Skip to main content

Quantifying Live Microbial Load in Human Saliva Samples over Time Reveals Stable Composition and Dynamic Load.

Marotz C, Morton JT, Navarro P, Coker J, Belda-Ferre P, Knight R, Zengler K

Publication: mSystems (2021) • Vol 6PMID: 33594005View on PubMed

Purpose

This paper compared the microbial composition of propidium monoazide (PMA)-treated and untreated saliva specimens obtained at nine different timepoints over the course of a single day and after swishing water, soda, antiseptic mouthwash, and alcohol-free mouthwash. The authors also compared the salivary flow rates at different times of day, after eating, in males and females, and in patients of different ages.

Conclusion

Glycerol concentration in the specimens varied between 8-38% but did not correlate with microbial diversity. The percentage of live cells varied from 1-100% among the timepoints with a highest percentage of live cells found in the specimen obtained after waking and the lowest after teeth brushing. Salivary flow rate was negatively correlated with microbial contamination and the number of live cells as measured by flow cytometry but not 16S rRNA gene copy number. More than 99% of the sequenced rRNA and shot-gun sequencing came from five phyla, with 16 genera accounting for >90% of the sequences and chloroplast DNA predominantly found in specimens obtained after eating. After PMA treatment, fewer Spirochaetes and less of the chloroplast 16s rRNA gene was found, indicating this DNA was not from live cells. Aitchison distance metric showed average differences in microbial communities as determined by rRNA sequencing and shot-gun sequencing were greater among participants than within individuals. Salivary flow rates were lower in the first two timepoints than later in the day and increased after eating but were comparable in males and females and not correlated with patient age. PMA treatment of the specimen increased the observed variability of the microbial community with time, showing that the presence of relic DNA not from intact cells may mask changes. Measured bacterial concentrations were significantly correlated between flow cytometry and real-time 16S rRNA real-time PCR.

The microbial load in saliva was significantly lower 15 min after swishing alcohol-free mouthwash than before but, two hours after swishing alcohol-free mouthwash, microbial loads were comparable to before swishing. There were significantly more dead microbial cells than live cells 2 h after swishing antiseptic mouthwash compared to 15 min after swishing, indicating a delayed effect of antiseptic mouthwash. Swishing water and drinking soda had no effect on the saliva microbial load.  There was no correlation between microbial load and alpha diversity. The UniFrac distance was higher in the alcohol-free mouthwash and soda groups than in the antiseptic mouthwash group when PMA treated and also significantly different from the water group when PMA-treated.

Studies (2)

Purpose

This study compared the microbial composition of PMA-treated and untreated saliva specimens obtained at nine different timepoints over the course of a single day and investigated the effects of collection timing, eating, and patient gender and age on salivary flow rates. Unstimulated saliva from 10 healthy people was self-collected for 5 min by funnel into tubes containing glycerol at nine different timepoints over the course of the day. Saliva was collected immediately after waking, again after teeth brushing, and then every 2-3 h; recording if the subject ate or brushed their teeth. Saliva was inverted ten times, frozen at -20°C, transported on ice, and stored again at -20°C. Saliva was thawed on ice and centrifuged at 3,000 x g for 1 minute and the volume and weight measured. Cells were diluted in PBS and filtered through a 5 µm syringe filter and treated with 10 µM PMA for 5 min or left untreated and then exposed to fluorescent light for 25 min. Saliva was then stained with Sybr-green, mixed with AccuCount fluorescent particles, and counted on a SH800 cell sorter. DNA was extracted from raw and PMA-treated saliva using the Qiagen PowerSoil MagAttract DNA Kit. DNA was quantified using Quant-iT PicoGreen. DNA was amplified using primers for the V4 region of the 16S rRNA gene and pair-end sequenced using the 150 V2 Kit on an Illumina MiSeq. Shot-gun pair-end sequencing was run on an Illumina HiSeq2500 v2. Taxonomy was assigned using the Human Oral Microbiome Database and alpha and beta-diversity calculated using QIIME 2.

Summary of Findings

Glycerol concentration in the specimens varied between 8-38% but did not correlate with microbial diversity. The percentage of live cells varied from 1-100% among the timepoints with a higher percentage of live cells found in the specimen obtained after waking, lower after teeth brushing, then fluctuated a bit, and dropped again when teeth were brushed a second time but was unaffected by eating. The number of live microbial cells collected in the 5 min timeframe displayed more than 3 order of magnitude intra- and inter-individual variability. Salivary flow rate was negatively correlated with microbial contamination (r=-0.326, P=0.009) and the number of live cells (r=-0.377, P=0.003) as measured by flow-cytometry but not 16S rRNA gene copy number. More than 99% of the sequenced rRNA came from five phyla, with 16 genera accounting for >90% of the sequences. Shot-gun sequencing confirmed that these five phyla and 16 genera represented >99% and 89% of the taxonomic hits, respectively. Fewer Spirochaetes, particularly Treponema, were found after PMA treatment, indicating this DNA was not from live cells. Chloroplast 16S rRNA gene was found in >49% of the specimens collected after eating but only 16% of the other specimens and the number of specimens with chloroplast rRNA declined after PMA treatment, indicating that the majority of the DNA was not from intact cells. For specimens in which the chloroplast rRNA gene was detected, the average abundance of chloroplast 16S rRNA gene was higher after eating than in other specimens (0.6% versus 0.04% of reads). Aitchison distance metric showed average differences in microbial communities as determined by rRNA sequencing and shot-gun sequencing were greater among participants than within individuals. Analysis of the functional pathways identified 355 pathways, which were more similar between timepoints in a participant than between participants. Salivary flow rates were lower at the first two timepoints than later in the day (P=0.015) and increased after eating (P=0.027) but were comparable in males and females (P=0.398) and were not correlated with patient age (r=-0.003, P=0.964). PMA treatment of the specimen increased the observed variability of the microbial community with time, showing that the presence of relic DNA not from intact cells may mask changes. Measured bacterial concentrations were significantly correlated between flow cytometry and real-time 16S rRNA real-time PCR in both raw specimens (r=0.272, P=0.011) and PMA-treated samples (r=0.333, P=0.002).

  • Biospecimens

    FluidSaliva
  • Preservation Types

    Frozen
  • Diagnoses

    Normal
  • Platforms

    DNANext generation sequencing
    Cell count/volumeMacroscopic observation
    Cell count/volumeFlow cytometry
    Cell count/volumeFluorescent microscopy
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    PreaquisitionPatient age
    Unspecified range with mean of 36.9 years and SD of 18.5 years
    PreaquisitionPatient gender
    Female
    Male
    Biospecimen AcquisitionTime of biospecimen collection
    9 different times
    After eating
    After teeth brushing
    After waking (before teeth brushing)
    Analyte Extraction and PurificationNucleic acid digestion
    PMA treated
    Untreated
    Next generation sequencing SpecificTechnology platform
    16s rRNA sequencing
    Shot-gun sequencing