The human microbiome, comprising diverse and complex microbial communities, plays a crucial role in health and disease (Aggarwal et al., 2023). Among these, the oral and oropharyngeal microbiome have garnered significant interest due to the growing evidence of its role beyond general oral health, such as the involvement in respiratory infections and even systemic diseases (Bao et al., 2020; Lee et al., 2021; Peng et al., 2022). The oropharynx therefore serves as a critical interface between the upper aerodigestive tract and the external environment, making it a relevant diagnostic and research target. Compared to the gut microbiome, the oral microbiome is relatively underexplored, particularly in large-scale sequencing studies. It also differs in composition, pH, host immune interaction, and exposure to environmental factors (Huttenhower et al., 2012; Ding and Schloss, 2014). Moreover, age plays an important role: several studies have shown that the oral microbiome evolves significantly from infancy to adulthood, both in terms of taxonomic composition and stability (Sampaio-Maia and Monteiro-Silva, 2014; Burcham et al., 2020; Kageyama and Takeshita, 2024). These differences are particularly relevant when interpreting population-level data sets or making comparative references. Careful differentiation between pediatric, adolescent, and adult populations is therefore necessary. Moreover, the specific niche within the upper aerodigestive tract plays a decisive role in microbiome composition and taxonomic representation, and different anatomical sites show relevant biological differences. The oropharynx and nasopharynx, while spatially adjacent, differ substantially in epithelial lining, microbial density, immune surveillance, and exposure to environmental factors such as food, saliva, and inhaled particles (Piters et al., 2020). The oropharynx harbors a more diverse and metabolically active microbiota, with higher bacterial biomass and greater ecological connectivity to both the oral and gastrointestinal compartments (Lemon et al., 2010; Charlson et al., 2011; Huttenhower et al., 2012). Beyond these biological aspects, there are also practical advantages to studying the oropharyngeal microbiome: sampling is less invasive and more acceptable in both clinical and non-clinical settings, which facilitates routine implementation. The higher bacterial biomass further enhances the robustness and consistency of 16S rRNA gene amplification.

The rapid expansion of microbiome research has been largely driven by next-generation sequencing (NGS) technologies, which enable comprehensive, high-throughput analysis of complex microbial communities at increasingly affordable cost and turnaround times (Malla et al., 2019). Depending on read length and chemistry, sequencing platforms can broadly be categorized into short-read and long-read technologies. Short-read sequencing, most notably Illumina’s MiSeq^® platform (2 × 250–300 base pair), has become the most widely used approach in large-scale microbiome studies due to its high basecalling accuracy and established pipelines (Huttenhower et al., 2012; McDonalda et al., 2018; Ravi et al., 2018). However, its limited read length typically restricts analyses to partial hypervariable regions of the 16S rRNA gene—most commonly the V3–V4 or V4 region—constraining taxonomic classification primarily to the genus level and complicating comparisons across studies that target different regions (Klindworth et al., 2013; Kim et al., 2024). Moreover, species-level resolution is rarely achieved without additional genomic or functional information. Third-generation sequencing technologies such as Oxford Nanopore Technologies (ONT) overcome this limitation by generating substantially longer reads - up to 15 kilobases - enabling full-length 16S rRNA gene sequencing and improving phylogenetic resolution (Deissová et al., 2023). This is particularly advantageous for profiling complex microbial ecosystems and distinguishing closely related species. Although ONT sequencing was initially hindered by high error rates of approximately 6%, continuous improvements in flow cell design (e. g., R10.4.1), sequencing chemistry (e. g., Q20+ kits), and basecalling algorithms have markedly improved accuracy, now achieving modal read accuracies below 1% error (Kim et al., 2024). In clinical and diagnostic microbiology, ONT platforms offer additional benefits: they are compact, scalable, and enable real-time sequencing and analysis. This makes them attractive for point-of-care applications, outbreak investigations, and settings with limited laboratory infrastructure. However, challenges remain in standardization, bioinformatics pipelines, and benchmarking against short-read or whole-genome sequencing (WGS) approaches, which are still considered the gold standard for strain-level characterization and resistance profiling.

