Specific TaqMan Probes for the Identification and Quantification of Lactobacilli in Pharmaceuticals

Within the last decades species of the genus Lactobacillus have been widely commercially used as they are believed to possess probiotic features and thereof resulting beneficial health effects. These strains are utilized in manufacturing fermented food from milk such as yoghurt and cheese [1]. As an example L. reuteri is producing Reuterin (3-hydroxypropionaldehyde [3-HPA]), which is water soluble, effective in a wide pH range and resistant to proteolytic and lipolytic enzymes [2,3]. Therefore, L. reuteri is used in therapeutic treatment being bioactive against bacteria, viruses and fungi [2-4].


Introduction
Within the last decades species of the genus Lactobacillus have been widely commercially used as they are believed to possess probiotic features and thereof resulting beneficial health effects. These strains are utilized in manufacturing fermented food from milk such as yoghurt and cheese [1]. As an example L. reuteri is producing Reuterin (3-hydroxypropionaldehyde ), which is water soluble, effective in a wide pH range and resistant to proteolytic and lipolytic enzymes [2,3]. Therefore, L. reuteri is used in therapeutic treatment being bioactive against bacteria, viruses and fungi [2][3][4].
Conclusively, these probiotic bacteria have also been used in a wide range of pharmaceuticals such as tablets, drops and granulate [1,5]. Other vehicles used are lozenges, powder, gelatin or straws [1]. As an example Klayraung et al. (2009) found that probiotic bacteria administered in liquid or semisolid formulations showed low cell viability after oral administration caused by harsh conditions in the stomach [6]. Thus, development of dry dosage forms, matrix forming excipients and attuned compression forces used for tablet preparation increases bacterial survival rates in low pH conditions [6].
The viability of the strains has to be assured for having beneficial effects on the host according the FAO/WHO guidelines ("live microorganisms that, when administered in adequate amounts, confer a health benefit on the host") [7,8]. However, these guidelines focused food; concluding biotherapeutic agents [9]. Nevertheless, the basic principles of this definition also apply for probiotic pharmaceuticals [10].
The adequate number of probiotic bacteria having health beneficial effects has to be accurately assessed, when used in food or pharmaceuticals [11]. As an example, the amount of probiotic bacterial cells needed to induce immune defensive benefits, reduction of cholesterol levels and preventing diarrhea and food allergies ranges between 10 6 and 10 8 cfu/ml [7,12,13].
Since the usage of lactobacilli is increasing steadily in medical treatment, two examples of commercially available probiotic species were included in the TaqMan ® labeled real-time PCR assay: Reuflor ® chewable tablets (Italchimici, Pomezia, Italy) containing approximately 10 8 cfu/tablet of viable L. reuteri DSM 17938 [14] and Milchsäure-Kulturen Bifido-Flor Kapseln ® (dm-drogeriemarkt, Karlsruhe, Germany) containing Bifidobacterium animalis subsp. lactis BB-12 ® and L. acidophilus -both in an amount of 2.0x10 6 cfu/capsule. Therefore, a rapid identification tool would be useful to identify and quantify probiotic bacteria in different products. As conventional PCR or MALDI-TOF MS lack the capacity to quantify bacterial loads in products, real-time PCR based approaches such as TaqMan ® probes are necessary.

Lactobacilli strains and isolates
In total, 77 different strains of the genera Lactobacillus (L.) and Bifidobacterium (B.) were used as positive and negative controls to confirm primer specificity (primer pairs: Table 2). In addition, species of the genera Lactococcus and Streptococcus were also chosen for comparison as they are often used as probiotics. A complete list of all tested species is published in Herbel et al. (2013) [15].

Confirmation of species identification
Reference strains and strains isolated from Reuflor ® tablets and Milchsäure-Kulturen Bifido-Flor Kapseln ® are listed in Table 2. In parallel to real-time PCR assay classical microbiological methods had been used to assure species identity [18]. In addition, species identification was confirmed by MALDI-TOF MS measurements carried out using a Microflex LT instrument (Bruker Daltonics, Bremen, Germany). The real-time classification tool of the Biotyper 3.0 software tool (Bruker Daltonics, Bremen, Germany) was utilized for culture-independent species identification [19]. Additionally, all amplicons processed by real-time assay utilizing universal primer pairs were fully sequenced (LGC Genomics, Berlin, Germany) and compared to public available sequence databasis (Supplement 1).

