Journal of Applied Microbiology ISSN 1364-5072
ORIGINAL ARTICLE
Probiotic bacteria survive in Cheddar cheese and modify populations of other lactic acid bacteria B. Ganesan1,2, B.C. Weimer3, J. Pinzon3, N. Dao Kong3, G. Rompato4, C. Brothersen1,2 and D.J. McMahon1,2 1 2 3 4
Dairy Technology and Innovation Laboratory, Western Dairy Center, Utah State University, Logan, UT, USA Department of Nutrition, Dietetics, and Food Sciences, Utah State University, Logan, UT, USA Department of Population Health and Reproduction, University of California, Davis, CA, USA Center for Integrated BioSystems, Utah State University, Logan, UT, USA
Keywords Cheddar cheese, low fat, nonculturability, probiotic, survival. Correspondence Balasubramanian Ganesan, Dairy Technology and Innovation Laboratory, Western Dairy Center, Utah State University, 8700 Old Main Hill, Logan, UT 84322, USA. E-mail: [email protected]; and Bart C. Weimer, University of California, Davis School of Veterinary Medicine 1089 Veterinary Medicine Dr. VM3B, Room 4023 Davis, CA 95616, USA. E-mail: [email protected] 2013/2047: received 8 October 2013, revised 6 January 2014 and accepted 12 February 2014 doi:10.1111/jam.12482
Abstract Aims: Starter lactic acid bacteria in Cheddar cheese face physico-chemical stresses during manufacture and ageing that alter their abilities to survive and to interact with other bacterial populations. Nonstarter bacteria are derived from milk handling, cheese equipment and human contact during manufacture. Probiotic bacteria are added to foods for human health benefits that also encounter physiological stresses and microbial competition that may mitigate their survival during ageing. We added probiotic Lactobacillus acidophilus, Lactobacillus casei, Lactobacillus paracasei and Bifidobacterium animalis subsp. lactis to full-fat, reduced-fat and low-fat Cheddar cheeses, aiming to study their survival over 270 days of ageing and to determine the role of the cheese matrix in their survival. Methods and Results: Probiotic and other lactic acid bacterial populations were enumerated by quantitative PCR using primers specifically targeting the different bacterial genera or species of interest. Bifidobacteria were initially added at 106 CFU g1 cheese and survived variably in the different cheeses over the 270-day ageing process. Probiotic lactobacilli that were added at 107 CFU g1 cheese and incident nonstarter lactobacilli (initially at 108 CFU g1 cheese) increased by 10- to 100-fold over 270 days. Viable bacterial populations were differentiated using propidium monoazide followed by species-specific qPCR assays, which demonstrated that the starter and probiotic microbes survived over ageing, independent of cheese type. Addition of probiotic bacteria, at levels 100-fold below that of starter bacteria, modified starter and nonstarter bacterial levels. Conclusions: We demonstrated that starter lactococci, nonstarter lactobacilli and probiotic bacteria are capable of surviving throughout the cheesemaking and ageing process, indicating that delivery via hard cheeses is possible. Probiotic addition at lower levels may also alter starter and nonstarter bacterial survival. Significance and Impact of the Study: We applied qPCR to study multispecies survival and viability and distinctly enumerated bacterial species in commercial-scale Cheddar cheese manufacture.
Introduction Probiotic bacteria are defined as ‘live micro-organisms which when administered in adequate amounts confer a 1642
health benefit on the host’ (FAO/WHO 2002; Morelli and Capurso 2012). The consumption of probiotic bacteria is reported to confer many health benefits such as preventing gut inflammation, immunomodulation, preventing
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
food allergy and also potentially provide anticarcinogenic activity (Kailasapathy 2002; Caglar et al. 2005; Mengheri 2008). As early as infancy, probiotic bacterial species that are components of the native microbiota of breast milk enter the human colon. Probiotic bacteria are also supplemented via infant formula (Moreno Villares 2008; de Vrese and Schrezenmeir 2008) to aid milk oligosaccharide digestion (Macfarlane et al. 2008), whereas adults typically source probiotics from fermented milks and nutritional supplements. The consumption of probioticsupplemented cheese boosts innate immunity of geriatric patients (Ibrahim et al. 2010). However, probiotic bacteria must survive in foods to reach the human gastrointestinal system and to further modify gut microbiota (Kramer et al. 2009; Yu et al. 2009). Guidelines for probiotic bacteria recommend that identity at the species level must be established (FAO/WHO 2002). Probiotic bacteria such as Bifidobacterium longum, Bif. lactis, Lactobacillus acidophilus, Lact. casei and Lact. paracasei are usually added to yoghurt and other fermented milks (Heller 2001), and more recently to cheese (Gardiner et al. 1998) as delivery vehicles for human consumption. Lactococcus lactis is used as a starter culture to produce lactic acid for pH reduction during Cheddar cheese manufacture, whereas nonstarter lactic acid bacteria (NSLAB) and other microbes are either present in cheese milk or are added purposefully (Peterson and Marshall 1990; Trepanier et al. 1991; HabibiNajafi and Lee 1996; Swearingen et al. 2001). Probiotic bacteria are usually added to cheese milk and thus sequentially undergo physico-chemical stresses such as heat, acid, salt and cold during initial manufacture, as well as changes in redox potential over storage and distribution (Rallu et al. 1996; van de Guchte et al. 2002), as do other adventitious or added lactic acid bacteria (LAB). When cheeses are manufactured from milk of different fat contents (Bertoni et al. 2001), the sequence of abiotic stresses varies due to the different processing conditions used (Fox and Wallace 1997; Banks 2007; Coolbear et al. 2011), which may alter subsequent LAB survival during ageing. As NSLAB survive in cheese and grow over ageing, of which lactobacilli are the dominant species, probiotic lactobacilli species may also remain viable in Cheddar cheese during ageing until consumption to provide health benefits. Estimates of bacterial viability in different foods and environments vary based on the enumeration techniques used. Growth media-based enumeration discounts possible alternate physiological states of bacteria, such as nonculturability (Fenelon et al. 2000; Ganesan et al. 2007). Such growth-based observations led to a previous hypothesis that starter bacteria die and lyse to subsequently provide substrates that accelerate NSLAB growth
LAB survival in probiotic-added cheese
(Branen and Keenan 1969; Crow et al. 1995; Buist et al. 1997, 1998). However, lactococci, NSLAB and brevibacteria become nonculturable in carbohydrate-depleted media while remaining metabolically active (Ganesan et al. 2004, 2007) and maintain substrate transport for producing energy from alternate substrates (Ganesan et al. 2007). The declining lactococcal counts in cheese may represent a subpopulation of replicating cells, while a nonculturable population of cells that is unable to divide and is hence not enumerated on growth media (Kilcawley et al. 2011) coexists. Additionally, the enumeration of an entire genus using selective growth media does not accurately reflect the metabolic contributions of multiple species of a diverse genus, such as Lactobacillus (Peterson and Marshall 1990; Trepanier et al. 1991), that shares metabolic capabilities with other Lactobacillales (Makarova et al. 2006). Hence, alternate population assessment methods are needed to better estimate LAB species diversity in cheese and survival of probiotic bacterial species in foods used for delivery. Multiple studies at laboratory- or pilot-scale cheese manufacture have demonstrated probiotic bacterial survival in hard cheeses such as Cheddar, a representative set of which are listed in Table 1. According to these studies, even the same strains or species survive variably, with one group showing survival throughout ageing, but another demonstrating loss of viability of the same in 6– 8 weeks. Some studies were conducted in smaller scale (10–20 l of milk), which do not represent issues in commercial-scale manufacture such as adventitious NSLAB and bacteriophages’ presence. Notably, none of these studies enumerated survival of probiotic bacteria at the species level. Neither did these studies assess the impact of adding a microbe not usually found in cheese on the starter or NSLAB of cheeses. In one study that used qPCR for detecting bacteria (Achilleos and Berthier 2013), only soft cheeses were assessed over 24 h and did not address the challenge of detecting bacteria in aged semi-hard cheeses. To meet regulatory requirements, bacteria are isolated from foods (Giraffa and Rossetti 2004; Rossetti and Giraffa 2005) prior to species identification based on 16S or 23S rRNA genes using organism-specific primers (Salama et al. 1991; Klijn et al. 1995), single-gene sequencing, random DNA amplification, use of restriction enzyme digestion patterns (Byun et al. 2004; Rossetti and Giraffa 2005), gel electrophoresis by varying electric fields (pulse field) (Klein et al. 1998; Henri-Dubernet et al. 2008), temperature (Vasquez et al. 2001; Ogier et al. 2002) or chemical gradients (denaturing gradient) (Fasoli et al. 2003); however, these methods do not enumerate bacteria nor conclusively establish species identity. With advances in PCR amplification technology and the use of fluorescent
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1643
1644
NS, not specified.
Sharp et al. (2008) Ong and Shah (2008) Achilleos and Berthier (2013)
Antibiotic resistance mutant
107
Lact. acidophilus and Lact. helveticus Lact. paracasei Not disclosed
20 l
Microbial plating on selective media Microbial plating on selective media and qPCR
NS 108
108
10 l
136 kg
108
450 l
Lact. paracasei, Bif. sp. BB-12 Lactobacillus acidophilus, Bifidobacterium sp., Lact. casei, Lact. paracasei and Lact. rhamnosus Lact. casei
RAPD-PCR/gel electrophoresis on DNA extracted from grown colonies Microbial plating on selective media, staining and confocal microscopy Microbial plating on selective media
108
450 l
Antibiotic resistance mutant
108
Bif. sp.
