Comparison of Quantitative Real-time PCR (qPCR) and Droplet Digital PCR (ddPCR) for Quantification of Lactobacillus reuteri in Human Feces

Abstract

The human gut microbiota, a complicated microbial community consisting of trillions of microorganisms, has been recognized as of great importance for maintaining human health. To establish direct linkages between a certain group of gut microorganisms and the physiological status of the host, it is critical to quantify their abundance with high sensitivity, accuracy, and reproducibility. Quantitative real-time PCR (qPCR) has been widely used in the absolute quantification of microorganisms by comparing PCR cycle numbers with those of a standard curve. Recently, the development of droplet digital PCR (ddPCR) has demonstrated the potential to handle samples with a complex background without standard curves. The goal of this work was to compare absolute quantification of a specific bacterial strain (Lactobacillus reuteri DSM 17938) in human stool between qPCR and ddPCR with three commonly used DNA extraction methods (QIAamp Fast Stool DNA Kit [QK], phenol chloroform [PC], protocol Q [PQ]). DNA extracted using QK and PQ had acceptable quantity and high quality while PC produced DNA with highest concentration but lower purity. Compared to the other two methods, PQ recovered the most substantial proportion of L. reuteri cells from feces. Generally, reproducibility was better in ddPCR than qPCR with methods QK and PC, but comparable with PQ. Within the detectable range, both qPCR and ddPCR presented better linearity with DNA extracted using methods QK and PQ than PC. For qPCR, the limit of detection (LOD) was 3.95 Log10, 4.86 Log10, 4.11 Log10 CFU/g feces with QK, PC and PQ. ddPCR exhibited a lower limit of quantification (LOQ) when compared to qPCR. The LOQ was 4.30 Log10 CFU/g feces when QK and ddPCR were combined and was slightly higher in use of PQ and qPCR (4.50 Log10 CFU/g feces). However, the cost of ddPCR per unit was 3 times higher than qPCR and is more time-consuming (6.5 h vs. 2.5 h). Therefore, the combination of PQ and qPCR is suggested as the best to detect L. reuteri in fecal samples. The second goal of my project was to use the information above and design strain/lineage-specific qPCR systems for the absolute quantification of L. reuteri PB-W1 and DSM 20016T in fecal samples. To achieve this, strain/lineage-specific primers were designed using a database approach. The LOD for PB-W1 measurement was 3.05 Log10 CFU/g feces, which was the lowest level reached in the field, and for DSM 20016T, it was 2.96 Log10 CFU/g feces, which was equally decent. Overall, this work successfully established approaches for a quantitative, selective, accurate, and sensitive quantification of bacterial strains in human fecal samples based on qPCR and ddPCR and provided information to select the appropriate methods for measurement and DNA extraction.