Application and Limitations of 16S rRNA Gene Sequencing for Identifying WHO Priority Pathogenic Gram-Negative Bacilli

Introduction

Since their discovery in the 20th century, antibiotics have revolutionized medicine, saving millions of lives over the decades¹. However, the widespread and often indiscriminate use of these drugs has driven the emergence of antibiotic-resistant bacterial strains.^1,2 Thus, antimicrobial resistance (AMR) is recognized as the most urgent global public health challenges of the 21st century, especially after the COVID-19 pandemic, when the misuse of antibiotics was even more intense.^3,4 This challenge is particularly critical in healthcare settings, where multidrug resistant pathogens are the primary cause of healthcare-associated infections, representing one of the greatest threats to the patient’s safety.^5,6

Healthcare-associated infections (HAIs) are a significant global burden, increasing hospital stay length, treatment costs, and, critically, patient morbidity and mortality.^7,8 This issue was further exacerbated during the COVID-19 pandemic, when hospitalization rates and antimicrobial use increased significantly.⁹ Langlete et al⁹ performed a retrospective registry-based study comparing HAIs and community-associated COVID-19 infections in 54,885 COVID-19 cases identified in patients hospitalized between January 1st, 2019, and January 1st, 2023. The authors reported that mortality rates were consistently higher among HAIs patients compared to community-associated COVID-19 infections patients, the difference being highest shortly after infection. The rise of AMR complicates this landscape, as it limits therapeutic options for infections that are already difficult to treat, leading to poorer clinical outcomes.^6,10 Gram-negative bacteria are already intrinsically resistant to a wide range of antibiotics, including β-lactams, quinolones and polymyxin, due to their outer membrane structure.¹¹ Among them, Acinetobacter baumannii, Pseudomonas aeruginosa and order Enterobacterales stand out as relevant infectious agents.^6,12,13 Along with their intrinsic resistance, these species also exhibit a remarkable capacity to acquire new resistance determinants, having rapid adaptation to selective pressure and high genetic plasticity that allows the accumulation of multiple factors, ultimately leading to multidrug resistance.^6,14,15

The World Health Organization (WHO) has included A. baumannii, P. aeruginosa, carbapenem-resistant Enterobacterales and third-generation cephalosporin-resistant Enterobacterales in its list of priority bacterial pathogens, highlighting the urgent need to prioritize research and studies focusing on these microorganisms.⁶ Actually, the order Enterobacterales possesses eight valid families: Budviciaceae, Enterobacteriaceae, Erwiniaceae, Gallaecimonadaceae, Hafniaceae, Morganellaceae, Pectobacteriaceae, and Yersiniaceae.¹⁶ Among all the genus and species from the order, members of Enterobacteriaceae are the most associated with human infections associated to AMR.¹⁷ Klebsiella pneumoniae, Enterobacter spp. and Escherichia coli belong to ESKAPEE (Enterococcus faecium, Staphylococcus aureus, K. pneumoniae, A. baumannii, P. aeruginosa, Enterobacter spp. and E. coli), a clinical relevant group of bacteria that can “escape” from the effects of standard antibiotics, thus making them fatal adversaries, especially in clinical settings.¹⁸ Due to the clinical relevance of these species, bacterial diagnostic testing is critical for the identification and characterization of bacterial infections, the prescription of targeted treatments, and the prevention of the spread of disease.^2,19

Although diagnostic technologies have advanced significantly, most individuals with infectious diseases continue to receive empiric treatment, which also leads to the excessive use of antibiotics.^2,10,20 Moreover, different species present different resistance profiles, and the incorrect identification may deprive the patient of receiving the most appropriate antibiotic choice.²¹ Conventional diagnostic methods for bacterial identification, including culture-based biochemical testing, are still widely used in clinical microbiology; however, these approaches are inherently time-consuming and often lack precision for closely related or fastidious species.¹⁹ The process typically requires 24 to 72 h for growth and phenotypic characterization, delaying appropriate therapeutic or containment measures.¹⁰ Furthermore, phenotypic variability and overlapping biochemical profiles among certain genera can lead to misidentification or inconclusive results.¹⁹ Therefore, the use of fast and accurate identification techniques is essential, as it would speed up the procedures of contamination control, the traceability of microorganisms, reduce the misuse of antibiotics and serve as a crucial tool for the control and prevention of healthcare-associated infections.²²