A critical source of variability in 16S rRNA gene-based microbiome profiling is the selection of primer pairs used for PCR amplification. Even minor mismatches between primer sequences and target regions - particularly in evolutionarily conserved but polymorphic regions - can introduce substantial amplification bias, leading to the preferential enrichment of certain taxa while underrepresenting others (Klindworth et al., 2013). This bias not only affects measures of alpha and beta diversity, but can also distort downstream taxonomic assignments, especially when comparing data across studies using different primer sets or targeting different regions of the gene (Deissová et al., 2023).

To address this issue, degenerate primers have been developed that incorporate nucleotide ambiguity codes at variable positions, thereby increasing coverage across a broader range of bacterial taxa. While this strategy can improve amplification inclusivity and reduce taxonomic dropout, it may also introduce challenges such as reduced amplification efficiency, increased non-specific binding, and the need for optimized PCR conditions (Frank et al., 2008).

In our previous study on human fecal samples, we systematically compared ONT’s standard 27F primer with a more degenerate variant and demonstrated that the latter resulted in significantly higher alpha diversity and a more balanced phylum-level distribution, with reduced overrepresentation of Firmicutes and Proteobacteria (Waechter et al., 2023).However, the extent to which these findings apply to other anatomical sites remains uncertain, as microbial composition, DNA extraction yield, and sequence conservation can vary widely between niches such as the gut, skin, and oral cavity (Huttenhower et al., 2012). The present study therefore extends our previous work to the oropharyngeal microbiome, a distinct and clinically relevant niche characterized by high microbial diversity and diagnostic potential. By systematically comparing primer sets in this anatomical context, our study contributes to the growing body of evidence on the influence of primer design in microbiome profiling and offer practical guidance for future studies of the oral and respiratory tract using long-read sequencing technologies.

Utilizing full-length 16S rRNA gene amplicon sequencing on the nanopore platform, we evaluated the efficiency of two primer sets: the standard 27F primer (designated as 27F-I) from ONT’s 16S Barcoding Kit (SQK-16S024) and a more degenerate variant (designated as 27F-II), designed to account for polymorphisms in conserved regions of the 16S rRNA gene. This comparison was conducted in the context of highly diverse bacterial communities from 80 human oropharyngeal swab samples. Demographic and baseline characteristics of the study cohort are summarized in Supplementary Table S2. The comparative primer strategy employed follows the four-primer PCR method outlined by Matsuo et al (Matsuo et al., 2021). This approach involves an initial PCR step utilizing a more degenerate 27F and 1492R primer pair [S-D-Bact-0008-c-S-20 and S-D-Bact-1492-a-A-22 (Klindworth et al., 2013)], followed by a barcoding PCR. Reads were aligned directly to the NCBI 16S database for taxonomic classification.

To globally compare the taxonomic profiles of the human oropharyngeal microbiota obtained with the two primer sets, the Pearson correlation coefficient (r) was calculated based on the average relative abundances of bacterial genera across all samples for each primer approach. The analysis showed only a moderate but statistically significant correlation (r = 0.67, p = 0.005) between the genera identified by the respective primer sets. To assess which primer more accurately represents the oropharyngeal microbiome, the taxonomic data generated using the 27F-I and 27F-II primers were compared to a reference dataset assembled by Ruan et al., which includes saliva samples from 1,989 healthy subjects (Ruan et al., 2022). The analysis revealed a strong and statistically significant correlation between the taxonomic profile of oral samples obtained with the 27F-II primer and the cited reference dataset (r = 0.86, p < 0.0001). In contrast, the correlation between the taxonomic profiles generated using the 27F-I primer and the reference dataset was weak and not statistically significant (r = 0.49, p = 0.06). Figure 1 presents a heatmap comparing the relative abundance of the 12 most prevalent genera identified by the two primer sets.

