 35 strains with the highest BTI could survive efficiently within the intestinal milieu by preventing the adverse effect of bile salt. It is important to mention here that although this study employed bile salt (Sigma-Aldrich) derived from porcine in selecting the bile-tolerant LAB, the relevant fish bile would be more appropriate in probiotic characterization of LAB for applying in fish. Besides, the exact bile tolerance level of isolates may be varied in the case of fish bile.

This study demonstrated that the non-neutralized supernatant of 26 (12 Cd- and 14 Pb-resistant) LAB strains showed the antimicrobial activity with a remarkable variations in lactic acid yield and pH level, whereas no antimicrobial effect was observed by neutralized supernatant of all examined LAB. With few exceptions, it is apparent that nonantimicrobial activity exhibiting LAB showed lower concentrations of lactic acid (57–289 mmol l−1) and higher pH (4·22–6·4) levels in their MRS broth than the antimicrobial activity exerting LAB (Fig. 1a,b). Correlation between the lactic acid concentrations and pH levels of the broth clearly demonstrated a significant negative relationship (r = −0·861) which indicated that the concentration of lactic acid is one of the major factors in affecting the pH level in LAB cultured MRS broth. In addition, a negative correlations (E. colir = −0·533, Salmonella sp. r = −0·664) were also seen between the pH levels and antimicrobial activities of the respective LAB strains, which suggested antimicrobial activity of isolated LAB strain was supposed to be associated with the function of lower pH level of supernatant developed by lactic acid but not with bacteriocin in this investigation. Previous studies showed that pH lower than 4·4 could inhibit the growth of E. coli (Alvarado et al. 2006) and Salmonella sp. (Sorrells and Speck 1970). Hwanhlem et al. (2010) isolated four strains of Lactobacillus plantarum and reported no strains produced bacteriocin. The above-mentioned propositions illustrated that antimicrobial activity was pH dependent and bacteriocin was not produced by the isolated LAB strains because of no antimicrobial activity in neutralized supernatant in this study. The antibacterial activity of LAB may often be due to the production of organic acids, with a consequent pH reduction, or to the production of hydrogen peroxide (Gonzalez et al. 2007). The results also clearly demonstrated that LAB with antimicrobial activity could survive within the intestinal milieu exerting antipathogenic effects as LAB strains were reported to inhibit the growth of pathogenic bacteria in several studies (Collado et al. 2005; Bernbom et al. 2006).

Antibiotic resistant profile clearly revealed that most of the selected LAB strains are highly resistant to antibiotics with some exception. Qing et al. (2007) has reported wider range (5–30 μg ml−1) of antibiotic resistance by Enterobacter cloacae against ampicillin, erythromycin, kanamycin and rifampicin. The resultant data elucidated that isolated LAB strains acquired the characteristically broad spectrum of antibiotic resistance along with metal-resistant ability. The antibiotic resistant properties also indicated that the isolated LAB strains would be able to survive in the environment and intestinal milieu by withstanding the undesirable situation occurred because of occasional high antibiotic concentrations.

The identified 26 Cd- and Pb-resistant LAB belonged to different species of genus Lactobacillus, Pediococcus and Enterococcus, which indicated that mud and sludge of effluents transporting channels, wastewater treatment plants and coastal habitats could be used as potential sources of various species of metal-resistant probiotic LAB. No clear metal-specific speciation of identified LAB species was prevalent in this investigation. Varieties of LAB such as Lactobacillus, Lactococcus, Enterococcus, Leuconostoc and Weissella genera are present in soil (Chen et al. 2005) and coastal sediment (Zamudio-Maya et al. 2008) samples.

The MIC is the lowest concentration of the heavy metals that completely inhibited the growth of the strains (Froidevaux et al. 2001). The high magnitude of the MIC values of 26 LAB strains against Cd (50 to >1000 mg l−1) and Pb (>2000 mg l−1) also implied that selected LAB acquired high metal resistance through exposure to metal loading environment and can easily survive in high Cd- and Pb-contaminating environment. Qing et al. (2007) reported the 1200 and 2000 mg l−1 MIC values of Cd in Bacillus cereus and Enterobacter cloacae, respectively. The ability of micro-organisms to resist antibiotics and metals seems to be the result of exposure to metal contaminated environments that cause coincidental selection of resistance factors for heavy metals and antibiotics (Foster 1983; Qing et al. 2007).