One of the techniques that can be applied for bacteria identification is the 16S ribosomal RNA (rRNA) gene sequencing.²³ This is one of the most consolidated and widely used molecular approaches for studying bacterial phylogeny and taxonomy.²⁴ The 16S rRNA gene is present in all bacteria, is approximately 1,500 base pairs (bp), and contains conserved and hypervariable regions (V1 to V9), which makes it ideal for the identification and classification of microorganisms.²⁵ The conserved regions allow the use of universal primers for amplification. The hypervariable regions (V1 to V9) allow differentiation between genera and, in some cases, species.²⁶

The 16S rRNA gene is encoded within the rRNA operon (rrn), which contains the genes responsible for the synthesis of ribosomal RNA, essential for the assembly of bacterial ribosomes (Figure 1). The 16S rRNA gene is not translated into proteins. It is transcribed into RNA and becomes part of the small subunit (30S) of the bacterial ribosome. The number of copies of the rrn operon can vary between bacterial species (from 1 to more than 15 copies).

Figure 1 Organization of the bacterial rrn operon.

Although the presence of multiple rrn operons enhances ribosomal synthesis and cellular adaptability, it also introduces complexity in molecular diagnostics. They are generally very similar to each other, but microvariations can exist. For example, Escherichia coli has seven copies of the rrn operon distributed along the chromosome.²⁷ The existence of several 16S rRNA gene copies within a single genome, often showing slight sequence heterogeneity, can complicate consensus sequence assembly and lead to ambiguous taxonomic assignments.²⁴ This intra-genomic variation may generate conflicting signals during sequence alignment or database comparison, occasionally resulting in misidentification at the species level. Recognizing this limitation is crucial for accurate interpretation of 16S rRNA-based analyses, particularly in clinical diagnostics where precision and reliability are essential.^25,26

Among the diverse molecular and proteomic techniques currently employed for bacterial identification, 16S rRNA gene sequencing remains one of the most consolidated and widely adopted approaches. While methods such as Matrix-Assisted Desorption/Ionization-Time of Flight Mass Spectrometry (MALDI-TOF MS) have gained prominence for their rapid and cost-effectiveness, 16S sequencing continues to provide more taxonomy resolution.^22,25

While 16S rRNA gene sequencing has revolutionized microbial ecology, there are recognized limitations for precise species-level identification, particularly within closely related taxa like the genus Bacillus and related genus.²⁵ These limitations arise from insufficient sequence diversity within the highly conserved 16S gene region, often leading to ambiguous speciation. This review, however, is specifically focused on the application of 16S rRNA gene sequencing to Gram-negative organisms associated with antibiotic resistance in clinical settings. In this context, this article aimed to review the use of 16S rRNA gene sequencing for identification of pathogenic Gram-negative bacilli included in WHO antimicrobial resistance priority list,⁶ spanning Sanger classical methods to innovative techniques such as next-generation sequencing (NGS). This data tries to elucidate the advantages and limitations of each approach, underscoring the urgent need for methodological standardization to enhance clinical diagnostics, epidemiological surveillance, and infection management caused by these emerging pathogens. By integrating data from standardized databases and international diagnostic guidelines, this work highlights the strengths and limitations of each approach in terms of resolution, reliability, and applicability in clinical microbiology. This comparative perspective aims to delineate the practical diagnostic value of each technique, offering a framework for evidence-based selection of identification strategies.