A noticeable discrepancy in relative abundance is evident even at the phyla level. Across all analyzed samples, the 27F-I primer yielded a significantly higher proportion of Proteobacteria (49.2% vs. 29.2%, p < 0.001) and lower abundances of Bacteroidota (5.1% vs. 19.2%, p < 0.001), Actinobacteria (0.1% vs. 1.3%, p < 0.001), and Verrucomicrobia (0.001% vs. 0.08%, p < 0.001) compared to the 27F-II primer. Figure 2 presents an overview of the relative abundance of different phyla, both as an average across all samples and at the individual sample level. Detailed quantitative data for all bacterial phyla are available in Supplementary Table 3.

At the genus level, substantial differences in relative abundance were observed for 125 genera. Focusing on the 10 genera with the most significant differences, the 27F-I primer led to a higher relative abundance of Haemophilus (33.6% vs. 12.1%, p < 0.001) and Campylobacter (3.8% vs. 1.4%, p < 0.001). In contrast, the 27F-II primer detected significantly higher levels of Prevotella (3.1% vs. 12.4%, p < 0.001), Porphyromonas (0.5% vs. 1.7%, p < 0.001), Faecalibacterium (0.4% vs. 1.5%, p < 0.001), Blautia (0.076% vs. 1.07%, p < 0.001), Bacteroides (0.009% vs. 0.1%, p < 0.001), Citrobacter (0.00003% vs. 0.07%, p < 0.001), Rothia (0.004% vs. 0.064%, p < 0.001) and Phascolarctobacterium (0.005% vs. 0.0058%, p < 0.001) compared to 27F-I (Figure 3). Comprehensive quantitative data for all genera are provided in Supplementary Table 4. Since the 16S Barcoding Kit (SQK-16S024) containing 27F-I is validated only for genus-level resolution, species-level classification was not conducted.

Beyond these taxonomic differences in the oropharyngeal microbiome, the choice of primer set also significantly influenced taxonomic diversity. The 27F-I primer detected fewer distinct amplicon sequence variants (ASV) in oropharyngeal swabs than the 27F-II primer, as reflected by a significantly lower Shannon index (1.850 vs. 2.684, p < 0.001), indicating reduced alpha diversity (Figure 4).

The advent of next-generation sequencing has transformed microbiology research, significantly enhancing our understanding of complex human gut bacterial communities. Among these technologies, nanopore sequencing has gained prominence due to its unique combination of cost efficiency, ease of use, high throughput, and superior taxonomic resolution, enabled by its ability to sequence long amplicons. Recent breakthroughs in sequencing accuracy have largely addressed one of the technology’s key limitations, marking a major milestone in the rapid evolution of the nanopore platform. These improvements have allowed nanopore sequencing to rival and, in some cases, surpass the capabilities of traditional short-read sequencing approaches. Additionally, the widely adopted 16S Barcoding Kit (SQK-16S024) from Oxford Nanopore Technologies (ONT) has further streamlined 16S rRNA gene sequencing, making it an accessible, fast, and cost-effective solution for microbiome research (Santos et al., 2020).

This study provides a systematic comparison of two primer sets with different levels of degeneracy for full-length 16S rRNA gene amplification from human oropharyngeal swab samples using nanopore sequencing. By adapting the approach of our previously published fecal microbiome study (Waechter et al., 2023) to the oral cavity, we extend the evidence that primer selection is a crucial determinant of sequencing outcome and diversity metrics in complex microbial environments.

The results demonstrate that the more degenerate primer set (27F-II) outperforms the standard ONT kit primer (27F-I) in capturing microbial diversity and in achieving a taxonomic composition more consistent with a reference dataset from nearly 2,000 healthy individuals’ salivary microbiota (Ruan et al., 2022). This is in line with previous findings from fecal samples, where the degenerate primer set led to higher biodiversity and a more balanced representation of key phyla, including Bacteroidota and Actinobacteria (Waechter et al., 2023). Our current data confirm this pattern in the oropharyngeal microbiome, suggesting that primer-induced amplification bias is not limited to gut environments but similarly affects oral microbial profiling.

