We designed and evaluated a new expanded-coverage bacterial quantitative real-time PCR assay targeting the 16
S rRNA gene. To accomplish this, we curated a set of high-quality 16
S rRNA gene sequences for assay design and evaluated the coverage of our primers and as a union (rather than as separate entities). In addition, we improved the quantitative capacity of our assay using a cloned plasmid standard. Our computational and laboratory analyses showed that BactQuant had superior in silico
taxonomic coverage while retaining favorable in vitro performance. As would be expected, the diverse gene sequences targeted by BactQuant have resulted in variable reaction efficiencies. Nevertheless, laboratory evaluation showed 100% sensitivity against perfect match species from the in silico
To allow researchers to determine whether BactQuant covers key organisms in their target community, we provided additional detailed OTU coverage information in the Supplemental Files. We have applied the logic that an OTU was covered if it contained at least one perfect match sequence in the in silico
S rRNA gene sequences with ambiguous or degenerate bases at the primer and probe sites were considered non-perfect matches, thus making our coverage estimates more conservative. Lastly, although we prohibited the use of a degenerate probe to maximize our assay’s quantitative ability, this approach may permit detection of specific taxa such as Chlamydia
For most studies, the desired measurement of bacterial load is the number of cells rather than 16
S rRNA gene copy number; however, the 16
S rRNA gene copy number varies among bacterial species and even among strains [29
]. The range of copy number is estimated at one to 14, with most non-spore forming species having fewer than 10 copies per genome [20
]. We use the average 16
S rRNA gene copy number per genome from rrnDB in our genomic equivalent estimation, but alternative approaches are possible. This, combined with logarithmic growth of bacteria, suggest that using estimated average copy number could be sufficient.
The in silico
analysis was an important component of our validation of BactQuant against diverse bacterial sequence types, even though sequence matching is not a perfect predictor of laboratory performance [31
]. Many factors are known to affect reaction efficiency, such as oligonucleotide thermodynamics, the type of PCR master mix used, and the template DNA extraction method. Concentration of background nontarget genomic DNA is another factor that can affect the quantitative parameters rRNA gene-based assays [32
]. The interference of background human DNA with BactQuant dynamic range reported in this paper was most likely due to cross-reactivity of human DNA with the probe, which targets a region conserved even among eukaryotic organisms, including in the human 18
S rRNA gene. This may be overcome by using an intercalating reporter dye in place of a fluorescent probe as a qPCR reporter mechanism; however, the loss of tertiary-level of specificity is a potential concern in direct application of an intercalating dye assay to specimens containing high amounts of nontarget DNA.
Exogenous bacterial DNA, particularly from biologically synthesized reagents such as Taq DNA polymerase are a known limitation for analyzing samples with low bacterial load [28
]. Recently, this issue has received renewed attention due to increased usage of next-generation sequencing and the frequent data contamination from exogenous bacterial DNA. Several methods have been evaluated for removing bacterial contaminants from Taq DNA polymerase, including UV irradiation [34
], DNAse I treatment, and ultrafiltration [36
]. The level of E. coli
contamination in Taq DNA polymerase has been estimated at 102
genome equivalents of bacterial DNA per unit of enzyme [28
]. This is consistent with the lowest amount of contamination we have observed in our experiments, which were 5 and 10 copies of 16
S rRNA gene in 5
μl and 10
μl reactions, respectively. The ubiquity of bacterial DNA also makes the determination of assay specificity challenging.
Our use of qPCR-quantified plasmid standards addressed a major limitation in the preparation of qPCR quantification standards. The conventional approach of quantifying bacterial genomic DNA or plasmid standards necessitates converting mass (i.e., nanograms per μl) to copy number (i.e., 108 copies per μl) and can introduce substantial error. Thus far, we have also successfully applied BactQuant to a diverse range of clinical specimens, including swab eluents, surgical specimens, and respiratory specimens, but we did not present these findings in this paper. To fully understand the likelihood of false negative results due to interference from human DNA or BactQuant’s limit of detection will require additional evaluations.