Methods

This study was conducted as an integrative review of the scientific literature. The following databases were searched: Embase, Web of Science, LILACS, MEDLINE, PubMed, and Scopus. Keywords used included: 16S rRNA, Gram-negative, sequencing, antimicrobial resistance, Enterobacterales, Pseudomonas aeruginosa, Acinetobacter baumannii, whole genome sequencing, MLST, MALDI TOF, carbapenem and cephalosporin. Inclusion criteria included articles that reported diagnostic methods used for identification of Gram-negative bacilli from clinical samples. Articles that were not related to Gram-negative bacilli identification were excluded. This methodological approach was chosen to comprehensively evaluate a broad range of scientific data and provide a broad understanding of 16S rRNA gene sequencing techniques. The primary objective of this review was to identify, analyze, and discuss the principal methodologies available for 16S rRNA gene sequencing, including their respective advantages, disadvantages, and applications in the identification of Gram-negative bacilli pathogenic identification included in WHO antimicrobial resistance priority list,⁶ including keywords, inclusion and exclusion criteria.

Articles unrelated to human infections or identified as duplicates were excluded. Studies meeting inclusion criteria were fully reviewed, Articles not meeting eligibility after full-text review were excluded with documented reasons.

Results and Discussion

16S rRNA Sequencing Methods

Over the years, 16S rRNA gene sequencing methodologies have evolved considerably, differing in terms of read length, cost, accuracy, and application. The 16S rRNA gene can be sequenced using Sanger, NGS or third-generation methodologies.²⁸

Sanger Method

Sanger sequencing, also known as chain termination sequencing, is ideal for the complete sequencing of the 16S rRNA gene, approximately 1,500 base pairs (bp), of a single isolate and has high accuracy.²⁹ Before sequencing, the DNA must first be amplified by PCR and then purified. The choice of primers in the amplification stage depends on the objective: to amplify the complete gene or only specific hypervariable regions (~250–500 bp). The pair of primers 27F (=PA) and 1492R is widely used to amplify the complete gene.³⁰ Partial amplification of the 16S rRNA gene consists of sequencing the hypervariable regions (V1-V9), and is widely used in microbiome studies and NGS sequencing.²³ Some examples of primers used for partial or complete amplification and sequencing of the 16S rRNA gene are shown in Table 1.

Table 1 Examples of Primers Used in the Partial or Complete Amplification and Sequencing of the 16S rRNA gene31–34

Despite its higher cost, full gene sequencing has more advantages over partial sequencing, as it covers all the hypervariable regions of the gene and consequently has a higher taxonomic resolution. Partial gene sequencing may not distinguish phylogenetically close species, such as Escherichia coli vs Shigella spp.²⁸ The complete gene is more reliable for robust phylogenetic analysis.^35,36 The phylogenetic tree can reveal evolutionary relationships between bacterial isolates and helps to discover new species, but the resolution of the 16S rRNA gene may not be enough to separate closely related species, requiring the use of additional genes, such as the housekeeping genes rpoB, gyrB, recA, among others, or whole genome sequencing.²⁵

There are currently standardized and validated commercial kits for partial or complete sequencing of the 16S rRNA gene using the Sanger method. These kits are widely used in clinical, microbiology laboratories, and include the identification of Gram-negative bacilli.^37–39

Thermo Fisher Scientific commercializes the MicroSEQ^TM 500 16S rDNA and MicroSEQ^TM Full Gene 16S rDNA kits (Thermo Fisher Scientific, USA), which are designed for bacterial identification and have their own database for automated identification. The difference between the two kits is the size of the 16S rRNA gene region sequenced and, consequently, the level of taxonomic resolution that each can achieve. According to the manufacturer, the first kit amplifies the V1-V2 region of the bacterial 16S rRNA gene, generating a fragment of approximately 500 bp. This region is widely used for bacterial identification due to its high conservation and variability.

For comparing the 16S gene sequences, the most common public databases used are: 1) the Genbank from the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/nucleotide/),⁴⁰ and 2) EzBioCloud Database (https://www.ezbiocloud.net/).^41,42 The “Guideline for Interpretive Criteria for Identification of Bacteria by DNA Target Sequencing” published by Clinical & Laboratory Standards Institute (CLSI)⁴³ establishes standardized interpretive criteria for the identification of bacteria from clinical cultures, using target regions such as the 16S rRNA. The criteria suggested by the Guideline are shown in Table 2.

