Rapid Determination of the Bacterial Composition of Commercial Probiotic Products by Terminal Restriction Fragment Length Polymorphism Analysis

Abstract

Label claims on probiotic products often do not represent the true constituents. With the increased use of probiotics in clinical studies, it is necessary to know the true composition of probiotic products to better interpret study outcomes. We used terminal restriction fragment length polymorphism analysis to rapidly determine the overall bacterial composition of 14 commercial probiotic products and validated the results with species-specific polymerase chain reaction. The results show that many probiotic products contain unadvertised additional lactobacilli and bifidobacteria, whereas others are missing species listed on the product label. In summary, terminal restriction fragment length polymorphism is a rapid method for profiling the microbial contents of probiotic products used in clinical studies.

Probiotics are marketed in the United States as dietary supplements containing viable beneficial microorganisms. There is an increasing interest in the use of probiotics to prevent allergies ⁽¹⁾, necrotising enterocolitis ⁽²⁾, diarrhea ⁽³⁾, and inflammatory bowel disease ⁽⁴⁾.

Unfortunately, a high percentage of commercial probiotics are erroneously labeled with respect to the actual bacterial contents. Both species-specific polymerase chain reaction (PCR) and community profiling techniques such as PCR-denaturing gradient gel electrophoresis (DGGE) have been used to identify the bacterial species present in commercial probiotics ^(5,6). Terminal restriction fragment length polymorphism (T-RFLP) is a culture-independent technique commonly used in the analysis of complex microbial communities ⁽⁷⁾.

T-RFLP uses a fluorescence-labeled primer in a PCR to amplify common rRNA gene sequences from DNA purified from a mixed culture. The resulting mixed amplicon is then digested with a restriction enzyme to create a mixture of fluorescence-tagged terminal restriction fragments (T-RFs) derived from the different microbes within the mixed culture. The exact size of each T-RF is then determined by standard capillary electrophoresis. The fragment sizes are then compared with existing rRNA sequence databases to predict the microbial contents of the original sample. By comparison to PCR-DGGE, T-RFLP is a more accessible method given that the DNA fragment separation is readily performed at service laboratories.

T-RFLP is more frequently used to characterize highly complex microbial communities in diverse environments, such as fecal or soil samples ⁽⁷⁾. Given the increased used of commercial probiotics in clinical trials, we sought to evaluate T-RFLP as a simple way to reveal the true bacterial contents of commercial probiotic products before their use in clinical studies. Our results indicate that T-RFLP is a readily accessible method to verify the bacterial species present in commercial probiotics and confirm that label claims often differ from the actual species present.

MATERIALS AND METHODS

Probiotic Products

A total of 14 commercial probiotic products were analyzed. These products included lyophilized preparations in tablet and capsule form, purchased from local vendors. All of the probiotic products were tested in duplicate.

T-RFLP Analysis

Pills were dissolved in 2 mL of trace element TE buffer (10 mol/L Tris, pH 7.5, 1 mol/L EDTA), and a DNeasy Blood and Tissue kit (Qiagen, Valencia, CA) was used for DNA isolation. The 16S rRNA genes were amplified by using the eubacterial universal primers 27F and 1492R ⁽⁸⁾. The forward primer 27F was labeled at the 5′ end with the fluorescent dye 6-FAM (6-carboxyfluorescein, Applied Biosystems, Foster City, CA). The PCRs were performed in a volume of 50 μL. Reaction temperatures and cycling were 95°C for 2 minutes, then 30 cycles of 95°C for 1 minute, 50°C for 1 minute, and 72°C for 2 minutes, and finally 72°C for 7 minutes. For each sample of DNA, 2 different PCRs were run and pooled together to minimize PCR biases. PCR products were purified using the Qiakit Purification Kit (Qiagen). About 30 μL of purified PCR products were digested separately with AluI and HaeIII. Digested PCR products were analyzed in Genescan mode on ABI Prism 3100 Capillary Electrophoresis Genetic Analyzer (Applied Biosystems). GS-500 ROX (Applied Biosystems) was the internal molecular weight standard. T-RFs with a peak height <25 fluorescence units were excluded from the analysis. The reproducibility of the method was confirmed by repeating each digestion twice, with only minor variation in peak heights. The identification of the T-RFs was done using the T-RFLP Analysis Program database ⁽⁹⁾. As described by other authors ⁽¹⁰⁾, we noted some minor differences between the predicted and observed T-RF lengths.