Antibiotic resistance mutant
4 9 108
450 l
Microbial plating on selective media
5 9 107
Ent. faecium
RAPD-PCR/gel electrophoresis on DNA extracted from grown colonies
NS
Microbial plating on selective media
106
450 l
Bifidobacterium infantis
Daigle et al. (1999) Gardiner et al. (1999a) Gardiner et al. (1999b) Mc Brearty et al. (2001) Auty et al. (2001) Phillips et al. (2006)
25, 450 l (2 reps 9 2 strains only) 250 l
100 kg
Detection method in cheese
Targeted level of probiotics (CFU g1)
Enterococcus faecium
Bifidobacterium bifidum (immobilized; added during salting) Lactobacillus salivarius, Lact. paracasei
Dinakar and Mistry (1994)
Gardiner et al. (1998)
Probiotic bacteria used
Study
Scale of cheese manufacture
Table 1 Previous studies of probiotic addition to Cheddar cheese
Microbial plating on selective media and qPCR
Microbial plating on selective media
Microbial plating on selective media
0
6
108 108
3
107
9
NS
6
9
3 9 108 105 to 108
15
4 9 108
3
8
108
107
6
Cheese ageing period (months)
106 to 107
Microbial plating on selective media, staining and confocal microscopy Microbial plating on selective media 103 to 108
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Viability assessment in cheese
Detected levels of probiotics (CFU g1)
200
25
25
25
108 (confocal)
25
25
25
25
25
25
Detection limits of assays (CFU g1)
LAB survival in probiotic-added cheese B. Ganesan et al.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
LAB survival in probiotic-added cheese
labelling, reliable estimates of bacterial populations for a single genus or species are easily obtained using quantitative PCR (qPCR) (Matsuki et al. 2004; Kramer et al. 2009). Recent studies have established the applicability of this technology to other fermented dairy products such as yoghurt (Fasoli et al. 2003; Rademaker et al. 2006; Sieuwerts et al. 2008; Garcia-Cayuela et al. 2009; Sheu et al. 2009), and its applicability to cheese has been demonstrated. Nucleic acid extraction directly from a food matrix was previously limited by high carbohydrate, protein and lipid contents, but methods are now available to extract high-quality DNA for direct use in identification strategies. Challenges in direct nucleic acid extraction from cheese were previously circumvented by prior microbial cell separation from the cheese matrix for nucleic acid extraction (Rudi et al. 2005; Fernandez et al. 2006; Monnet et al. 2006; Fukushima et al. 2007) or using large sample quantities (≥10 g cheese) (Monnet et al. 2006, 2008; Sohier et al. 2012). Else, the methods only provided higher detection limits (≥1000 CFU g1) (Baruzzi et al. 2005; Rantsiou et al. 2008; Sheu et al. 2009) or are incapable of distinguishing added probiotic lactobacilli from NSLAB lactobacilli (Gardiner et al. 1998). Such approaches cannot determine changes of species levels in long-term aged cheeses. However, using genus- and species-specific qPCR, population changes of distinct groups of LAB specifically induced by probiotic addition can be determined. We hypothesized that the addition of specific probiotic bacteria to cheese during manufacture modifies starter and NSLAB lactobacilli survival in Cheddar cheese at different fat levels. In this study, three cheese types that had
different fat contents were made to determine the role of the physico-chemical conditions in such cheeses in allowing survival of probiotic lactobacilli and bifidobacteria by applying a qPCR-based bacterial detection method. We also assessed whether addition of probiotics at levels below that of starter bacteria altered starter or NSLAB levels. Further, the viability of the three groups of bacteria was also determined using propidium iodide-based qPCR assays. DNA extraction from small quantities of cheese (02–1 g) was optimized by coupling physicochemical and enzymatic lysis of bacteria to release nucleic acids, followed by phenol–chloroform-based applications to remove protein and lipid. We found that starter, NSLAB lactobacilli and probiotic bacteria survive throughout Cheddar cheese ageing with limited reduction in viability and that probiotic addition even at levels lower than starter LAB, altered levels of starter lactococci and NSLAB lactobacilli. Materials and methods Probiotic-added cheese manufacture Starter culture, probiotic adjunct, suppliers and the amount of culture used are indicated in Table 2. The targeted range of probiotic bacteria in the finished cheese was 36 9 106–107 CFU g1 (based on the requirement of ≥108 CFU per 28 g serving), and direct-vat-set cultures were added proportionally based on manufacturer’s specifications and expected cheese yields to achieve the target levels. The control cheese was made with Lc. lactis DVS850 starter culture only, while the probiotic cheeses each contained DVS850 and one probiotic culture.
Table 2 Starter culture, probiotic adjunct, suppliers and the amount of freeze-dried culture used Amount used in cheese (g)*
Supplier
Organism
Name
Full fat (160 kg cheese milk)
Reduced fat (135 kg cheese milk)
Low fat (135 kg cheese milk)
Chr. Hansen, Milwaukee, WI, USA Cargill Inc., Waukeshaw, WI, USA Chr. Hansen Chr. Hansen DSM Food Specialties, Logan, UT, USA DSM Food Specialties Chr. Hansen Chr. Hansen
Lactococcus lactis
DVS850
170
300
260
Bifidobacterium lactis
Bif-6
140
120
120
Bif. lactis Lactobacillus acidophilus Lact. acidophilus
BB-12 LA-5 L10
140 140 78
120 120 67
120 120 67
Lactobacillus casei Lact. casei Lactobacillus paracasei subsp. paracasei
L26 CRL-431 F19
64 140 350
55 120 300
55 120 300
*Inoculation levels were adjusted by milk quantity and manufacturer-provided information about cultures on numbers of viable bacteria in freeze-dried cultures to achieve the desired target ranges of 107 to 108 CFU g1 probiotic bacteria in cheese.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1645
LAB survival in probiotic-added cheese
B. Ganesan et al.
Manufacturing procedures for full-fat, reduced-fat and low-fat Cheddar cheeses are described in the supplementary materials and methods section, along with proximate analysis procedures for cheese.
proteinase K at 50°C for 1 h). This was followed by phenol–chloroform extraction of DNA and qPCR for enumerating lactococci using genus-specific primers. This lysis approach was also applied to cheeses tested for viable bacterial levels (see below).
Sampling After ageing for 5 days, the cheese was cut into 10 blocks of c. 1 kg each, vacuum packaged and stored at 3°C. For each replicate treatment, one block was randomly chosen for analysis at 5 days, 1, 2, 3, 4, 6 and 9 month of age.
DNA quality and yield DNA quality and yield were checked on a Nanodrop spectrophotometer (Agilent Technologies, Santa Clara, CA, USA). All samples used in qPCR had a A260/A280 ratio ≥18.
Extraction of genomic DNA from cheese To extract DNA from all three types of cheese, finely grated cheese (025 g) was suspended in TrizolLS (500 ll; Invitrogen, Carlsbad, CA, USA) along with glass beads (03 g, size 01 mm; sterile, acid-washed; BioSpec Products Inc., Bartlesville, OK, USA). The cheese-TrizolLS suspension was disrupted on a Mini-Beadbeater (BioSpec Products) for 30 s to lyse bacteria and further shaken with chloroform (200 ll) for 30 s to allow phase separation. The top aqueous phase was removed after centrifugation (12 000 g for 15 min at room temperature), and DNA was precipitated from the lower organic phase with 100% ethanol for 3 min. DNA was collected by centrifugation (2000 g for 5 min at 4°C), washed twice with 01 mol l1 sodium citrate in 10% ethanol (30 min at room temperature), once with 75% ethanol (15 min at room temperature), pelleted, air-dried and resuspended in sterile double-deionized water (ddH2O) This DNA was used to measure population changes of lactobacilli and bifidobacteria. Due to poor efficacy of glass bead lysis for lactococcal detection (Fig. S2), lactococcal lysis was separately accomplished by an enzymatic approach in which finely grated cheese (025 g) was treated with bacterial lytic enzymes (initially 50 mg ml1 lysozyme + 10 U ml1 mutanolysin at 37°C for 1 h, followed by 9 U ml1
Quantitative PCR Bacterial levels in cheese were determined by qPCR in triplicate reactions. Compliance to MIQE guidelines was ensured throughout the process of qPCR analyses, data extraction and data analysis (Table S5). All bacterial primers were selected from previous studies that designed genus- and species-specific primers for the starter and probiotic organisms (Table 3) and were subsequently validated by assays against the strains used in this study prior to use (see Fig. S1). Briefly, the total DNA extracted from cheese was used in a 25-ll reaction that included qPCR master Mix (HotStart-ITâ SYBRâ Green; USB Corp., Cleveland, OH, USA) and genus- and species-specific primers targeting the 16s rRNA gene (Table 3). Each final reaction mixture contained: template DNA, 500 ng, MgCl2, 25 mmol l1, primers, 10 pmol and each dNTP, 04 mmol l1. The qPCR was performed on a DNA Engine OPTICON2 (Bio-Rad Labs Inc., Hercules, CA, USA) with initial enzyme activation at 95°C for 5 min, followed by 40 cycles of: denaturation at 95°C for 15 s, annealing at 50°C for 30 s and extension at 72°C for 1 min. All DNA samples from probiotic-added cheeses were always assayed simultaneously with DNA samples from control cheeses (no probiotic added) using speciesspecific primers to confirm lack of amplification. Following