Table 2 Evaluation of 16S rRNA Gene Sequencing Analysis for Pathogenic Gram-Negative Bacilli Identification Using BLAST Analysis in NCBI Database According to Clinical & Laboratory Standards Institute Guideline43

For interpretation, when using EzBioCloud, a similarity of ≥98.7% identity and >90% coverage against a reference 16S sequence is considered sufficient for identification at the species level.⁴² An analysis in EzBioCloud Database Update 2025.04.21 (https://www.ezbiocloud.net/), selecting valid names only, and using the first 500 bp and the full gene sequence (~1,500 bp) of the type strain sequence of each bacilli Gram-negative species and genus (for Enterobacterales order) in WHO Priority List⁶ is shown in Table 3. Although the EzBioCloud database criterion for species identification considers a similarity percentage of ≥98.7%, Table 3 also include the CLSI parameters according to the genus/species analyzed.

Table 3 Evaluation of 16S rRNA Gene Sequencing Analysis for Identification of Type Species of Pathogenic Gram-Negative Bacilli Described in World Health Organization Priority List6 Using EzBioCloud Database Update 2025.04.21 and Clinical & Laboratory Standards Institute Guideline43

Alternative Methodologies to 16S rRNA Sequencing

MALDI-TOF MS can be a helpful and faster alternative to 16S rRNA sequencing in bacteria identification. It has already been a reliable tool for clinical microbiology laboratories due to its fast, reliable and effective results. It has reduced in 24 h the time to obtain a microbiological diagnosis in comparison to conventional biochemical automatic systems,⁴⁴ which can make all the difference in patients’ treatment, especially in cases of life-threatening infections or in cases of slow-growing strains.⁴⁵ In the last few years, MALDI-TOF MS has been also used to the rapid detection of antibiotic resistance^46–48 which represents a promising solution for nosocomial infections improvements. However, the effectiveness of this methodology depends directly on the robustness of its database, which is related to the presence of a given pathogen spectra and the ability to differentiate closely related species.

The main MALDI-TOF MS instruments used worldwide are the Vitek^® MS Prime (bio-Mérieux, Marcy l’Etoile, France) and the MALDI Biotyper^® Sirius (Bruker Daltonics GmbH, Bremen, Germany). Despite their database and algorithms differing considerably,^44,49 many studies show that there are no significant differences in the identification of clinically important pathogens when using the two devices.^49,50

The databases of Vitek MS Prime (v. 3.3, 2022) and MALDI Biotyper Sirius (Revision G, 2023) were consulted to assess their robustness in relation to the type species of the genera presented in the WHO priority list (Table 3). When the species is listed in the database, MALDI-TOF MS analysis is a faster and cheaper alternative than 16S for identifying these bacterial species. However, among the 87 type species of the genera analyzed, 24 (27.6%) (Acerihabitans arboris, Apirhabdus apintestini, Biostraticola tofi, Buchnera aphidicola, Chania multitudinisentens, Chimaeribacter arupi, Dryocola boscaweniae, Duffyella gerundensis, Enterobacillus tribolii, Gallaecimonas pentaromativorans, Gibbsiella quercinecans, Huaxiibacter chinensis, Intestinirhabdus alba, Limnobaculum parvum, Mangrovibacter plantisponsor, Phaseolibacter flectens, Prodigiosinella aquatilis, Rosenbergiella nectarea, Saccharobacter fermentatus, Shigella dysenteriae, Silvania hatchlandensis, Symbiopectobacterium purcellii, Tenebrionibacter intestinalis, and Tenebrionicola larvae) are not included in the database of either device (Table 4), showing that its continuous improvement is essential for the identification of clinically important pathogens. In any case, all bacterial species in the ESKAPEE group¹⁸ are included in both databases. However, from the 27 species validity species described in Enterobacter genera, only 8 (29,6%) and 9 (33,3%) are listed in the MALDI Biotyper and VITEK MS Prime databases, respectively, showing that their continuous improvement is essential for the identification of clinically important pathogens. For Cronobacter, an emerging bacterial pathogen associated with infections such as necrotizing enterocolitis, sepsis, and meningitis in neonates and infants,⁵¹ MALDI Biotyper database only identifies the species at genus level, whereas Vitek MS database possesses six of the seven valid species described.¹⁶