DNA extraction from pure lactobacilli cultures and tablets
DNA extraction from pure cultures was performed as described by Herbel et al. (2013) [15,20]. The protocol for tablets followed the one given for yoghurt. In brief, tablets were dissolved in 500 µl 0.4 M sodium hydroxide (Roth, Karlsruhe, Germany) and 150 µl of 40 % trisodium citrate di-hydrate (Roth, Karlsruhe, Germany). The mixture was carefully shaken, incubated for 15 min at room temperature and centrifuged for 2 min (15.000 x g). Following, supernatant was removed, the pellet was washed and treated with mutanolysin (Roth, Karlsruhe, Germany) as described by Lick et al. (1996) [21]. In addition, sonication was done using UP100H Hielscher Ultrasound Technology (Teltow, Germany) followed by DNA isolation as described in Herbel et al. (2013) [15].

Primer design
The heat shock protein region GroEL (synonymes hsp60, cpn60, groL) is a single copy target gene in the genome of all lactobacilli [22][23][24]. The principal usability of this region as a target for species identification has been already shown in several publications [15,24,25]. Partial sequences of this region were retrieved from the NCBI database. In addition, the heat shock protein gene regions of all reference strains and the lactobacilli isolated from products were sequenced (Table  1). Sequence alignments were performed by Megalign ® alignment suite (Lasergene DNA Star, Madison, Wisconsin, USA) applying the ClustalW algorithm [26]. The alignments were used to design lactobacilli universal primer pairs and to identify species-specific primers within the universal primer pair region. Furthermore, all designed primers were screened for biophysical similarities and dimer formations using BLAST algorithm and Oligoanalyzer 3.0 software (Integrated DNA Technologies, Coralville, Iowa, USA) [27]. Ensuring their specificity for the GroEL region of the mentioned species, all primers were verified using the BLAST algorithm of the NCBI database [27][28][29]. √, species-specific amplification was possible using this primer in combination with universal primer set g n.t., not tested The predicted amplicon size for the universal primer pair was 124 bp and each primer set was evaluated by melting curve analysis on a StepOnePlus TM Real-Time PCR Systems (Applied Biosystems, Darmstadt, Germany) using Power SYBR ® Green (Applied Biosystems, Darmstadt, Germany) (data not shown). The primers listed in Table 2 showed the best specificity of all evaluated oligonucleotides of this study and were therefore used.

Quantitative Real-time PCR
All universal primer pairs and specific TaqMan ® probes developed during this study were tested with DNA samples (20 ng/µl) of the genera Lactobacillus, Bifidobacterium, Lactococcus and Streptococcus, which were used as positive and negative controls.
Each real-time PCR included two technical repeats and the results were analyzed by using the Roche ® Light Cycler 480 ® software.

Determination of the real-time PCR detection limit using pure lactobacilli DNA
To assess the sensitivity of the assay and to compare real-time PCR results to Colony Forming Units (CFU) of bacteria obtained by plating, DNA was isolated from serial dilutions of lactobacilli, which were plated on agar, respectively. Therefore, liquid cultures (three technical repeats) of each strain utilizing 0.9 % sodium chloride solution (Roth, Karlsruhe, Germany) were set to a concentration of 10 8 cfu/ml using 0.5 McFarland Standard in a Sensititre ® Inoculator (Thermo Scientific, Dreieich, Germany) and diluted from 10 8 cfu/ml to 10 3 cfu/ml. 200 µl of each dilution were plated on MRS, LBS, COL and CHOC agar plates (origin as described above) and incubated for 24 h to 48 h (37° C; 5 % CO 2 ). Colonies were counted and used for extrapolating the cfu/ml. One milliliter of each serial dilution of the strains was used for parallel DNA isolation using a protocol published by Herbel et al. (2013) [15,20,22].
The theoretical number of genome equivalents (GE) was calculated based on the DNA isolated from 10 8 cfu/ml L. reuteri (160 ng/μl) and the published genome size of L. reuteri (1,969,869 bp) [30].