Table 3 Sequences of primers used for genus- or species-specific 16s ribosomal gene qPCR for different cheese bacteria Organism
Specificity
Primer sequence
Cheeses applied to
References
Lactococcus lactis
Genus
All
Klijn et al. (1995)
Bifidobacterium lactis
Species
Bif-6, BB-12
Ventura et al. (2001)
Lactobacillus
Genus
Control, LA-5, L10, L26, F19
Byun et al. (2004)
Lactobacillus acidophilus
Species
LA-5, L10
Delroisse et al. (2008)
Lactobacillus casei/paracasei
Species
F: 50 - GCGGCGTGGCTAATACATGC-30 R: 50 - CTGCTGCGTCCCGTAGGAGT-30 F: 50 -GTGGAGACACGTTTCCC-30 R: 50 -CACACACACAATCAATAC-30 F: 50 -TGGAAACAGRTGCTAATACCG-30 R: 50 -GTCCATTGTGGAAGATTCCC-30 F: 50 -GAGGCAGCAGTAGGGAATCTTC-30 R: 50 -GGCCAGTTACTACCTCTATCCTTCTTC-30 F: 50 -GCACCGAGATTCAACATGG-30 R: 50 -GGTTCTTGGATYTATGCGGTATTAG-30
L26, F19
Byun et al. (2004)
1646
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
MIQE-compliant procedures, results from all qPCR assays were extracted using OPTICON monitor software v2.2 (Bio-Rad). Viable bacterial qPCR assay using propidium monoazide The viability of bacteria present in cheese was verified by a qPCR-based assay, using the DNA-chelating reagent propidium monoazide (PMA; Biotium Inc., Hayward, CA, USA). Briefly, the assay was conducted as follows: a small portion of cheese (025 g) was thoroughly minced, suspended in an equal amount (w/v) of sterile ddH2O, and a stock of 20 mmol l1 PMA was added to the cheese suspension to obtain a final concentration of 50 lmol l1, as described by Nocker et al. (2007). The addition and mixing was carried out in light-transparent tubes that were then kept in dark for 10 min to allow PMA-nucleic acid binding, followed by exposure to a 650 W halogen lamp for 5 min, with samples kept on ice, to quench the reaction. After light exposure, samples were treated with lysozyme–mutanolysin and DNA was extracted for qPCR-based enumeration. For live-dead assays by qPCR using PMA, the difference in Cq times with and without PMA addition, denoted as DCq, was monitored over cheese age. Statistical analysis All cheeses were made in two replicates that were manufactured from different batches of milk. Cheese samples were collected for qPCR enumeration at 5 days, 1, 2, 3, 4, 6 and 9 months. Each sample was analysed by qPCR in triplicate, and the results of bacterial counts from qPCR assays were analysed as a nested factorial repeatedmeasures design (Ganesan et al. 2007) to understand the impact of cheese type on survival of different added probiotics and the impact of cheese type and strain on survival of NSLAB. JMP v7 statistical software (The SAS Institute, Cary, NC, USA) was used for repeated-measures statistical analyses. For the time series analyses, significance was assigned at a = 005/n, where n is the number of time points. Two-sample statistical comparisons within a particular condition when necessary were performed by two-tailed Student’s t-tests with Microsoft Excel software (Redmond, CA, USA), and significance was assigned at a = 005. Results
LAB survival in probiotic-added cheese
bacterial species were added to the different cheeses. Changes in specific LAB populations over ageing were assessed by qPCR using a species primer set that corresponded to the species added to the cheeses (validation and variability results in supplementary material). We developed a sensitive qPCR assay based on previously validated primers that provided reliable detection even at 10–500 CFU g1 from Cheddar cheese (Table S2). The targeted levels of probiotics in cheese were at least 10 000-fold higher than our detection limits. Primarily, probiotic bacteria were added targeting a final level of 36 9 106–107 CFU g1 cheese, which was accomplished in all full- and low-fat cheeses (Fig. S3). Interestingly, added probiotic lactobacilli populations were consistently lower than the target level by 10-fold initially in reduced-fat Cheddar cheeses, but rose to attain levels of c. 107–108 CFU g1 cheese within a month of ageing (Fig. S2), indicating that they grew in the cheese by 10- to 100-fold. However, at any time, probiotic lactobacilli levels were only 1–10% of that of total lactobacilli (Figs S2 and S4), confirming that added probiotic and NSLAB lactobacilli can be distinctly enumerated using qPCR. Probiotic Lact. casei/paracasei survived through 270 days of ageing in all Cheddar cheeses either at comparable levels or grew by 10- to 70-fold (Table 4), with differences in final levels attained within each fat level (P ≤ 005). Added Lact. acidophilus exhibited a 16- to 37fold reduction in population (P ≤ 005) from the initial level of addition during ageing in low-fat cheese (Table 4). However, Lact. acidophilus strains did not differ in survival patterns over time across the different fat
Table 4 Fold change of different probiotic bacterial populations in cheeses over 270 days of ageing Fold change in populations (CFU g1 ratio of 270 days/0 days)* Probiotic organism Lactobacillus acidophilus LA-5 Lact. acidophilus L10 Lactobacillus casei CRL-431 Lact. casei L26 Lactobacillus paracasei F19 Bifidobacterium lactis Bif-6 Bif. lactis BB-12
Full fat
Reduced fat
Low fat
19 23
13 09
16 04
23 10 41 20
230 09 93 09
37 14 71 06
16 04 15 05
74 02 22 10
35 03 52 08
12 10
28 11
6900 10
1500 10
27 10
3200 100
LAB population changes during ageing Three types of Cheddar cheeses—full, reduced and low fat—were manufactured (Table S1), and seven probiotic
Negative sign indicates reduction. *Values are indicated followed by standard deviations after the ‘’ sign.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1647
LAB survival in probiotic-added cheese
B. Ganesan et al.
levels (P > 005). In contrast, Bif. lactis strains exhibited varying survival patterns (P ≤ 005) within each fat level, with the Bif-6 strain populations declining by 6900-fold in low-fat cheese within 270 days, whereas the BB-12 strain declined by 3200-fold in full-fat cheese (Table 4). Bif-6 remained at similar levels to initial inoculum in reduced-fat cheese (P > 005), whereas BB-12 increased by 1500-fold (P ≤ 005) (Table 4). Notably, estimates of bifidobacterial levels varied by 100- to 1000-fold within replicates for 43% of cheeses across age and fat level, which precluded any strong conclusions about their population variations. Such variation was much less prevalent in probiotic lactobacilli (9% cheeses with 100- to 1000fold variation). However, bifidobacteria were exclusively detected by qPCR throughout cheese ageing, which confirms their viability in cheese. To understand the role of cheese fat content on microbial survival, we directly compared the probiotic levels in reduced-fat and low-fat cheeses. Interestingly, the lower fat level appeared to reduce bacterial survival, as highlighted by significant decrease (P ≤ 005) in numbers of Lact. acidophilus over ageing in low-fat cheese. However, Lact. casei/paracasei grew in both reduced- and low-fat cheeses. These results indicated that fat in cheese may aid survival of selected species of lactobacilli during ageing. As probiotic bacteria were added at levels of 107 CFU 1 g and survived variably, they may alter the levels of starter LAB and NSLAB in ageing Cheddar cheese. We assessed the effect of probiotic addition by measuring changes in overall lactobacilli populations using a primer set targeting the Lactobacillus genus. Notably, in all cheeses tested, levels of total lactobacilli were 10- to 100fold higher than levels of added probiotic, and the patterns of survival of these two groups were different (P ≤ 005) (Figs S2 and S4). However, survival and growth of total lactobacilli in the probiotic lactobacilliadded cheeses differed due to the different strains added and fat level (P ≤ 005). While the overall populations of total lactobacilli remained at 108 CFU g1 or higher in all cheeses, the addition of Lact. acidophilus La-5 and Lact. casei CRL431 significantly (P ≤ 005) altered the levels of total lactobacilli, dependent on cheese fat level. For example, at 270 days of ageing, La-5 addition reduced total lactobacilli by eightfold in the reduced-fat cheese, but there was no change in the low-fat cheese (P ≤ 005; Table 5). Addition of CRL431 decreased total lactobacilli by sevenfold in reduced-fat Cheddar cheese and increased total lactobacilli by sevenfold in low-fat cheese (P ≤ 005; Table 5). Addition of other probiotic organisms did not alter total lactobacilli levels (P > 005), but cheese fat level altered (P ≤ 005) the total lactobacilli levels. One such example is the cheeses to which Lact. paracasei F19 was added, in which total lactobacilli levels 1648
Table 5 Fold change of total lactobacilli populations in cheeses added with different probiotic lactobacilli over 270 days of ageing Fold change in populations (CFU g1 ratio of 270 days/0 days)*
Probiotic organism Lactobacillus acidophilus LA-5 Lact. acidophilus L10 Lactobacillus casei CRL-431 Lact. casei L26 Lactobacillus paracasei F19
Full fat (n = 2)
Reduced fat (n = 2)
Low fat (n = 2)
16 27
81 13
14 04
35 07 16 08
83 10 69 07
60 16 69 07
58 05 16 04
17 02 56 07
15 04 342 13
Negative sign indicates reduction. *Values are indicated followed by standard deviations after the ‘’ sign.
