Page 129 - Read Online

Page 14 of 19 Wang et al. Microbiome Res Rep. 2025;4:23 https://dx.doi.org/10.20517/mrr.2024.94

The proposed DNA spike-in quantification method offers several advantages. Compared to flow cytometry,
the spike-in method is faster, easier to perform, and provides better sample reproducibility. Flow cytometry,
on the other hand, is labor-intensive and time-consuming, with optimal counting accuracy only within the
7
[14]
5
range of 10 to 10 bacteria per mL , which may limit its reliability in low-biomass or antibiotic-affected
samples such as infant feces. Commercial products such as the ZymoBIOMICS Spike-in Control (Zymo
TM
Research) have adopted a related strategy for absolute quantification, using non-native bacterial cells added
directly to fecal samples prior to DNA extraction. In contrast, our approach adds purified spike-in DNA
after extraction, which avoids biases introduced by variable cell lysis efficiency. This also enables absolute
quantification for previously extracted or archived samples, which would be incompatible with cell-based
spike-in strategies. Together, these features highlight the methodological flexibility of spike-in–based
quantification, as reflected in both commercial and research settings.

A limitation of the spike-in method is its reliance on copy number as the measurement standard, which can
design make them less suitable for high-throughput, whole-community analyses. Future strategies may
lead to the propagation of PCR amplification errors from the spike-in bacterium to other taxonomic
units . This occurs because different organisms have varying 16S rRNA gene copy numbers (GCNs),
[46]
[47]
causing sequence variant counts to be biased toward clades with higher GCNs . While this primarily
affects comparisons between different genera, it has less impact on comparisons within the same genus, as
[15]
genera-specific amplification errors tend to cancel out .
To mitigate this effect, using multiple spike-in bacteria with fixed copy numbers across samples and
averaging or summing their counts can be more effective. In this study, we selected two phylogenetically
distinct spike-in strains (Gram-positive and Gram-negative) to average out potential taxon-specific
amplification effects, thereby enhancing the robustness of absolute quantification. Furthermore, based on
both previous studies and manufacturer guidelines (ZymoBIOMICS Spike-in Control), it is recommended
that spike-in DNA should constitute between 0.1% and 10% of the total microbial DNA. Maintaining the
spike-in proportion within this range helps ensure sufficient detection sensitivity while minimizing
potential perturbations to the native microbial community structure. In our study, spike-in concentrations
were adjusted based on preliminary testing to consistently fall within this optimal range across all samples.
These tests showed that adding spike-in DNA corresponding to approximately 10 -10 copies per sample
7
6
typically yielded a final proportion within the 0.1%-10% range across a variety of microbial loads. This
strategy further reduces the likelihood that amplification biases from spike-in bacteria significantly affect the
accuracy of absolute quantification.

Additionally, single-copy housekeeping genes (e.g., rpoB, recA) offer a theoretically more stable alternative
for cell-based quantification, but their limited phylogenetic coverage and the need for taxon-specific primer

combine spike-in normalization for total load calibration, targeted detection of single-copy genes for key
taxa, and metagenomic approaches to avoid PCR amplification bias altogether.

Another limitation arises from the detection threshold of 16S rRNA sequencing, where low-abundance taxa
(approximately fewer than 10 cells) may be reported as having zero abundance due to insufficient
5
sequencing depth . In this study, one infant sample exhibited a zero abundance of Bifidobacterium in the
[48]
16S rRNA sequencing results, while qPCR analysis detected approximately 10 copies, reflecting the higher
3
sensitivity of qPCR. Consequently, using DNA spike-in with 16S rRNA sequencing may lead to an
underestimation of the total microbial load. However, employing DNA spike-in with shotgun metagenomic
sequencing can provide a more accurate and comprehensive quantification of microbial abundance.