Sample storage and preparation of whole milk samples for determination of microbiota – #JournalClub no.054

Improved assessments of bulk milk micro-biota composition via sample preparation and DNA extraction methods

DNA sequencing is the most common method for studying microbes in humans, animals, food and the environment. The workflow for sample preparation and DNA sequencing consists of numerous steps that are prone to bias and contamination problems. Numerous bacterial species are present in the diverse microbiota of high-quality whole milk. Within the study by Xue et al. presented here, the following bacterial strains were used:

  • Bacillus subtilis
  • Clostridium tyrobutyricum
  • Corynebacterium bovis
  • Enterococcus faecalis
  • Escherichia coli
  • Lactococcus lactis
  • Pseudomonas fluorexcens
  • Staphylococcus aureus
  • Staphylococcus agalactiae

Bakterielle Cell-Mock-Community (BCMC)

Nine strains in BCMC1 were grown to near stationary phase and counted directly under the microscope. The BCMC2 pools just produced were either used immediately for DNA extraction or stored at -80°C. To obtain the bacteria, BCMC1 and BCMC2 were inoculated with ultra-high temperature pasteurised milk and centrifuged at 13,000 for five minutes at 4°C before DNA extraction. They were stored for seven days at -20°C in PBS or PBS containing 25% v/v glycerol.

Methods for cell lysis and DNA purification

It is possible that Escherichia coli and Pseudomonas are more sensitive to DNA shearing and lysis than Gram-positive bacteria. B5 (tapping at 4 m/sec for 10 seconds) and V3 (vortexing at 1800 rpm for 30 seconds) were two of the milder methods tried to break the cells. Three MagMAX DNA purification kits Total, Core and Ultra2 were compared and the Total kit gave the most accurate results.

Impact of the BCMC preparation method on the estimation of community composition

For each BCMC, the 16S rRNA V4 region was sequenced and analysed to determine the proportions that were „observed“. Escherichia and Pseudomonas were significantly lower than expected, and Bacillus and Corynebacterium were significantly higher than expected for BCMC1.

Effects of milk sample volume and cell count on bacterial composition estimates

Larger milk samples resulted in lower alpha diversity and fewer unexpected taxa assignments or bacteria that did not belong to the mock community. Most of these unexpected bacteria were Micrococcus or Tepidimonas.

Effects of the storage method on the detection of sham communities

Regardless of whether the cells were frozen in PBS or in water with 25 % v/v glycerol at -20 °C, the bacterial proportions of BCMC did not change.

Identification and tracking of bacteria

With the help of 16S rRNA gene sequence surveys, it is possible to identify and track bacterial populations. However, bias can occur at almost every stage of the process, from sampling to data analysis and interpretation. The most accurate method for automatic identification of bacteria consists of at least 10 grams of milk, mild cell lysis, proteinase K treatment and magnetic bead-based DNA purification.

Storage conditions of the collected cells and cell lysis method

The mock community remained unchanged despite different freeze-thaw cycles and storage conditions. It can be assumed that other upstream sample preparation steps have a greater influence than this one previously discovered. Although not tested directly, it has been previously demonstrated that liquid milk samples can be frozen directly when rapid cold transfer between collection sites is not possible.

Application of cell lysis and DNA extraction methods to assess the microbiota of raw and pasteurised milk

Although cell lysis techniques can bias the results of 16S rRNA gene amplicon DNA sequencing, they are unlikely to interfere with comparative microbiota studies of different milk samples. The variation within a sample caused by methodological adjustments was significantly less than the variation between samples. Microbial DNA extraction and analysis may require different methodological approaches for different types of cow’s milk samples.

Summary

Examining the microbiota of dairy products requires careful sample preparation and analysis to minimise bias and contamination problems. The use of a bacterial cell mock community (BCMC) can help to verify and improve the accuracy of results. When selecting cell lysis and DNA purification methods, milder techniques should be preferred to avoid bias. Storage conditions of the collected cells should also be carefully controlled to avoid changes in bacterial composition. The use of 16S rRNA gene sequencing can help identify and track bacterial populations, but possible biases must be considered.

Tags: No tags

Comments are closed.