had increased 34-fold in 270-day-old low-fat cheese (P ≤ 005), but had decreased sixfold in reduced-fat cheese (Table 5). These results indicate that fat reduction in cheeses and probiotic addition alters survival of NSLAB lactobacilli over cheese ageing. As NSLAB levels were altered by addition of probiotic lactobacilli and bifidobacteria, we also measured changes in lactococcal populations using PCR targeting Lc. lactis species in all three cheese types in a limited set of samples. Lactococcal populations in the control (no probiotic added) cheeses were compared to levels in Lact. acidophilus L-10-added and Bif. lactis Bif-6-added cheeses. Lactococcal counts in control cheeses ranged between 75 9 107 and 8 9 108 CFU g1 and did not change over time. Addition of different probiotics significantly (P ≤ 005) altered lactococcal populations and also the effects varying within each fat level. For example, addition of Bif-6 did not change lactococcal levels within 270 days in low-fat cheeses, but decreased lactococcal populations in reduced-fat cheeses by 10-fold compared to without Bif-6 addition (P ≤ 005; Fig. 1), whereas L-10 addition consistently allowed lactococci to survive at 10to 100-fold higher levels than without probiotic addition in low-fat cheeses (P ≤ 005), but did not alter lactococcal levels in reduced- and full-fat cheeses. Viability of LAB in cheese LAB that do not survive the harsh cheese conditions may die, and their cell walls and membranes may rupture, releasing DNA and intracellular contents into the cheese matrix where it may remain stable and be detected using PCR, which could artificially elevate the bacterial estimates during qPCR assays. Such extracellular DNA can be removed by treating samples with DNase enzyme prior
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
LAB survival in probiotic-added cheese
10
CFU gm–1
9 8 7 6 5
0
90
120
270 Days
10
CFU gm–1
9 8 7 6 5
0
90
120
270 Days
10
detection using species-specific primers, did not differ (P > 005) in qPCR Cq times from nontreated cheeses (Fig. 2), indicating that extracellular DNA was not available for PCR detection and did not contribute to the total count using qPCR. This indicated that qPCR assays are exclusively based on DNA from live bacteria (including bacteria that are nonculturable) and that DNA from dead bacteria is not being enumerated. Secondarily, bacterial cells may exist in a physiologically dead state wherein they are partially autolysed or their cell walls or membranes are partly compromised to create a leaky membrane, but still contain intact DNA that could be co-extracted along with live cells’ DNA and cause overestimation. To assess this potential interference, we conducted qPCR assays on all three cheese types at 5 days, 4, 6 and 9 months age, to which PMA, a DNAchelating dye, was added prior to DNA extraction. PMA selectively permeates across and binds DNA in dead/compromised cells and thus limits PCR amplification from the chelated DNA, which consequently will increase the DCq over ageing if cell membrane is compromised. The procedure of PMA treatment for cheese was validated by adding 107–108 logarithmic growth phase cells to a cheese sample obtained from another source prior to PMA treatment, and DNA extraction was followed by qPCR enumeration. With and without PMA treatment, the DCq between samples was similar (P > 005), which proved that the added cells were live (data not shown). We conducted qPCR assays targeting the starter lactococci in cheese without added probiotics and two cheeses with
9 CFU gm–1
25 8
20 7
15
5
Cq
6
10 0
90
120
270 Days
Figure 1 Lactococcal starter bacteria population levels in Cheddar cheeses estimated by qPCR. Vertical error bars represent standard deviations of log10 (CFU g1) assessed from two replicates of cheese. Top panel, full-fat cheese, middle panel, reduced-fat cheese, and bottom panel, low-fat cheese. (●) Control—probiotic not added; (■) probiotic Bifidobacterium lactis Bif-6 added; (♦) probiotic Lactobacillus acidophilus L10 added.
to cellular lysis to degrade prereleased DNA. However, DNase-treated 3-month-old cheeses containing added Lact. casei L26 and Bif. lactis Bif-6, when tested by qPCR
5
0 Genus Species Lactobacillus primer type Figure 2 Quantitative PCR Cq times of Lactobacillus genus and species determined from cheeses treated with DNase enzyme prior to bacterial lysis to remove extracellular DNA. Cheese age was 3 months and contained added probiotic Lactobacillus casei L26. White bar, no DNase treatment, grey bar, DNase treated prior to lysis and DNA extraction.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1649
LAB survival in probiotic-added cheese
B. Ganesan et al.
different probiotic species (Lact. acidophilus La-5 and Bif. lactis Bif-6) in full-, reduced- and low-fat cheeses over 9 months of cheese ageing. In almost all cases, the DCq did not change significantly (P > 005) over cheese age for lactococci (Fig. 3a), suggesting that over ageing up to 9 months, starter lactococci survive the ageing process without loss in viability no change in the membrane permeability, despite declines in plate counts as demonstrated by other studies (Crow et al. 1995; Buist et al. 1998; Oberg et al. 2011). The DCq for lactococci did, however, increase slightly in the low-fat cheese with added bifidobacteria between 0 and 120 days (Fig. 3a) and remained at similar levels up to 270 days, suggesting that the membrane permeation increase was insignificant over long-term ageing of cheese. Lactococci survived even after probiotic addition without loss in viability, suggesting that the impact of added probiotics is likely minimal. This approach also verified that LAB levels assessed by qPCR identified viable and intact cells that were present in the cheese matrix and not dead or compromised cells. Assessing probiotic survival using the PMA-added cheese samples showed that added lactobacilli also survived in Cheddar cheese over age, demonstrated by either a significant reduction in DCq (P ≤ 005, full-fat cheese,
14
Fig. 3b) or modest increases in DCq (P ≤ 005, reducedand low-fat cheeses, Fig. 3b). Similarly, small but significant changes in DCq (P ≤ 005) were also observed with bifidobacteria-added cheese (Fig. 3c), showing that bifidobacteria also survived, with some loss of viability on one strain, which matches assessment by standard qPCR that showed some loss of bifidobacterial viability in 270day-old Cheddar cheeses (Table 4 and Fig. S3). Additional bifidobacterial species need to be assessed for their viability in cheese. However, application of the PMA assay confirmed that qPCR detected viable LAB and bifidobacteria in the cheese matrix and that starter lactococci, NSLAB and probiotic lactobacilli, and bifidobacteria remained viable throughout 270 days of Cheddar cheese ageing. Discussion To consistently produce commercial cheeses, the manufacturing process of aged semi-hard and hard cheeses is designed to physico-chemically control their microbial diversity. For example, Cheddar cheese is initially inoculated with high numbers of starter lactococci (105– 106 CFU ml1 milk) that replicate rapidly during cheese
(a)
ΔCq (with - without PMA)
12 10 8 6 4 2
0
9 (b) 8 7 6 5 4 3 2 1 0
50
0
180 Time (days)
1650
100
150 Time (days)
ΔCq (with - without PMA)
ΔCq (with - without PMA)
0
200
18 (c) 16 14 12 10 8 6 4 2 0
250
0
300
180 Time (days)
Figure 3 Comparisons of live lactococcal (a), probiotic Lactobacillus acidophilus (b), and probiotic Bifidobacterium lactis (c) populations by determining qPCR Cq times from cheese samples’ DNA extracted with and without propidium monoazide addition. Data are presented as the difference in Cq times, DCq, as explained in Materials and Methods. Vertical error bars represent standard deviations of DCq assessed from two replicates. In panel a, □ full-fat control (probiotic bacteria not added during cheesemaking), ■, low-fat control (probiotic bacteria not added during cheesemaking), 4, Bif. lactis Bif-6 added to full-fat cheese, ♢ Bif. lactis Bif-6 added to low-fat cheese, ○, Lactobacillus casei L10 added to full-fat cheese, ●, Lact. casei L10 added to low-fat cheese. In panel b, white bars, full-fat cheese, grey bars, reduced-fat cheese, and black bars, low-fat cheese, all containing Lact. acidophilus LA-5 added to cheese. In panel c, white bars, full-fat cheese, grey bars, reduced-fat cheese, and black bars, low-fat cheese, all containing Bif. lactis Bif-6 added to cheese. Statistical significance over time was determined only within each cheese.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
manufacture and chiefly degrade lactose to produce lactic acid and reduce cheese pH during early ageing. When adjunct or probiotic bacteria are included, the balance of LAB populations may be altered. Chiefly, the added probiotics must survive in sufficient numbers (target 36 9 106 to 36 9 107 CFU g1 cheese, to deliver 109 bacteria per serving) throughout long-term ageing of cheese to provide desired health benefits upon human consumption (Medici et al. 2004; Ouwehand et al. 2012). Among the added lactobacilli, Lact. casei and Lact. paracasei are commonly found in Cheddar cheese, are even used as adjuncts and are expected to survive throughout cheese ageing (Peterson and Marshall 1990). To detect added probiotic bacteria, we applied primers previously designed in other studies (Table 3) after selecting for specificity against the strains used in our study. Better sensitivity was achieved compared to previous studies by improving DNA extraction from Cheddar cheese using phenol–chloroform-based methods (min. 100 CFU g1 detected; Fig. S1). By applying genus- and species-specific primers, we were able to distinctly enumerate probiotic lactobacilli and NSLAB lactobacilli. The consistent presence of NSLAB lactobacilli at 10- to 100fold higher levels precluded any interference by added probiotic lactobacilli in NSLAB enumeration. The use of qPCR also provided a specific and reliable quantitative assay throughout cheese ageing, as opposed to growth media-based enumeration techniques (McSweeney et al. 1993; Gardiner et al. 1998; Crow et al. 2001; Oberg et al. 2011) where the distinction between starter and NSLAB becomes difficult beyond 90 days of age. At certain cheese ages, bacterial levels estimated by qPCR varied by as much as 10- to 1000-fold, which appears to be common in the cheese matrix based on other studies that used media-based enumeration approaches (Table 1). Total lactobacilli were present at much higher levels of 108 to 109 CFU g1 initially than expected (500 kg milk per batch), equipment is frequently subject to cleaning-in-place, which provides excellent control over unwanted NSLAB presence. However, even at this high initial NSLAB level, we could still distinguish added probiotic from NSLAB, which demonstrated the strength of using qPCR to detect added bacterial species. Added probiotic lactobacilli survived in cheese over 270 days of ageing, even growing 10 to 1000-fold (Table 4, Fig. S2). This observation is not surprising, considering that NSLAB lactobacilli are acquired from different sources and only found at low levels (