Results of MRE quantified the μg metal removed per hour by per wet weight of LAB cell that simply helps to compare the metal-removal abilities amongst the LAB under described experimental conditions without depicting the mass–balance relationship in removing metals. It demonstrated that (i) higher Cd-removal efficiency classes (>0·13 to ≤0·145 μg h−1 mg−1) occurred at higher percentage (66·6%) amongst the Cd-resistant LAB than amongst the Pb-resistant LAB and (ii) the cent per cent strains of higher Pb-removal efficiency classes (>0·5 to ≤0·602 μg h−1 mg−1) were noticeably found amongst the Pb-resistant LAB groups (Fig. 2a,b). From this result, it is obvious that LAB with elevated Cd- and Pb-removal efficiency may be obtained at higher percentage from the Cd- and Pb-resistant LAB strains, respectively. Irrespective of tested metal species (Cd and Pb), no clear relationship was found between the resistant patterns and metal-removal efficiencies, indicating the existence of variations in resistant mechanism amongst the examined LAB. Moreover, in consideration of two metal species, each LAB showed higher degree of Pb-resistance and Pb-removal efficiencies than that of the Cd. It might also be signified that high metal-resistance and MRE are specific properties of the identified LAB strains. They, thus, can be used for removing heavy metals from aqueous environment. However, 7 (Lact. amylovorus– Cd54-2, Ped. acidilactici Cd61-7, Lact. reuteri Cd69-12 and Cd70-13, Ped. acidilactici Pb82-8, Lact. dextrinicus Pb96-19 and Ent. faecium Cd109-16) and 4 (Lact. reuteri– Pb71-1 and Pb73-2, Ent. faecium Pb85-9 and Lact. dextrinicus Pb96-19) strains pronounced relatively higher Cd- and Pb-removal efficiencies, respectively, than that of the rest LAB (Fig. 2a,b). Metal resistance and excellent removal properties suggested that these selected LAB strains could have huge potential for treating heavy metal-contaminated water especially for drinking water treatment. According to Halttunen et al. (2007), LAB effectively removes heavy metals from water.

Furthermore, the study of metal removal from MRS media demonstrated that Cd70-13 and Pb71-1 LAB strains have high efficiency in removing heavy metals suggesting that they would be a potential candidate for removing Cd and Pb, respectively (Fig. 3a,b). The increment (in Cd70-13) or unchanged (in Pb71-1) metal content on OSM at 48 h indicating the mechanism of heavy metals entrapped by extracellular polysaccharides on the OSM and rapidly transported towards ICS of membrane. The concentration of metals was decreased in ICS of both strains with time that signified the mechanism of gradual release of heavy metal after first step of rapid and huge uptake. The cell walls of the Gram-positive bacteria are efficient metal chelators, and in Bacillus subtilis, the carboxylic group of the glutamic acid of peptidoglycan was the major site of metal deposition (Gadd 1990). In addition to metal removal proficiency, selected LAB (Cd70-13 and Pb71-1) showed excellent adhesion ability to fish intestinal mucus that is the prerequisite for colonization in intestinal milieu. Therefore, it may also be concluded that selected Cd70-13 and Pb71-1 strains could be applied as potential Cd and Pb removing probiotic to uptake metals from food and intestinal environment. Phylogenetic analysis revealed that Cd70-13 and Pb71-1 are two strains of Lact. reuteri.Ibrahim et al. (2006) showed that the Lactobacillus rhamnosus LC-705, Propionibacterium freudenreichii ssp. shermanii JS strains and their combinations were found to bind Cd and Pb efficiently at low concentration ranges commonly observed in foods.

In this study, probiotic and the best metal removal properties of these selected 7 and 4 LAB strains were characterized, indicating their high potential to be able to survive in the intestinal milieu and uptake respective metals from the ambience. In addition, properties of excellent metal removal from MRS medium and adhesion ability to fish mucus demonstrated that the newly selected Lact. reuteri Cd70-13 and Pb71-1 strains could be applied as efficient probiotic candidates for food and in vivo challenge study in intestinal milieu of fish to uptake Cd and Pb from intestinal milieu, respectively, to solve the problems of heavy metal contamination and bioaccumulation. As this study focused on heavy metals removal from water and MRS medium, metal removal of selected probiotic in challenge experiments in natural gastro-intestinal conditions should be studied in further experiments to elucidate the exact removal capacity and mechanism. It is pertinent to mention here that our pilot trial for the application of selected metal-resistant LAB in fish exhibited a promising viability and metal removal effect to lower the heavy metal bioaccumulation within fish tissue (data not shown). Moreover, the potential probiotic strains were isolated from nonfood grade environmental (mud and sludge) samples; therefore, it can be safely used in fish and other animals. The food safety and pathogenic studies are essential before possible application for human purposes, because probiotics those are generally recognized as safe for human are isolated from foods or human faeces. However, a great effort is also needed to investigate potential metal uptaking LAB strains more characteristically, considering the samples of various metal polluted sites where chance factor will be high for opportunistic obtaining of such potential metal uptaking LAB. Finally, it is also obvious that the exploration of such characteristically potential probionts from heavy metals polluted environmental samples will open a new biotechnological approach to protect lives from the problems of heavy metal bioaccumulation and biomagnification.


The authors are thankful to Govt. of Japan for supporting the Grant-in-Aid for Scientific Research Fund. Dr Bhakta is also especially grateful to JSPS for providing the fellowship under the ‘FY2009 JSPS postdoctoral fellowship for foreign researcher’.