Species-specific amplification of DNA isolated from lactobacilli
By using the BLAST algorithm we were able to confirm that the primer sequences solely targeted the genome of the lactobacilli species of interest. No other genes were found which showed a comparable DNA sequence to the used primers, which could have led to false positives.
First, the universal primer set was evaluated using DNA from reference strains and the strains isolated from tablets. A set of two diverse forward and eight reverse universal primers had been tested. Finally, the universal primer set of UniLactoHsp60FOR1 and UniLactoHsp60REV1 had been chosen, because it was amplifying the highest number of different species belonging to the genus Lactobacillus (L. acidophilus, L. brevis, L. helveticus and L. reuteri) ( Table 2). Following, species-specific TaqMan ® probes targeting regions within the universal primer amplicon region were established. Evaluation of the specific TaqMan ® probes revealed no amplification by using the DNA of the wide range of negative controls. However, the TaqMan ® probes were able to specifically amplify positive control DNA and therefore detect the species L. acidophilus and L. reuteri in a singleplex real-time PCR run. Unfortunately, we were not able to perform a duplex detection assay of both species utilizing the same universal primers and specific TaqMan ® probes.
In parallel, we always confirmed these results with an established SYBR® Green real-time PCR including melting curve analysis. As shown in supplement 2 TaqMan ® and the SYBR ® Green real-time PCR using L. reuteri standard curves and DNA isolated from two batches of Reuflor ® tablets (10 6 or 10 9 cfu/tablet) showed comparable results. Besides melting curve analysis standard curves had been generated for all shown real-time PCRs (Figure 1, Supplement 2). Additionally to assure the real-time PCR results L. reuteri had been isolated from tablet material and was confirmed by MALDI-TOF MS analysis (data not shown). Furthermore, the amplicon of the universal primer pair has been sequenced resulting in a 124 bp long sequence, which confirmed L. reuteri (Supplement 1). In this study, the C t value the C t values of each real-time PCR run in this study were ranging between 10 to 32 threshold cycles and declared as unspecific signals if they started from the 34th cycle onwards. Thus, analysis of C t value allowed a distinct species-specific identification of the tested lactobacilli species. So summing up to test the specificity of the assay, we tested DNA isolated from pure cultures as well as from probiotic Reuflor ® tablets containing L. reuteri. We were able to identify L. reuteri within the DNA obtained from cultures and tablets using our assay in a singleplex real-time TaqMan ® PCR approach (Figure 1).

Detection limit, identification and quantification of Lactobacillus isolates from tablets
Using standard curves of DNA of known concentrations and correlate these results to the colony forming unit counting in the parallel plating experiments we observed a detection limit of 10 4 cfu/ml for our TaqMan ® approach and SYBR ® Green assay (Figure 1).
The calculated correlation coefficient of TaqMan ® real-time PCR of DNA isolated from Reuflor ® tablets (10 9 cfu/tablet) was R 2 = 0,99458 and its equation of the linear regression line is y = -3.6571x + 35.467 which is comparable to the SYBR ® Green real-time PCR (R 2 = 0.99576; y = -4.7429x + 38.993) (Supplement 2). The theoretical number of genome equivalents (GE) was calculated to be 7.23x10 7 GE/µl for 10 8 cfu/ml pure culture and 8.23x10 7 GE/µl for the DNA isolated from Reuflor ® tablets (10 9 cfu/tablet, (batch number: 2TSA122, Figure 1)). We also used this standard curves to estimate the number of bacteria present in a tablet. By correlating of the amplification curves of DNA extracts from Reuflor ® tablets, we quantified the number of L. reuteri at 10 9 cfu/tablet (batch number: 2TSA122, Figure 1). The parallel plating and counting of serial dilutions derived from tablets of the same batch revealed that viable cell number of L. reuteri was 6.34x10 9 cfu/tablet. However, in another run using a different batch of tablets (2TSA156) we detected 10 8 cfu/tablet (batch number: 2TSA156, both data not shown). As Reuflor ® tablets are designated to contain 10 8 cfu/tablet of L. reuteri both batches seem to carry an adequate number of probiotic bacteria. In addition, one package of Reuflor ® tablets (batch number: 1TSA051) was stored at ambient temperature for a duration of two years, which was still within the expiring date of the product, when DNA had been isolated. Using our real-time PCR assay based on DNA isolated from the stored tablets, we quantified the number of bacteria in the stored tablet to 10 6 cfu/tablet only (data not shown).