ORIGINAL ARTICLE
Probiotic bacteria survive in Cheddar cheese and modify populations of other lactic acid bacteria B. Ganesan1,2, B.C. Weimer3, J. Pinzon3, N. Dao Kong3, G. Rompato4, C. Brothersen1,2 and D.J. McMahon1,2 1 2 3 4
Dairy Technology and Innovation Laboratory, Western Dairy Center, Utah State University, Logan, UT, USA Department of Nutrition, Dietetics, and Food Sciences, Utah State University, Logan, UT, USA Department of Population Health and Reproduction, University of California, Davis, CA, USA Center for Integrated BioSystems, Utah State University, Logan, UT, USA
Keywords Cheddar cheese, low fat, nonculturability, probiotic, survival. Correspondence Balasubramanian Ganesan, Dairy Technology and Innovation Laboratory, Western Dairy Center, Utah State University, 8700 Old Main Hill, Logan, UT 84322, USA. E-mail: [email protected]; and Bart C. Weimer, University of California, Davis School of Veterinary Medicine 1089 Veterinary Medicine Dr. VM3B, Room 4023 Davis, CA 95616, USA. E-mail: [email protected] 2013/2047: received 8 October 2013, revised 6 January 2014 and accepted 12 February 2014 doi:10.1111/jam.12482
Abstract Aims: Starter lactic acid bacteria in Cheddar cheese face physico-chemical stresses during manufacture and ageing that alter their abilities to survive and to interact with other bacterial populations. Nonstarter bacteria are derived from milk handling, cheese equipment and human contact during manufacture. Probiotic bacteria are added to foods for human health benefits that also encounter physiological stresses and microbial competition that may mitigate their survival during ageing. We added probiotic Lactobacillus acidophilus, Lactobacillus casei, Lactobacillus paracasei and Bifidobacterium animalis subsp. lactis to full-fat, reduced-fat and low-fat Cheddar cheeses, aiming to study their survival over 270 days of ageing and to determine the role of the cheese matrix in their survival. Methods and Results: Probiotic and other lactic acid bacterial populations were enumerated by quantitative PCR using primers specifically targeting the different bacterial genera or species of interest. Bifidobacteria were initially added at 106 CFU g1 cheese and survived variably in the different cheeses over the 270-day ageing process. Probiotic lactobacilli that were added at 107 CFU g1 cheese and incident nonstarter lactobacilli (initially at 108 CFU g1 cheese) increased by 10- to 100-fold over 270 days. Viable bacterial populations were differentiated using propidium monoazide followed by species-specific qPCR assays, which demonstrated that the starter and probiotic microbes survived over ageing, independent of cheese type. Addition of probiotic bacteria, at levels 100-fold below that of starter bacteria, modified starter and nonstarter bacterial levels. Conclusions: We demonstrated that starter lactococci, nonstarter lactobacilli and probiotic bacteria are capable of surviving throughout the cheesemaking and ageing process, indicating that delivery via hard cheeses is possible. Probiotic addition at lower levels may also alter starter and nonstarter bacterial survival. Significance and Impact of the Study: We applied qPCR to study multispecies survival and viability and distinctly enumerated bacterial species in commercial-scale Cheddar cheese manufacture.
Introduction Probiotic bacteria are defined as ‘live micro-organisms which when administered in adequate amounts confer a 1642
health benefit on the host’ (FAO/WHO 2002; Morelli and Capurso 2012). The consumption of probiotic bacteria is reported to confer many health benefits such as preventing gut inflammation, immunomodulation, preventing
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
food allergy and also potentially provide anticarcinogenic activity (Kailasapathy 2002; Caglar et al. 2005; Mengheri 2008). As early as infancy, probiotic bacterial species that are components of the native microbiota of breast milk enter the human colon. Probiotic bacteria are also supplemented via infant formula (Moreno Villares 2008; de Vrese and Schrezenmeir 2008) to aid milk oligosaccharide digestion (Macfarlane et al. 2008), whereas adults typically source probiotics from fermented milks and nutritional supplements. The consumption of probioticsupplemented cheese boosts innate immunity of geriatric patients (Ibrahim et al. 2010). However, probiotic bacteria must survive in foods to reach the human gastrointestinal system and to further modify gut microbiota (Kramer et al. 2009; Yu et al. 2009). Guidelines for probiotic bacteria recommend that identity at the species level must be established (FAO/WHO 2002). Probiotic bacteria such as Bifidobacterium longum, Bif. lactis, Lactobacillus acidophilus, Lact. casei and Lact. paracasei are usually added to yoghurt and other fermented milks (Heller 2001), and more recently to cheese (Gardiner et al. 1998) as delivery vehicles for human consumption. Lactococcus lactis is used as a starter culture to produce lactic acid for pH reduction during Cheddar cheese manufacture, whereas nonstarter lactic acid bacteria (NSLAB) and other microbes are either present in cheese milk or are added purposefully (Peterson and Marshall 1990; Trepanier et al. 1991; HabibiNajafi and Lee 1996; Swearingen et al. 2001). Probiotic bacteria are usually added to cheese milk and thus sequentially undergo physico-chemical stresses such as heat, acid, salt and cold during initial manufacture, as well as changes in redox potential over storage and distribution (Rallu et al. 1996; van de Guchte et al. 2002), as do other adventitious or added lactic acid bacteria (LAB). When cheeses are manufactured from milk of different fat contents (Bertoni et al. 2001), the sequence of abiotic stresses varies due to the different processing conditions used (Fox and Wallace 1997; Banks 2007; Coolbear et al. 2011), which may alter subsequent LAB survival during ageing. As NSLAB survive in cheese and grow over ageing, of which lactobacilli are the dominant species, probiotic lactobacilli species may also remain viable in Cheddar cheese during ageing until consumption to provide health benefits. Estimates of bacterial viability in different foods and environments vary based on the enumeration techniques used. Growth media-based enumeration discounts possible alternate physiological states of bacteria, such as nonculturability (Fenelon et al. 2000; Ganesan et al. 2007). Such growth-based observations led to a previous hypothesis that starter bacteria die and lyse to subsequently provide substrates that accelerate NSLAB growth
LAB survival in probiotic-added cheese
(Branen and Keenan 1969; Crow et al. 1995; Buist et al. 1997, 1998). However, lactococci, NSLAB and brevibacteria become nonculturable in carbohydrate-depleted media while remaining metabolically active (Ganesan et al. 2004, 2007) and maintain substrate transport for producing energy from alternate substrates (Ganesan et al. 2007). The declining lactococcal counts in cheese may represent a subpopulation of replicating cells, while a nonculturable population of cells that is unable to divide and is hence not enumerated on growth media (Kilcawley et al. 2011) coexists. Additionally, the enumeration of an entire genus using selective growth media does not accurately reflect the metabolic contributions of multiple species of a diverse genus, such as Lactobacillus (Peterson and Marshall 1990; Trepanier et al. 1991), that shares metabolic capabilities with other Lactobacillales (Makarova et al. 2006). Hence, alternate population assessment methods are needed to better estimate LAB species diversity in cheese and survival of probiotic bacterial species in foods used for delivery. Multiple studies at laboratory- or pilot-scale cheese manufacture have demonstrated probiotic bacterial survival in hard cheeses such as Cheddar, a representative set of which are listed in Table 1. According to these studies, even the same strains or species survive variably, with one group showing survival throughout ageing, but another demonstrating loss of viability of the same in 6– 8 weeks. Some studies were conducted in smaller scale (10–20 l of milk), which do not represent issues in commercial-scale manufacture such as adventitious NSLAB and bacteriophages’ presence. Notably, none of these studies enumerated survival of probiotic bacteria at the species level. Neither did these studies assess the impact of adding a microbe not usually found in cheese on the starter or NSLAB of cheeses. In one study that used qPCR for detecting bacteria (Achilleos and Berthier 2013), only soft cheeses were assessed over 24 h and did not address the challenge of detecting bacteria in aged semi-hard cheeses. To meet regulatory requirements, bacteria are isolated from foods (Giraffa and Rossetti 2004; Rossetti and Giraffa 2005) prior to species identification based on 16S or 23S rRNA genes using organism-specific primers (Salama et al. 1991; Klijn et al. 1995), single-gene sequencing, random DNA amplification, use of restriction enzyme digestion patterns (Byun et al. 2004; Rossetti and Giraffa 2005), gel electrophoresis by varying electric fields (pulse field) (Klein et al. 1998; Henri-Dubernet et al. 2008), temperature (Vasquez et al. 2001; Ogier et al. 2002) or chemical gradients (denaturing gradient) (Fasoli et al. 2003); however, these methods do not enumerate bacteria nor conclusively establish species identity. With advances in PCR amplification technology and the use of fluorescent
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1643
1644
NS, not specified.
Sharp et al. (2008) Ong and Shah (2008) Achilleos and Berthier (2013)
Antibiotic resistance mutant
107
Lact. acidophilus and Lact. helveticus Lact. paracasei Not disclosed
20 l
Microbial plating on selective media Microbial plating on selective media and qPCR
NS 108
108
10 l
136 kg
108
450 l
Lact. paracasei, Bif. sp. BB-12 Lactobacillus acidophilus, Bifidobacterium sp., Lact. casei, Lact. paracasei and Lact. rhamnosus Lact. casei
RAPD-PCR/gel electrophoresis on DNA extracted from grown colonies Microbial plating on selective media, staining and confocal microscopy Microbial plating on selective media
108
450 l
Antibiotic resistance mutant
108
Bif. sp.