The discrepancies observed between the two primer sets are particularly pronounced at both the phylum and genus levels. The 27F-I primer set yielded a microbial profile dominated by Proteobacteria and Haemophilus, reflecting overamplification of specific taxa. In contrast, the degenerate primer set enabled a more diverse detection spectrum, revealing increased levels of clinically relevant genera such as Prevotella, Porphyromonas, and Faecalibacterium, which are often underrepresented in datasets generated with less degenerate primers. These findings highlight the risk of skewed taxonomic inference when using primers with limited degeneracy, especially in environments with high microbial variability such as the oropharyngeal cavity.

Importantly, the more degenerate primer set also led to significantly higher alpha diversity, as indicated by the Shannon index. This reinforces the notion that primer degeneracy enhances the detection of low-abundance taxa, contributing to a more comprehensive and ecologically valid microbiome profile. The correlation with the large-scale salivary microbiome dataset by Ruan et al. further strengthens the validity of the degenerate primer set for oral microbial community profiling and supports its use as a methodological standard in future studies (Klindworth et al., 2013).

Our findings also have important implications beyond primer performance. While full-length 16S rRNA gene sequencing using long-read platforms such as ONT allows comprehensive coverage across all nine hypervariable regions (V1–V9), this does not inherently guarantee higher taxonomic resolution for all bacterial clades. Several studies have demonstrated that targeted short-read sequencing, particularly of the V1 - V4 or V3 - V4 regions using Illumina technology, can outperform full-length approaches in certain contexts. This is especially true for taxa whose discriminative nucleotide signatures are concentrated in specific regions of the gene, such as Bifidobacterium, Lactobacillus, or Enterobacteriacea members, where short-read methods have shown better genus- or even species-level concordance with whole-genome data (Janssen et al., 2018; Macip et al., 2025).

Moreover, the error profile of long-read sequencing, although significantly improved in recent ONT chemistry (e.g., Q20+ kits), may still impair taxonomic resolution at lower ranks when not properly corrected (Liu-Wei et al., 2024). This is particularly relevant in clinical diagnostics, where misclassification of near-neighbor taxa may lead to false-positive or false-negative results (Gu et al., 2018). As such, the decision between short- and long-read platforms should be guided by the biological context, the expected diversity and complexity of the sample type, and the resolution required for the intended application. For instance, high-throughput surveillance studies may prioritize cost-effective short-read platforms with robust pipelines, whereas exploratory profiling of under-characterized niches may benefit from the broader coverage of full-length sequencing (Wenger et al., 2019). Clinical applications may require additional benchmarking or validation with mock communities to ensure sufficient taxonomic precision and reproducibility.

Recent improvements in ONT sequencing chemistry and basecalling have substantially increased the accuracy of full-length 16S rRNA gene sequencing, thereby enhancing its potential to resolve complex microbial communities with higher taxonomic resolution than conventional short-read approaches. Given its cost-efficiency, scalability, and ability to sequence full-length amplicons in real time, the ONT platform is poised to gain increasing importance in oral and oropharyngeal microbiome research.

Our study presents a comparative analysis of two primer sets with different levels of degeneracy for nanopore-based 16S rRNA gene sequencing of human oropharyngeal swabs. We demonstrate that the widely used standard 27F primer (27F-I) introduces measurable amplification bias, whereas a more degenerate variant (27F-II) yields richer and more representative taxonomic profiles. These findings underscore the critical role of primer selection in shaping microbiome readouts and support the broader use of degenerate primers for accurate and unbiased profiling in complex oral environments.

Looking ahead, future studies should aim for greater methodological harmonization, particularly in the design and selection of primer sets. The current lack of interoperability among primer strategies remains a major obstacle to reproducibility and cross-study comparability. Establishing community-wide standards for primer choice, as well as unified guidelines for the selection of taxonomic reference databases across anatomical niches and sequencing platforms, will be essential for advancing microbiome research toward clinical and translational applications.