Discussion
Herein we describe a singleplex TaqMan ® labeled real-time PCR approach for the rapid and culture independent detection and quantification of probiotic lactobacilli directly from tablets. The heat shock protein GroEL has been demonstrated being a suitable target region for DNA based species identification [15,24]. Additionally, we used this region for establishing a TaqMan ® real-time PCR system and it turned out to be highly specific in our assay as well as we obtained no false positive results while evaluating a wide spectrum of reference strains. The species-specificity of the assay was evaluated in parallel based on classic culture methods and culture independent using MALDI-TOF MS.
One of the advantages of this assay is its rapidity, allowing a speciesspecific identification of lactobacilli species within 7 hours without any prior cultivation step. The detection limit measured by SYBR ® Green and TaqMan ® real-time PCR assay was at a level of 10 4 cfu/ml, which is adequate to quantify strains used in tablets, capsules or granulate formulations containing 10 6 to 10 10 cfu per dosage in order to exert probiotic activity [7].
In case of the Reuflor ® tablets the stated number of bacteria in the package insert was 10 8 cfu/tablet, however, by evaluating tablets from two different batches we detected a cell number between 10 8 and 10 9 cfu/tablet actually. However, in case of probiotic strains this does not seem to be problematic as health beneficial effect for the consumer could be expected (10 6 to 10 8 ) [7,12]. The actual number of cells in a batch of tablets might differ due to manufacturing reasons causing a cell loss by freeze-drying or incorporating the cells in tablet matrix. Additionally, DNA extraction method used in our assay might contribute to these findings as the calculated gene equivalents DNA from the tablets (8.23x10 7 GE) was lower than the actual colony forming units we obtained by plating (6.34x10 9 cfu/ml). In contrast to that for pure cultures the numbers fitted well (7.23x10 7 GE vs 1.47x10 8 cfu/ml).
As we detected more or less comparable numbers of cells culture depended (plating) and culture independent (real-time PCR) this point toward a high viability of the lactobacilli in the tablets. In case of high rates of dead bacteria in the tablets the numbers detected in the PCR assay should have been higher than the results from plating and counting colony-forming units.
The evaluation of tablets that have been stored for two years at room temperature led to the detection of 10 6 cfu/tablet in total, however, according to the package insert there should be 10 8 cfu/tablet. There are several possible reasons for this low number of probiotic bacteria such as a loss or varying numbers of L. reuteri cells during lyophilization or the long storage time including DNA extraction within the expiration date of the product [31]. It seems rather likely that this lower number of bacteria was already present when the tablets were manufactured. A loss of viable bacteria due to the long storage time seems to be unlikely as bacterial DNA is relatively stable under dry conditions such as in a tablet and should have been detected by real-time PCR as well. Indeed a detected amount of 10 6 cfu/tablet is of high importance as recent studies have shown that differing doses of probiotic cells have different effects in the host [32,33]. Thus, our TaqMan ® probe based real-time PCR was highly specific and sensitive in the singleplex assay we had difficulties in establishing a multiplex real-time PCR approach. This might have been caused by primer-primer interactions [35].
In conclusion, our TaqMan ® labeled real-time PCR system was useful for the detection of L. acidophilus and L. reuteri with high specificity and sensitivity directly in tablets without prior cultivation. This, in contrast to other methods such as physiological or morphology testing, contributes to the rapidness of the assay by working cultureindependent. Additionally, besides an identification real-time PCR also allows a proper quantification of Lactobacillus sp. directly from tablets within seven hours. Therefore, this TaqMan ® real-time PCR assay could be a useful tool for the detection of Lactobacillus sp. strains in drugs and food for regulation and quality management purposes.