Antibiotic resistance mutant
4 9 108
450 l
Microbial plating on selective media
5 9 107
Ent. faecium
RAPD-PCR/gel electrophoresis on DNA extracted from grown colonies
NS
Microbial plating on selective media
106
450 l
Bifidobacterium infantis
Daigle et al. (1999) Gardiner et al. (1999a) Gardiner et al. (1999b) Mc Brearty et al. (2001) Auty et al. (2001) Phillips et al. (2006)
25, 450 l (2 reps 9 2 strains only) 250 l
100 kg
Detection method in cheese
Targeted level of probiotics (CFU g1)
Enterococcus faecium
Bifidobacterium bifidum (immobilized; added during salting) Lactobacillus salivarius, Lact. paracasei
Dinakar and Mistry (1994)
Gardiner et al. (1998)
Probiotic bacteria used
Study
Scale of cheese manufacture
Table 1 Previous studies of probiotic addition to Cheddar cheese
Microbial plating on selective media and qPCR
Microbial plating on selective media
Microbial plating on selective media
0
6
108 108
3
107
9
NS
6
9
3 9 108 105 to 108
15
4 9 108
3
8
108
107
6
Cheese ageing period (months)
106 to 107
Microbial plating on selective media, staining and confocal microscopy Microbial plating on selective media 103 to 108
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Microbial plating on selective media
Viability assessment in cheese
Detected levels of probiotics (CFU g1)
200
25
25
25
108 (confocal)
25
25
25
25
25
25
Detection limits of assays (CFU g1)
LAB survival in probiotic-added cheese B. Ganesan et al.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
LAB survival in probiotic-added cheese
labelling, reliable estimates of bacterial populations for a single genus or species are easily obtained using quantitative PCR (qPCR) (Matsuki et al. 2004; Kramer et al. 2009). Recent studies have established the applicability of this technology to other fermented dairy products such as yoghurt (Fasoli et al. 2003; Rademaker et al. 2006; Sieuwerts et al. 2008; Garcia-Cayuela et al. 2009; Sheu et al. 2009), and its applicability to cheese has been demonstrated. Nucleic acid extraction directly from a food matrix was previously limited by high carbohydrate, protein and lipid contents, but methods are now available to extract high-quality DNA for direct use in identification strategies. Challenges in direct nucleic acid extraction from cheese were previously circumvented by prior microbial cell separation from the cheese matrix for nucleic acid extraction (Rudi et al. 2005; Fernandez et al. 2006; Monnet et al. 2006; Fukushima et al. 2007) or using large sample quantities (≥10 g cheese) (Monnet et al. 2006, 2008; Sohier et al. 2012). Else, the methods only provided higher detection limits (≥1000 CFU g1) (Baruzzi et al. 2005; Rantsiou et al. 2008; Sheu et al. 2009) or are incapable of distinguishing added probiotic lactobacilli from NSLAB lactobacilli (Gardiner et al. 1998). Such approaches cannot determine changes of species levels in long-term aged cheeses. However, using genus- and species-specific qPCR, population changes of distinct groups of LAB specifically induced by probiotic addition can be determined. We hypothesized that the addition of specific probiotic bacteria to cheese during manufacture modifies starter and NSLAB lactobacilli survival in Cheddar cheese at different fat levels. In this study, three cheese types that had
different fat contents were made to determine the role of the physico-chemical conditions in such cheeses in allowing survival of probiotic lactobacilli and bifidobacteria by applying a qPCR-based bacterial detection method. We also assessed whether addition of probiotics at levels below that of starter bacteria altered starter or NSLAB levels. Further, the viability of the three groups of bacteria was also determined using propidium iodide-based qPCR assays. DNA extraction from small quantities of cheese (02–1 g) was optimized by coupling physicochemical and enzymatic lysis of bacteria to release nucleic acids, followed by phenol–chloroform-based applications to remove protein and lipid. We found that starter, NSLAB lactobacilli and probiotic bacteria survive throughout Cheddar cheese ageing with limited reduction in viability and that probiotic addition even at levels lower than starter LAB, altered levels of starter lactococci and NSLAB lactobacilli. Materials and methods Probiotic-added cheese manufacture Starter culture, probiotic adjunct, suppliers and the amount of culture used are indicated in Table 2. The targeted range of probiotic bacteria in the finished cheese was 36 9 106–107 CFU g1 (based on the requirement of ≥108 CFU per 28 g serving), and direct-vat-set cultures were added proportionally based on manufacturer’s specifications and expected cheese yields to achieve the target levels. The control cheese was made with Lc. lactis DVS850 starter culture only, while the probiotic cheeses each contained DVS850 and one probiotic culture.
Table 2 Starter culture, probiotic adjunct, suppliers and the amount of freeze-dried culture used Amount used in cheese (g)*
Supplier
Organism
Name
Full fat (160 kg cheese milk)
Reduced fat (135 kg cheese milk)
Low fat (135 kg cheese milk)
Chr. Hansen, Milwaukee, WI, USA Cargill Inc., Waukeshaw, WI, USA Chr. Hansen Chr. Hansen DSM Food Specialties, Logan, UT, USA DSM Food Specialties Chr. Hansen Chr. Hansen
Lactococcus lactis
DVS850
170
300
260
Bifidobacterium lactis
Bif-6
140
120
120
Bif. lactis Lactobacillus acidophilus Lact. acidophilus
BB-12 LA-5 L10
140 140 78
120 120 67
120 120 67
Lactobacillus casei Lact. casei Lactobacillus paracasei subsp. paracasei
L26 CRL-431 F19
64 140 350
55 120 300
55 120 300
*Inoculation levels were adjusted by milk quantity and manufacturer-provided information about cultures on numbers of viable bacteria in freeze-dried cultures to achieve the desired target ranges of 107 to 108 CFU g1 probiotic bacteria in cheese.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1645
LAB survival in probiotic-added cheese
B. Ganesan et al.
Manufacturing procedures for full-fat, reduced-fat and low-fat Cheddar cheeses are described in the supplementary materials and methods section, along with proximate analysis procedures for cheese.
proteinase K at 50°C for 1 h). This was followed by phenol–chloroform extraction of DNA and qPCR for enumerating lactococci using genus-specific primers. This lysis approach was also applied to cheeses tested for viable bacterial levels (see below).
Sampling After ageing for 5 days, the cheese was cut into 10 blocks of c. 1 kg each, vacuum packaged and stored at 3°C. For each replicate treatment, one block was randomly chosen for analysis at 5 days, 1, 2, 3, 4, 6 and 9 month of age.
DNA quality and yield DNA quality and yield were checked on a Nanodrop spectrophotometer (Agilent Technologies, Santa Clara, CA, USA). All samples used in qPCR had a A260/A280 ratio ≥18.
Extraction of genomic DNA from cheese To extract DNA from all three types of cheese, finely grated cheese (025 g) was suspended in TrizolLS (500 ll; Invitrogen, Carlsbad, CA, USA) along with glass beads (03 g, size 01 mm; sterile, acid-washed; BioSpec Products Inc., Bartlesville, OK, USA). The cheese-TrizolLS suspension was disrupted on a Mini-Beadbeater (BioSpec Products) for 30 s to lyse bacteria and further shaken with chloroform (200 ll) for 30 s to allow phase separation. The top aqueous phase was removed after centrifugation (12 000 g for 15 min at room temperature), and DNA was precipitated from the lower organic phase with 100% ethanol for 3 min. DNA was collected by centrifugation (2000 g for 5 min at 4°C), washed twice with 01 mol l1 sodium citrate in 10% ethanol (30 min at room temperature), once with 75% ethanol (15 min at room temperature), pelleted, air-dried and resuspended in sterile double-deionized water (ddH2O) This DNA was used to measure population changes of lactobacilli and bifidobacteria. Due to poor efficacy of glass bead lysis for lactococcal detection (Fig. S2), lactococcal lysis was separately accomplished by an enzymatic approach in which finely grated cheese (025 g) was treated with bacterial lytic enzymes (initially 50 mg ml1 lysozyme + 10 U ml1 mutanolysin at 37°C for 1 h, followed by 9 U ml1
Quantitative PCR Bacterial levels in cheese were determined by qPCR in triplicate reactions. Compliance to MIQE guidelines was ensured throughout the process of qPCR analyses, data extraction and data analysis (Table S5). All bacterial primers were selected from previous studies that designed genus- and species-specific primers for the starter and probiotic organisms (Table 3) and were subsequently validated by assays against the strains used in this study prior to use (see Fig. S1). Briefly, the total DNA extracted from cheese was used in a 25-ll reaction that included qPCR master Mix (HotStart-ITâ SYBRâ Green; USB Corp., Cleveland, OH, USA) and genus- and species-specific primers targeting the 16s rRNA gene (Table 3). Each final reaction mixture contained: template DNA, 500 ng, MgCl2, 25 mmol l1, primers, 10 pmol and each dNTP, 04 mmol l1. The qPCR was performed on a DNA Engine OPTICON2 (Bio-Rad Labs Inc., Hercules, CA, USA) with initial enzyme activation at 95°C for 5 min, followed by 40 cycles of: denaturation at 95°C for 15 s, annealing at 50°C for 30 s and extension at 72°C for 1 min. All DNA samples from probiotic-added cheeses were always assayed simultaneously with DNA samples from control cheeses (no probiotic added) using speciesspecific primers to confirm lack of amplification. Following
