Differential effects of frozen storage on the molecular detection of bacterial taxa that inhabit the nasopharynx

Abstract

Background: Frozen storage often precedes metagenomic analysis of biological samples; however, the freezing process can have adverse effects on microbial composition. The effect of freezing on the detection of bacteria inhabiting the infant nasopharynx, a major reservoir of bacterial pathogens, was investigated.

Methods: 16S ribosomal RNA (rRNA) gene-based terminal restriction fragment length polymorphism (T-RFLP) analysis of nasopharyngeal (NP) swabs from twelve Gambian infants was employed. NP swabs were analysed within hours of collection and then after 30 days of storage at -70°C.

Results: There was substantial heterogeneity among subjects with respect to the effect of freezing on the number of operational taxonomic units (OTUs) detected. Nevertheless, the mean number of OTUs decreased after frozen storage and the relative abundance for 72% of the OTUs changed by less than 0.5% after deep frozen storage. There were differences in the odds of detection and relative abundance of OTUs matched with Moraxella sp., Haemophilus sp./Burkholderia sp., and Pseudomonas sp. A strong interaction between sex and the effect of freezing was found, whereby there was no significant change observed for males while the mean number of OTUs significantly declined among female infants following frozen storage.

Conclusions: Although frozen storage of biological samples is often necessary for archiving and logistic purposes, the potential effects on the number of taxa (composition) detected in microbial community studies are significant and should not be overlooked. Moreover, genetic factors such as sex may influence the integrity of nucleic acids during the freezing process.

Figure 1

Effect of frozen storage on the composition of bacterial OTUs found in the infant nasopharynx. The composition (mean number of OTUs) per subject against 16S rRNA-based T-RFLP before and after freezing at -70°C dichotomized by sex. Dotted lines represent females, and the bold dotted line shows the mean change. The difference in composition pre and post freezing was significant for female (p = 0.0014) but not for male infants (p = 0.56).

Figure 2

Bar graph showing proportion change in relative abundance of bacterial OTUs after frozen storage of NP swabs at -70°C analysed by 16S rRNA T-RFLP. The difference in relative abundance before and after frozen storage was expressed as a proportion of the relative abundance for direct analysis. Most of the 42 OTUs detected showed minimal (<0.5%) change and OTUs that had proportion change >0.5% are shown here.

Figure 3

Bar Graph showing the relative distribution of the bacterial OTUs detected before and after frozen storage of NP swabs at -70°C amongst male and female infants by 16S rRNA-based T-RFLP. Partial 16S rRNA gene sequences from infant nasopharyngeal clone libraries were BLASTED to identify the microbes (>97% sequence similarity) and in silico T-RFLP analysis was used to match them to the OTUs.