Table 4 Genera and Species Listed MALDI Biotyper and Vitek MS Databases

Next-Generation Sequencing Methods

Next-Generation Sequencing (NGS) methods allow millions of DNA fragments to be sequenced simultaneously and have revolutionized microbiome studies by enabling large-scale analysis of bacterial communities using the 16S rRNA gene. Platforms such as Illumina (Illumina Inc, CA, USA) and Ion Torrent (Thermo Fisher Scientific, USA) use short paired-end reads, usually covering partial regions of the gene, such as V3-V4 or V4, ensuring a high depth of sequencing and a reduced cost per sample. These technologies allow the simultaneous detection of thousands of taxa, making them an excellent choice for studies of microbial diversity, ecology and clinical biomarkers, which can be applied to mixed cultures or faecal material. However, because they are limited to 200–500 bp fragments, they can have limited resolution at the species level, as well as being subject to primer bias and the loss of phylogenetic information. Kits such as the “16S Metagenomic Sequencing Library Preparation” (Illumina Inc, CA, USA), which amplify and sequence the V3-V4 regions, are used in amplicon metagenomics for studies of the microbiome of various environments, whether clinical, environmental or industrial. For data processing, bioinformatic tools such as QIIME (Quantitative Insights into Microbial Ecology), a pioneering program for analyzing amplicon sequencing data, mainly the 16S rRNA gene, it was developed to process raw data for microbial diversity analysis, taxonomic classification, and visualization of results. QIIME2 is the successor to QIIME and was developed as a modular platform that analyzes amplicon, metagenome, and transcriptome sequences.⁵² Both are integrated with databases such as SILVA, with cured ribosomal RNA sequences, widely used for taxonomic classification of ASVs/OTUs (amplicon sequence variants/operational taxonomic units) of microorganisms in amplicon-based metagenomics studies.⁵³ Even so, the robustness, scalability, and well-established protocols consolidate NGS methods as the gold standard for large-scale analysis in microbiology. Recent reviews highlight that, despite the growth of long-read technologies, second-generation platforms continue to be widely used for their reliability and cost-effectiveness.^23,54

In recent years, advances in sequencing technologies have revolutionized genomic and metagenomic research, significantly expanding the accuracy and scope of microbial analyses. Currently, the Illumina platform is one of the most widely used, producing short, high-quality reads that are ideal for amplicon-based microbial community profiling and whole genome sequencing, combining high throughput and low cost. More recently, third-generation sequencing technologies, such as PacBio SMRT (Pacific Biosciences Single Molecule Real-Time) and Oxford Nanopore Technologies (ONT), have made it possible to obtain long reads, allowing the assembly of complete genomes, plasmids, and operons without extensive fragmentation of the assembly, overcoming several limitations of NGS, especially in the resolution of repetitive genomic regions and the recovery of complete genomes, better described in the following topic.^55,56

Third-Generation Sequencing Methods

Third-generation sequencing technologies are DNA reading methods that allow individual molecules to be analyzed in real time, without the need for prior large-scale amplification. Among the main platforms are PacBio SMRT and Oxford Nanopore Technologies. In general, both platforms produce long reads, improving taxonomic resolution and offering significant advances in 16S rRNA gene analysis, allowing long reads to be obtained that cover the entire gene region (V1-V9). These technologies provide more accurate taxonomic resolution, especially at the species level, compared to second-generation approaches. Recent studies have shown that in a comparative analysis between Illumina, PacBio and Oxford Nanopore, the latter two platforms offered better resolution at the species level, although they still present challenges related to taxonomic annotation due to limitations in reference databases.^55–57 One of the most important advances in third-generation sequencing is PacBio HIFI (High-Fidelity Reads), which combines long reads (>25 kb) with high accuracy (99.9%), allowing for the complete analysis of complex regions, circumventing the biases intrinsic to amplification-based approaches. This occurs due to the repeated sequencing of the same molecule by DNA polymerase, and multiple reads of the same molecule are combined to generate circular consensus sequences (CCS). Similarly, continuous improvements in ONT’s chemistry and basecalling algorithms have increased accuracy and throughput, as well as making ONT’s long reads more reliable for assembling complete genomes.⁵⁸

