Page 94 - Read Online

Longhi et al. Microbiome Res Rep 2024;3:4 https://dx.doi.org/10.20517/mrr.2023.02 Page 9 of 15

However, the same limited impact on the bacterial DNA content was not observed for other biological
matrices such as biopsies (gastric antrum and body), saliva, and sputum samples for which the same
saponin-based extraction method induced a substantial alteration of the whole microbial taxonomic profile.
In this case, the taxonomic variation index between treated samples and controls was 190.3 for the gastric
antrum, 180.5 for the gastric body biopsies, and 129.7 for the sputum, respectively [Supplementary Table 2].
Consistently, 14 bacterial taxa with a prevalence between 0.5% to 25.7% were lost in the saponin-treated
biopsy of the gastric antrum, 33 in the biopsy of the gastric body, 17 in the saliva, and 12 in the treated
sputum [Supplementary Table 3 and Supplementary Figure 2]. Although with a relative abundance greater
than 2%, the most represented bacterial taxa in the saponin-treated samples ranged from 4 bacterial taxa in
the biopsies to 15 and 11 in the saliva and sputum, respectively [Supplementary Table 3 and Supplementary
Figure 2]. In saliva, sputum, and biopsy specimens not treated with saponin, several members of
Haemophilus, Neisseria, Prevotella, and Veilonella genera were present with a relative abundance higher
than 10% and were the most representative bacterial taxa of these biological samples, while in the respective
samples treated with saponin, they fell under the limit of detection [Supplementary Table 3 and
Supplementary Figure 2].

In contrast, several bacterial taxa not detected in the untreated specimens appeared in the profile of
saponin-treated samples, with few cases (such as Streptococcus spp. and Actinomyces spp.) where these taxa
even became dominant in the retrieved taxonomic profiles [Supplementary Figure 2 and Supplementary
Table 3]. Altogether, these data highlighted the impact of the saponin-based eukaryotic DNA depletion
method, which can markedly alter the retained bacterial DNA profile.

In this context, we noticed that the depletion protocol less compromised samples with a substantially high
bacterial load. Conversely, when processing samples poor in bacterial DNA, such as mucosa, saliva and
sputum, the saponin treatment tended to reduce the already low bacterial population, resulting in the
sequencing of a lower number of bacterial DNA sequences, causing the loss of the minor component of the
bacterial population in the taxonomic profiles retrieved after data analysis. Therefore, these results
significantly impact studies in which the purpose is represented by comparing and analyzing complex
bacterial populations constituted by many microbial species at low relative abundance. For this reason,
preliminary testing of biological matrices included in each study should be performed.

On the other hand, it should be considered that studies focused specifically on the detection of one or few
target microbial taxa in simple microbial communities are less affected by the use of saponin depletion
protocol, which may represent an extremely useful tool by increasing the microbial/host DNA ratio. For
example, in the case of pathogen detection for clinical diagnosis, the use of saponin would be highly
recommended to remove the large amount of human DNA present in the biological samples, in particular
when applying shotgun or PCR/amplicon DNA sequencing approaches targeting and amplifying specific
microbial taxa.

Saponin/DNase treatment affects the Gram-positive/Gram-negative ratio of the retrieved
metagenomic profiles
Microbial profiling revealed a differential response to saponin treatment of Gram-positive and Gram-
negative bacteria in the biological samples analyzed. More specifically, although recent literature
demonstrates a potential sensitivity of Gram-positive bacteria to saponin treatment, especially of microbes
recovered from blood cultures [31,32] , in this study employing a range of different matrices, saponin seemed to
have an impact on the relative abundance of Gram-negative bacteria that prompted a concomitant increase
in the proportion of Gram-positive microbes with respect to the whole population. It has been generally
accepted that bacteria differ in their susceptibility to cell lysis based on the molecular structures composing