Species-specific PCRs

Species-specific PCR primers were used to detect the following microbial species: Bifidobacterium bifidum⁽¹¹⁾, B infantis, B longum⁽¹²⁾, Lactobacillus acidophilus⁽¹³⁾, Lb helveticus⁽¹⁴⁾, Lb rhamnosus, Lb casei/paracasei, Lb salivarius⁽¹⁵⁾, Lb plantarum⁽¹⁶⁾, Streptococcus thermophilus⁽¹⁷⁾, and Lactococcus lactis subsp. lactis ⁽¹⁸⁾. Probiotics were tested by species-specific PCR to validate the label claims. The PCR reaction mixture for each pair of oligonucleotides was 50 μL. The PCR conditions were 95°C for 2 minutes, followed by 30 cycles consisting of 95°C for 1 minute, 50°C for 1 minute, and 72°C for 1 minute, and a final extension period of 72°C for 7 minutes.

T-RFLP Detection Limit

Ten-fold serial dilutions of B infantis UCD272 (10⁸ to 10² colony-forming units [CFU]/mL) were prepared and the cells added to a probiotic product previously determined not to contain B infantis. DNA was extracted from each sample by using a DNeasy Blood and Tissue kit. Each DNA sample was then evaluated by T-RFLP as described above.

RESULTS AND DISCUSSION

Detection Limits of T-RFLP

To test the lower limit of detection within the probiotic products, we added different levels of B infantis into an existing probiotic product and performed T-RFLP to identify the resulting B infantis fragments. The lowest detection limit at which B infantis was detected was 10³ CFU/mL. This detection limit is lower than that determined using PCR-DGGE ⁽¹⁹⁾.

T-RFLP Analysis of Commercial Probiotic Pills

The results of the T-RFLP analysis of 14 probiotic products are presented in Table 1. A sample T-RFLP profile is shown for product 2 in Fig. 1. Surprisingly, only 1 product (no. 9) contained the exact species stated on the label (B longum and B bifidum). Most of the probiotic cocktails—such as products 2, 4, 8, 10, 11, 13, and 14—contained all of the microorganisms that were specified on the label, but had additional microbial constituents as well. Analysis of products 1, 3, 5, 6, and 7 indicated that each was missing 1 species claimed on the label. Species-specific PCR revealed the missing bacterial constituent, suggesting that the particular microorganism in question was present, but at a level below 10³ cells/mL, which is the lower T-RFLP threshold.

TABLE 1: Overview of terminal restriction fragment length polymorphism (T-RFLP) and species-specific polymerase chain reaction (PCR) analysis of 14 probiotic products

FIG. 1: Terminal restriction fragment (T-RF) length polymorphism (LP) profiles of 16S rRNA genes from strains present in probiotic product 2. The 16S rRNA genes were amplified using the universal primers 27F and 1492R. Two different restriction enzymes (AluI and HaeIII) were used for the analysis. T-RF sizes in base pairs are shown in parentheses. T-RF pattern data from the 2 enzyme digestions were matched to the T-RFLP Analysis Program database, allowing +3 bp of the predicted T-RF sizes. Confirmation of species identification required the presence of definitive peaks in both electropherograms. T-RFs that could not be assigned to any bacterial species are indicated by asterisks. These unidentified fragments likely derive from species not included in the database, or from pseudo T-RF generated in the polymerase chain reaction as previously discussed by Matsumoto et al ⁽²⁰⁾.

Others have reported that the contents of different commercial probiotic products do not always correspond to the label claims ^(6,21). Given the increased use of probiotics in clinical trials, and the fact that the microbial content in various probiotic cocktails may influence study outcomes, clear documentation of the bacterial species used in such studies is critical. The aim of this work was to evaluate T-RFLP as a tool to rapidly determine the composition of commercial probiotic products. T-RFLP has significant advantages over PCR-DGGE, the other commonly used profiling technique. T-RFLP is far easier to perform, has higher throughput, greater resolution, and provides results in the form of a digital output that allows for more rapid data processing ⁽⁹⁾.

T-RFLP clearly has some limitations. For example, the protocol used in this study does not distinguish among Lb plantarum, Lb alimentarius, and Lb casei. Similarly, Lb johnsonii and Lb gasseri match the same profile in the T-RFLP Analysis Program database, as do Lb rhamnosus and Lb paracasei. Another limitation is that, like many DNA-based community profiling methods, T-RFLP analysis does not give information regarding the viability or relative numbers of the probiotics. In spite of these limitations, we have demonstrated that T-RFLP is a useful approach to rapidly identify the microbial composition of probiotic products used in clinical trials, and that the majority of probiotic labels do not accurately reflect product composition.