In addition to the partial or complete sequencing of the 16S rRNA gene by NGS and third-generation methodologies, respectively, it is fully possible to extract 16S rRNA gene sequences directly from genomes to carry out taxonomic identification and phylogenetic inference analyses. This approach circumvents the limitations associated with sequencing partial regions per amplicon, taking advantage of the complete 16S gene (~1.5 kb) for greater resolution at the species level. Bioinformatics tools such as Contest16S⁵⁹ can automatically locate and extract 16S genes from genome FASTA files, which can then be aligned, compared with reference database, and used to build robust phylogenetic trees, inferring evolutionary relationships between bacteria. This procedure is particularly useful in microbial taxonomy studies, characterization of clinical or environmental isolates, as well as allowing cross-validation with amplicon sequencing data, consolidating bacterial identification in a more complete and reliable way.

The choice of sequencing technology for bacterial identification based on the 16S rRNA gene depends on factors such as accuracy, read length, and data transfer speed. Sanger sequencing offers high-quality reads, ideal for low-yield studies, while next-generation sequencing produces higher yields but generates shorter reads, which can decrease taxonomic resolution. Third-generation platforms, such as Oxford Nanopore and PacBio, produce long reads that increase resolution across the entire 16S gene. Therefore, the selection of a platform should consider the balance between accuracy, detail, and the actual limitations of each sequencing method.^29,56

A recent study compared three sequencing platforms – Illumina MiSeq, PacBio HIFI, and Oxford Nanopore Technologies - to characterize the gut microbiome of rabbits based on the 16S rRNA gene. On the Illumina platform, short reads from the V3–V4 regions of the 16S gene were sequenced, producing many reads but lower taxonomic resolution (48% identification at the species level). PacBio and ONT, which generate long reads, achieved 63% and 76% taxonomic resolution at the species level, respectively. Despite this, the three platforms showed significant differences in microbial composition and diversity, and many species-level classifications were ambiguous (“uncultured_bacterium”).⁵⁵ In another study, a comparative evaluation of 16S rRNA gene sequencing in soil microbiomes was performed using the Illumina (V4 and V3-V4 regions), PacBio (full sequences and truncated V3-V4/V4 regions), and ONT (full sequences) platforms. There was significant variation in detection sensitivity between platforms. PacBio showed a slight advantage in identifying low-abundance taxa, while ONT provided results very similar to those of PacBio, indicating that ONT’s typical sequencing errors have limited impact on the interpretation of well-represented taxa. These results highlight the importance of choosing the appropriate sequencing platform to achieve the desired taxonomic resolution and meet specific research objectives.⁵⁶

Conclusion

The analysis presented in this review underscores the importance of 16S rRNA gene sequencing as a valuable tool for the identification of pathogenic Gram-negative bacilli listed in the WHO antimicrobial resistance priority list. Its conserved and hypervariable regions provide a solid framework for phylogenetic studies and for distinguishing a wide range of clinically relevant taxa. Nevertheless, this method also presents intrinsic limitations, particularly in differentiating closely related species, or in genera with low taxonomic resolution. Furthermore, the accuracy of identification is highly dependent on the choice of primers, sequencing approach, and, critically, the quality and comprehensiveness of reference databases. Thus, while 16S rRNA sequencing remains a cornerstone for bacterial taxonomy and clinical diagnostics, it should be complemented by additional molecular targets or whole genome sequencing to achieve reliable species-level resolution. The integration of these strategies will enhance diagnostic accuracy, infection control, AMR identification, support epidemiological surveillance, and strengthen infection control practices against multidrug-resistant pathogens.