Table 3 Sequences of primers used for genus- or species-specific 16s ribosomal gene qPCR for different cheese bacteria Organism
Specificity
Primer sequence
Cheeses applied to
References
Lactococcus lactis
Genus
All
Klijn et al. (1995)
Bifidobacterium lactis
Species
Bif-6, BB-12
Ventura et al. (2001)
Lactobacillus
Genus
Control, LA-5, L10, L26, F19
Byun et al. (2004)
Lactobacillus acidophilus
Species
LA-5, L10
Delroisse et al. (2008)
Lactobacillus casei/paracasei
Species
F: 50 - GCGGCGTGGCTAATACATGC-30 R: 50 - CTGCTGCGTCCCGTAGGAGT-30 F: 50 -GTGGAGACACGTTTCCC-30 R: 50 -CACACACACAATCAATAC-30 F: 50 -TGGAAACAGRTGCTAATACCG-30 R: 50 -GTCCATTGTGGAAGATTCCC-30 F: 50 -GAGGCAGCAGTAGGGAATCTTC-30 R: 50 -GGCCAGTTACTACCTCTATCCTTCTTC-30 F: 50 -GCACCGAGATTCAACATGG-30 R: 50 -GGTTCTTGGATYTATGCGGTATTAG-30
L26, F19
Byun et al. (2004)
1646
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
MIQE-compliant procedures, results from all qPCR assays were extracted using OPTICON monitor software v2.2 (Bio-Rad). Viable bacterial qPCR assay using propidium monoazide The viability of bacteria present in cheese was verified by a qPCR-based assay, using the DNA-chelating reagent propidium monoazide (PMA; Biotium Inc., Hayward, CA, USA). Briefly, the assay was conducted as follows: a small portion of cheese (025 g) was thoroughly minced, suspended in an equal amount (w/v) of sterile ddH2O, and a stock of 20 mmol l1 PMA was added to the cheese suspension to obtain a final concentration of 50 lmol l1, as described by Nocker et al. (2007). The addition and mixing was carried out in light-transparent tubes that were then kept in dark for 10 min to allow PMA-nucleic acid binding, followed by exposure to a 650 W halogen lamp for 5 min, with samples kept on ice, to quench the reaction. After light exposure, samples were treated with lysozyme–mutanolysin and DNA was extracted for qPCR-based enumeration. For live-dead assays by qPCR using PMA, the difference in Cq times with and without PMA addition, denoted as DCq, was monitored over cheese age. Statistical analysis All cheeses were made in two replicates that were manufactured from different batches of milk. Cheese samples were collected for qPCR enumeration at 5 days, 1, 2, 3, 4, 6 and 9 months. Each sample was analysed by qPCR in triplicate, and the results of bacterial counts from qPCR assays were analysed as a nested factorial repeatedmeasures design (Ganesan et al. 2007) to understand the impact of cheese type on survival of different added probiotics and the impact of cheese type and strain on survival of NSLAB. JMP v7 statistical software (The SAS Institute, Cary, NC, USA) was used for repeated-measures statistical analyses. For the time series analyses, significance was assigned at a = 005/n, where n is the number of time points. Two-sample statistical comparisons within a particular condition when necessary were performed by two-tailed Student’s t-tests with Microsoft Excel software (Redmond, CA, USA), and significance was assigned at a = 005. Results
LAB survival in probiotic-added cheese
bacterial species were added to the different cheeses. Changes in specific LAB populations over ageing were assessed by qPCR using a species primer set that corresponded to the species added to the cheeses (validation and variability results in supplementary material). We developed a sensitive qPCR assay based on previously validated primers that provided reliable detection even at 10–500 CFU g1 from Cheddar cheese (Table S2). The targeted levels of probiotics in cheese were at least 10 000-fold higher than our detection limits. Primarily, probiotic bacteria were added targeting a final level of 36 9 106–107 CFU g1 cheese, which was accomplished in all full- and low-fat cheeses (Fig. S3). Interestingly, added probiotic lactobacilli populations were consistently lower than the target level by 10-fold initially in reduced-fat Cheddar cheeses, but rose to attain levels of c. 107–108 CFU g1 cheese within a month of ageing (Fig. S2), indicating that they grew in the cheese by 10- to 100-fold. However, at any time, probiotic lactobacilli levels were only 1–10% of that of total lactobacilli (Figs S2 and S4), confirming that added probiotic and NSLAB lactobacilli can be distinctly enumerated using qPCR. Probiotic Lact. casei/paracasei survived through 270 days of ageing in all Cheddar cheeses either at comparable levels or grew by 10- to 70-fold (Table 4), with differences in final levels attained within each fat level (P ≤ 005). Added Lact. acidophilus exhibited a 16- to 37fold reduction in population (P ≤ 005) from the initial level of addition during ageing in low-fat cheese (Table 4). However, Lact. acidophilus strains did not differ in survival patterns over time across the different fat
Table 4 Fold change of different probiotic bacterial populations in cheeses over 270 days of ageing Fold change in populations (CFU g1 ratio of 270 days/0 days)* Probiotic organism Lactobacillus acidophilus LA-5 Lact. acidophilus L10 Lactobacillus casei CRL-431 Lact. casei L26 Lactobacillus paracasei F19 Bifidobacterium lactis Bif-6 Bif. lactis BB-12
Full fat
Reduced fat
Low fat
19 23
13 09
16 04
23 10 41 20
230 09 93 09
37 14 71 06
16 04 15 05
74 02 22 10
35 03 52 08
12 10
28 11
6900 10
1500 10
27 10
3200 100
LAB population changes during ageing Three types of Cheddar cheeses—full, reduced and low fat—were manufactured (Table S1), and seven probiotic
Negative sign indicates reduction. *Values are indicated followed by standard deviations after the ‘’ sign.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1647
LAB survival in probiotic-added cheese
B. Ganesan et al.
levels (P > 005). In contrast, Bif. lactis strains exhibited varying survival patterns (P ≤ 005) within each fat level, with the Bif-6 strain populations declining by 6900-fold in low-fat cheese within 270 days, whereas the BB-12 strain declined by 3200-fold in full-fat cheese (Table 4). Bif-6 remained at similar levels to initial inoculum in reduced-fat cheese (P > 005), whereas BB-12 increased by 1500-fold (P ≤ 005) (Table 4). Notably, estimates of bifidobacterial levels varied by 100- to 1000-fold within replicates for 43% of cheeses across age and fat level, which precluded any strong conclusions about their population variations. Such variation was much less prevalent in probiotic lactobacilli (9% cheeses with 100- to 1000fold variation). However, bifidobacteria were exclusively detected by qPCR throughout cheese ageing, which confirms their viability in cheese. To understand the role of cheese fat content on microbial survival, we directly compared the probiotic levels in reduced-fat and low-fat cheeses. Interestingly, the lower fat level appeared to reduce bacterial survival, as highlighted by significant decrease (P ≤ 005) in numbers of Lact. acidophilus over ageing in low-fat cheese. However, Lact. casei/paracasei grew in both reduced- and low-fat cheeses. These results indicated that fat in cheese may aid survival of selected species of lactobacilli during ageing. As probiotic bacteria were added at levels of 107 CFU 1 g and survived variably, they may alter the levels of starter LAB and NSLAB in ageing Cheddar cheese. We assessed the effect of probiotic addition by measuring changes in overall lactobacilli populations using a primer set targeting the Lactobacillus genus. Notably, in all cheeses tested, levels of total lactobacilli were 10- to 100fold higher than levels of added probiotic, and the patterns of survival of these two groups were different (P ≤ 005) (Figs S2 and S4). However, survival and growth of total lactobacilli in the probiotic lactobacilliadded cheeses differed due to the different strains added and fat level (P ≤ 005). While the overall populations of total lactobacilli remained at 108 CFU g1 or higher in all cheeses, the addition of Lact. acidophilus La-5 and Lact. casei CRL431 significantly (P ≤ 005) altered the levels of total lactobacilli, dependent on cheese fat level. For example, at 270 days of ageing, La-5 addition reduced total lactobacilli by eightfold in the reduced-fat cheese, but there was no change in the low-fat cheese (P ≤ 005; Table 5). Addition of CRL431 decreased total lactobacilli by sevenfold in reduced-fat Cheddar cheese and increased total lactobacilli by sevenfold in low-fat cheese (P ≤ 005; Table 5). Addition of other probiotic organisms did not alter total lactobacilli levels (P > 005), but cheese fat level altered (P ≤ 005) the total lactobacilli levels. One such example is the cheeses to which Lact. paracasei F19 was added, in which total lactobacilli levels 1648
Table 5 Fold change of total lactobacilli populations in cheeses added with different probiotic lactobacilli over 270 days of ageing Fold change in populations (CFU g1 ratio of 270 days/0 days)*
Probiotic organism Lactobacillus acidophilus LA-5 Lact. acidophilus L10 Lactobacillus casei CRL-431 Lact. casei L26 Lactobacillus paracasei F19
Full fat (n = 2)
Reduced fat (n = 2)
Low fat (n = 2)
16 27
81 13
14 04
35 07 16 08
83 10 69 07
60 16 69 07
58 05 16 04
17 02 56 07
15 04 342 13
Negative sign indicates reduction. *Values are indicated followed by standard deviations after the ‘’ sign.
