We report preliminary data showing that deep freezing NP swabs in STGG medium at -70°C may have a modest effect on the fingerprint of bacterial communities and a differential effect on the detection of bacterial OTUs found in the infant nasopharynx. Although the relative proportions of some OTUs changed, overall, the relative distributions were comparable before and after frozen storage (Figure 3). Previous studies which investigated the effects of frozen storage on the detection of various microbes in biological samples have shown minimal or no significant effects [1, 21, 27–29]. Abdullahi et al. (2007) reported that the recovery of S. pneumoniae from fresh and frozen NP swabs in STGG was consistent, but there were differences in the serotype distributions . This could be attributed to a differential capacity to survive the freezing process among S. pneumoniae serotypes. In another culture-based study, there was no effect of freezing on the recovery of S. pneumoniae and S. aureus from milk samples frozen at -20°C [27, 28], but there was an increase in the detection of coagulase negative staphylococci and a decrease in the recovery of Escherichia coli and Actinomyces pyogenes . Likewise, a recent 16S rRNA gene-based study of Black Band Disease showed that frozen storage increased the proportion of Proteobacteria phylotypes while direct analysis promoted the detection of cyanobacterial and sulfur-oxidizing bacteria .
These reports suggest that there may be differential survival capacities to frozen storage among bacterial taxa from the same community. As such, the deep freezing and thawing processes may alter the odds of detection and relative abundance of some, but not all OTUs in a biological sample. In this study, frozen storage significantly altered the odds of detecting a small proportion (<10%) of the bacterial OTUs found (Table 1) and the relative abundance of a couple of OTUs changed by more than 5% after frozen storage (Figure 2). This could be explained by DNA degradation amongst some bacterial taxa; Suomalainen and colleagues demonstrated that the freezing process results in the disintegration of the Flavobacterium columnare cell wall, associated with the release of large quantities of DNAse, lysases and proteases. There is further evidence that the structure and stability of bacterial cells influence cryo-preservation of nucleic acids .
Interestingly, the observed decrease in the number of OTUs found after frozen storage was significant among female (p = 0.0014) but not among male infants (p = 0.56) (Figure 1). This preliminary data suggests that an effect of freezing on microbial detection may also be differential across sexes. Interestingly, several OTUs detected in both sexes in fresh NP swabs were not detected or were detected at much lower frequencies among female infants post frozen storage. Although sex has been shown to be an important factor in colonization by various bacterial pathogens [31, 32], this finding has not been reported elsewhere to our knowledge. It is unclear how an effect of deep freezing on bacterial detection could be differential depending on the sex of a subject, which necessitates further investigations.
Frozen storage of biological samples is necessary for archiving and often done for logistic purposes where real-time processing of samples is not practicable. With the widespread use of vaccines targeting commensals of the respiratory mucosae, it is essential to effectively monitor non-vaccine serotypes and species replacement disease [33, 34]. Tracking intra-species serotype replacement and/or switching has overshadowed comprehensive research into the long term effects of vaccination on the microbiome, which may influence health, predisposition to disease and the pathogenesis of various infections [35, 36]. Loss or increased detection of particular taxa due to deep frozen storage may have a bearing on microbial ecology and species replacement surveillance. Furthermore, the intricate competitive relations between bacterial phylotypes may not be fully understood if deep freezing alters the fingerprint of microbial communities, albeit modestly [7–10].
In this study, we investigated the effect of frozen storage on NP swabs stored in STGG from 12 infants. However, the sample size and study design limit the validity of the findings. Broad investigations of different biological specimens, storage media, storage duration and microbial detection tools are needed to validate these findings. An effect of frozen storage on microbial detection using culture-based and culture-independent approaches needs to be studied. Finally, further investigation is needed to determine the precise mechanisms by which shifts in microbial community structure occur following frozen storage.