Introduction
Streptococcus thermophilus is a cornerstone of the global dairy industry, second only to Lactococcus lactis in its importance for producing yogurt and cheese. While industrial production relies on a limited set of well-characterized strains, "wild" strains found in traditional fermented products across regions like Mongolia and China harbor significant genetic diversity. This study provides a comprehensive genomic map of 185 such strains, aiming to bridge the gap between genetic architecture and phenotypic traits like acidification speed.
Genomic Diversity and Phylogenetic Structure
The researchers sequenced 185 S. thermophilus strains (65 from China, 120 from Mongolia) using the Illumina HiSeq 4000 platform. Pan-genome analysis revealed a total of 7,629 genes, with only 827 (approximately 10%) constituting the "core" genome shared by all strains. This highlights the vast diversity within the accessory genome of the species.
Phylogenetic analysis categorized these strains into four distinct clades: A, B, C, and D.
Clades A and B were found to be significantly enriched with "H-Acid" (high acid-producing) strains, identifying them as prime candidates for industrial application. Conversely, Clade D exhibited notable "genome decay," characterized by the loss of genes related to quorum sensing and bacteriocin production. This suggests an evolutionary specialization for the nutrient-rich, stable environment of milk, where complex survival mechanisms required in the wild become redundant.
The "Growth Score": A Novel Predictive Tool
A major challenge in dairy science is the labor-intensive nature of screening strains for fermentation efficiency. To address this, the authors developed a "growth score" derived from static genomic data. This method leverages the biological fact that in a growing bacterial population, DNA near the replication origin is more abundant than DNA near the terminus.
By calculating the correlation between sequencing read depth and the distance from the replication origin, the researchers created a proxy for replication speed. Validation experiments on 85 strains confirmed that this growth score significantly correlates with actual acid production speed (Pearson’s r = 0.4), demonstrating that genomic "snapshots" can predict dynamic fermentation behavior.
Identification of the AcnA Biomarker
Using Genome-Wide Association Studies (GWAS) and stability selection, the study linked specific Single Nucleotide Polymorphisms (SNPs) to fermentation phenotypes. A key finding was a missense mutation (G1118698T) in the AcnA gene, which encodes aconitate hydratase A.
Strains carrying this mutation were consistently slower acidifiers. This makes the AcnA mutation a robust negative-selection biomarker, allowing researchers to quickly filter out low-performing strains during the initial genomic screening phase without the need for time-consuming culture experiments.
Environmental Impact and Antibiotic Resistance
The study also uncovered evidence of human industrial impact on traditional food microbiomes. The chloramphenicol-resistance gene CatB8 was identified specifically in strains isolated from regions with high environmental antibiotic emissions. This finding underscores how anthropogenic activity influences the genetic makeup of bacteria even in traditional, non-industrial fermentation settings.
Conclusion and Future Directions
This research provides a high-throughput framework for identifying high-performing dairy starter cultures. By utilizing the growth score and specific genetic markers like the AcnA mutation, the industry can move toward more efficient, genomics-based strain selection.
Future work aims to adapt the growth score for mixed-culture fermentations (e.g., S. thermophilus grown with Lactobacillus bulgaricus). While current read-mapping ambiguities in metagenomic samples present a challenge, focusing on species-unique sequences could enable real-time monitoring of complex industrial fermentations.
Summary updated