had increased 34-fold in 270-day-old low-fat cheese (P ≤ 005), but had decreased sixfold in reduced-fat cheese (Table 5). These results indicate that fat reduction in cheeses and probiotic addition alters survival of NSLAB lactobacilli over cheese ageing. As NSLAB levels were altered by addition of probiotic lactobacilli and bifidobacteria, we also measured changes in lactococcal populations using PCR targeting Lc. lactis species in all three cheese types in a limited set of samples. Lactococcal populations in the control (no probiotic added) cheeses were compared to levels in Lact. acidophilus L-10-added and Bif. lactis Bif-6-added cheeses. Lactococcal counts in control cheeses ranged between 75 9 107 and 8 9 108 CFU g1 and did not change over time. Addition of different probiotics significantly (P ≤ 005) altered lactococcal populations and also the effects varying within each fat level. For example, addition of Bif-6 did not change lactococcal levels within 270 days in low-fat cheeses, but decreased lactococcal populations in reduced-fat cheeses by 10-fold compared to without Bif-6 addition (P ≤ 005; Fig. 1), whereas L-10 addition consistently allowed lactococci to survive at 10to 100-fold higher levels than without probiotic addition in low-fat cheeses (P ≤ 005), but did not alter lactococcal levels in reduced- and full-fat cheeses. Viability of LAB in cheese LAB that do not survive the harsh cheese conditions may die, and their cell walls and membranes may rupture, releasing DNA and intracellular contents into the cheese matrix where it may remain stable and be detected using PCR, which could artificially elevate the bacterial estimates during qPCR assays. Such extracellular DNA can be removed by treating samples with DNase enzyme prior
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
LAB survival in probiotic-added cheese
10
CFU gm–1
9 8 7 6 5
0
90
120
270 Days
10
CFU gm–1
9 8 7 6 5
0
90
120
270 Days
10
detection using species-specific primers, did not differ (P > 005) in qPCR Cq times from nontreated cheeses (Fig. 2), indicating that extracellular DNA was not available for PCR detection and did not contribute to the total count using qPCR. This indicated that qPCR assays are exclusively based on DNA from live bacteria (including bacteria that are nonculturable) and that DNA from dead bacteria is not being enumerated. Secondarily, bacterial cells may exist in a physiologically dead state wherein they are partially autolysed or their cell walls or membranes are partly compromised to create a leaky membrane, but still contain intact DNA that could be co-extracted along with live cells’ DNA and cause overestimation. To assess this potential interference, we conducted qPCR assays on all three cheese types at 5 days, 4, 6 and 9 months age, to which PMA, a DNAchelating dye, was added prior to DNA extraction. PMA selectively permeates across and binds DNA in dead/compromised cells and thus limits PCR amplification from the chelated DNA, which consequently will increase the DCq over ageing if cell membrane is compromised. The procedure of PMA treatment for cheese was validated by adding 107–108 logarithmic growth phase cells to a cheese sample obtained from another source prior to PMA treatment, and DNA extraction was followed by qPCR enumeration. With and without PMA treatment, the DCq between samples was similar (P > 005), which proved that the added cells were live (data not shown). We conducted qPCR assays targeting the starter lactococci in cheese without added probiotics and two cheeses with
9 CFU gm–1
25 8
20 7
15
5
Cq
6
10 0
90
120
270 Days
Figure 1 Lactococcal starter bacteria population levels in Cheddar cheeses estimated by qPCR. Vertical error bars represent standard deviations of log10 (CFU g1) assessed from two replicates of cheese. Top panel, full-fat cheese, middle panel, reduced-fat cheese, and bottom panel, low-fat cheese. (●) Control—probiotic not added; (■) probiotic Bifidobacterium lactis Bif-6 added; (♦) probiotic Lactobacillus acidophilus L10 added.
to cellular lysis to degrade prereleased DNA. However, DNase-treated 3-month-old cheeses containing added Lact. casei L26 and Bif. lactis Bif-6, when tested by qPCR
5
0 Genus Species Lactobacillus primer type Figure 2 Quantitative PCR Cq times of Lactobacillus genus and species determined from cheeses treated with DNase enzyme prior to bacterial lysis to remove extracellular DNA. Cheese age was 3 months and contained added probiotic Lactobacillus casei L26. White bar, no DNase treatment, grey bar, DNase treated prior to lysis and DNA extraction.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
1649
LAB survival in probiotic-added cheese
B. Ganesan et al.
different probiotic species (Lact. acidophilus La-5 and Bif. lactis Bif-6) in full-, reduced- and low-fat cheeses over 9 months of cheese ageing. In almost all cases, the DCq did not change significantly (P > 005) over cheese age for lactococci (Fig. 3a), suggesting that over ageing up to 9 months, starter lactococci survive the ageing process without loss in viability no change in the membrane permeability, despite declines in plate counts as demonstrated by other studies (Crow et al. 1995; Buist et al. 1998; Oberg et al. 2011). The DCq for lactococci did, however, increase slightly in the low-fat cheese with added bifidobacteria between 0 and 120 days (Fig. 3a) and remained at similar levels up to 270 days, suggesting that the membrane permeation increase was insignificant over long-term ageing of cheese. Lactococci survived even after probiotic addition without loss in viability, suggesting that the impact of added probiotics is likely minimal. This approach also verified that LAB levels assessed by qPCR identified viable and intact cells that were present in the cheese matrix and not dead or compromised cells. Assessing probiotic survival using the PMA-added cheese samples showed that added lactobacilli also survived in Cheddar cheese over age, demonstrated by either a significant reduction in DCq (P ≤ 005, full-fat cheese,
14
Fig. 3b) or modest increases in DCq (P ≤ 005, reducedand low-fat cheeses, Fig. 3b). Similarly, small but significant changes in DCq (P ≤ 005) were also observed with bifidobacteria-added cheese (Fig. 3c), showing that bifidobacteria also survived, with some loss of viability on one strain, which matches assessment by standard qPCR that showed some loss of bifidobacterial viability in 270day-old Cheddar cheeses (Table 4 and Fig. S3). Additional bifidobacterial species need to be assessed for their viability in cheese. However, application of the PMA assay confirmed that qPCR detected viable LAB and bifidobacteria in the cheese matrix and that starter lactococci, NSLAB and probiotic lactobacilli, and bifidobacteria remained viable throughout 270 days of Cheddar cheese ageing. Discussion To consistently produce commercial cheeses, the manufacturing process of aged semi-hard and hard cheeses is designed to physico-chemically control their microbial diversity. For example, Cheddar cheese is initially inoculated with high numbers of starter lactococci (105– 106 CFU ml1 milk) that replicate rapidly during cheese
(a)
ΔCq (with - without PMA)
12 10 8 6 4 2
0
9 (b) 8 7 6 5 4 3 2 1 0
50
0
180 Time (days)
1650
100
150 Time (days)
ΔCq (with - without PMA)
ΔCq (with - without PMA)
0
200
18 (c) 16 14 12 10 8 6 4 2 0
250
0
300
180 Time (days)
Figure 3 Comparisons of live lactococcal (a), probiotic Lactobacillus acidophilus (b), and probiotic Bifidobacterium lactis (c) populations by determining qPCR Cq times from cheese samples’ DNA extracted with and without propidium monoazide addition. Data are presented as the difference in Cq times, DCq, as explained in Materials and Methods. Vertical error bars represent standard deviations of DCq assessed from two replicates. In panel a, □ full-fat control (probiotic bacteria not added during cheesemaking), ■, low-fat control (probiotic bacteria not added during cheesemaking), 4, Bif. lactis Bif-6 added to full-fat cheese, ♢ Bif. lactis Bif-6 added to low-fat cheese, ○, Lactobacillus casei L10 added to full-fat cheese, ●, Lact. casei L10 added to low-fat cheese. In panel b, white bars, full-fat cheese, grey bars, reduced-fat cheese, and black bars, low-fat cheese, all containing Lact. acidophilus LA-5 added to cheese. In panel c, white bars, full-fat cheese, grey bars, reduced-fat cheese, and black bars, low-fat cheese, all containing Bif. lactis Bif-6 added to cheese. Statistical significance over time was determined only within each cheese.
Journal of Applied Microbiology 116, 1642--1656 © 2014 The Society for Applied Microbiology
B. Ganesan et al.
manufacture and chiefly degrade lactose to produce lactic acid and reduce cheese pH during early ageing. When adjunct or probiotic bacteria are included, the balance of LAB populations may be altered. Chiefly, the added probiotics must survive in sufficient numbers (target 36 9 106 to 36 9 107 CFU g1 cheese, to deliver 109 bacteria per serving) throughout long-term ageing of cheese to provide desired health benefits upon human consumption (Medici et al. 2004; Ouwehand et al. 2012). Among the added lactobacilli, Lact. casei and Lact. paracasei are commonly found in Cheddar cheese, are even used as adjuncts and are expected to survive throughout cheese ageing (Peterson and Marshall 1990). To detect added probiotic bacteria, we applied primers previously designed in other studies (Table 3) after selecting for specificity against the strains used in our study. Better sensitivity was achieved compared to previous studies by improving DNA extraction from Cheddar cheese using phenol–chloroform-based methods (min. 100 CFU g1 detected; Fig. S1). By applying genus- and species-specific primers, we were able to distinctly enumerate probiotic lactobacilli and NSLAB lactobacilli. The consistent presence of NSLAB lactobacilli at 10- to 100fold higher levels precluded any interference by added probiotic lactobacilli in NSLAB enumeration. The use of qPCR also provided a specific and reliable quantitative assay throughout cheese ageing, as opposed to growth media-based enumeration techniques (McSweeney et al. 1993; Gardiner et al. 1998; Crow et al. 2001; Oberg et al. 2011) where the distinction between starter and NSLAB becomes difficult beyond 90 days of age. At certain cheese ages, bacterial levels estimated by qPCR varied by as much as 10- to 1000-fold, which appears to be common in the cheese matrix based on other studies that used media-based enumeration approaches (Table 1). Total lactobacilli were present at much higher levels of 108 to 109 CFU g1 initially than expected (500 kg milk per batch), equipment is frequently subject to cleaning-in-place, which provides excellent control over unwanted NSLAB presence. However, even at this high initial NSLAB level, we could still distinguish added probiotic from NSLAB, which demonstrated the strength of using qPCR to detect added bacterial species. Added probiotic lactobacilli survived in cheese over 270 days of ageing, even growing 10 to 1000-fold (Table 4, Fig. S2). This observation is not surprising, considering that NSLAB lactobacilli are acquired from different sources and only found at low levels (
