While some species prefer specific habitats, others are found in multiple microniches within the oral cavity. According to physical and morphological criteria, the oral cavity can be divided into several major ecological communities, including the palate, dorsum of the tongue, floor of the mouth, buccal epithelium, supragingival plaque, and periodontal pocket. For example, P. gingivalis, P. intermedia, A. actinomycetemcomitans, Spirochetes are able to locate all of these areas (Marsh et al., 2009). A specific example revealed P. melaninogenica not only in pockets and supragingival plaques, but also in the saliva, tonsils, and other mucosal sites (Quirynen et al., 2001). The findings show that intraoral translocation of periodontopathogens inhibits guided tissue regeneration and locally applied antibiotics, thus jeopardizing the outcome of periodontal therapy (Quirynen et al., 2001).

A pure microbial culture is essential for research on its pathogenesis, antibiotic resistance, and invasive potential, to improve the knowledge and therapy of related infectious diseases. Overall, there are four primary elements involved in the growth of bacteria: nutrients, temperature, atmosphere, and incubation time. However, culture should be more pertinently accommodated for particular species in some cases to provide an optimal environment for their growth. Despite prominent advances in bacterial culture over the past decades, several unsatisfactory results remain. On the one hand, there are some limitations to displaying microbial diversity. However, a large number of organisms may not be cultivated by the conventional method because of their association with other microbiota. Instead of isolation, oral bacteria survive and develop in a complicated community called biofilms (Jenkinson, 2011; Soro et al., 2014). To date, almost 50% of oral microbial species detected through culture-independent gene sequencing techniques remain resistant to cultivation (Sizova et al., 2012; Siqueira and Rocas, 2013; Wade, 2013).

Polymerase chain reaction is a revolutionary in vitro replication technique that enables the amplification of targeted genes up to millions in a 2–3 h timeframe. Owing to its high sensitivity/specificity, simplicity of use, and rapid turnaround time, PCR has been comprehensively used in diagnostic microbiology. In recent decades, multiple PCR variations have been developed for various purposes, including nested PCR, asymmetric PCR, qualitative PCR, and reverse transcription-PCR (Sakamoto et al., 2005). For example, Deng et al. (2012) established a novel strategy for rapid colorimetric analysis of Bacillus anthracis by combining asymmetric PCR with gold nanoparticles. To improve the positivity rate of skeletal tuberculosis diagnosis, which is usually performed by histopathology, researchers have investigated the clinical utility of qPCR-based diagnosis of skeletal tuberculosis using formalin-fixed paraffin-embedded tissues. Compared to that of traditional acid-fast bacillus staining, qPCR offers superior accuracy for mycobacterial detection in clinical samples (He et al., 2022). With the advent of the 2019 coronavirus disease (COVID-19) pandemic, RT-PCR is currently considered the gold standard for COVID-19 detection (El-Kafrawy et al., 2021). Importantly, real-time PCR/RT-PCR can determine experimental results without post-PCR procedures, such as gel electrophoresis, minimizing potential cross-contamination. Notably, molecular methods cannot be used to differentiate between dead and living organisms. In addition, PCR is not well suited for identifying novel microbes. In addition, it only detects a single or small number of gene regions/loci of interest at a time, as with multiplex PCR (Lochman et al., 2019). Therefore, high-throughput approaches are required.

As discussed above, culture-dependent methods and PCR-based techniques have limitations in the characterization of bacterial diversity. Owing to developments in sequencing-based approaches, such as 16 s rRNA gene sequencing, this weakness can now be overcome. The 16S rRNA gene is one of the most commonly analyzed marker genes because its functional consistency of 16S rRNA enables it to serve as an evolutionary clock and taxonomic signature in bacterial systematics (Woo et al., 2008). 16S rRNA gene sequencing has been extensively used to determine microbial diversity and for phylogenetic analyses (Janda and Abbott, 2007). PCR-amplified 16S fragments were cloned, sequenced, and clustered based on sequence similarity to infer likely taxonomy. Olson et al. (2011) identified different bacterial signatures in the subgingival plaque of subjects with low or high oral disease by broad-range PCR amplification of the 16S rRNA gene with universal bacterial primers followed by clone-by-clone sequencing. They found that high oral disease exhibited markedly increased bacterial diversity and included an elevated frequency of Clostridiales cluster bacteria, indicating that this atypical bacterial signature is associated with high oral disease (Olson et al., 2011). Another study revealed that anaerobes and oral bacteria are more frequently identified in patients with community-acquired pneumonia than previously thought. These findings strongly suggest that these bacteria may play important in pneumonia (Yamasaki et al., 2013).

Notably, the use of 16S rRNA gene sequencing for bacterial identification can usually be determined accurately at the genus level and not at the species level (Kilian et al., 2016). In addition, this technique is limited to bacteria and archaea, but cannot detect viruses and fungi. Nuclear ribosomal Internal Transcribed Spacer (ITS) regions, including ITS1 and ITS2, are relatively conserved among fungi. Therefore, DNA metabarcoding markers can be used to characterize the diversity and composition of fungal communities (Mbareche et al., 2020).

In contrast to low-throughput molecular techniques, which generally discover a limited number of targets simultaneously, DNA microarrays offer the benefits of broad coverage, speed, and moderate cost. In recent decades, microarrays have become a viable alternative, enabling simultaneous detection and quantification of hundreds or thousands of targets in a single assay. As a three-dimensional array, suspension microarrays have faster hybridization kinetics and more flexibility in array construction than solid-phase arrays. Briefly, the procedure consists of broad-range PCR, probe design, probe immobilization to microspheres/beads, and molecular hybridization.

Wang et al. (2002) constructed a DNA microarray system, designated ViroChip, for virus identification. Later, GreeneChip and MDA microarrays were developed to detect several thousand pathogens, including viruses, bacteria, fungi, and protozoa (Palacios et al., 2007; Gardner et al., 2010). Common features of these platforms are the use of long (60-mer) oligonucleotide probes and random amplification of nucleic acids. Recently, Huang et al. (2013) developed a high-density microarray platform for large-scale vertebrate pathogen discovery based on similar strategies. This EOPM Chip can detect and distinguish all 2,554 known vertebrate virus species, 124 bacterial genera, 38 fungal genera, and 47 parasitic genera at the species or genus level, nearly all of which are known. Despite its lower sensitivity than that of PCR, this system provides sufficient sensitivity to identify the key pathogens responsible for the clinical symptoms. The authors successfully identified adenovirus causing an outbreak of flu-like infections, as well as cardiovirus in a hand-foot-and-mouth juvenile patient in clinical samples, demonstrating its potential for clinical applications. Theoretically, this cutting-edge platform can be applied for the molecular detection of oral microbes.

The Human Oral Microbe Identification Microarray (HOMIM), a 16S rRNA-based mid-density array containing more than 400 probes, was developed and utilized to analyze oral bacteria. To improve our in-depth understanding of the relationship between the oral microbiota and human health, oral diseases such as periodontitis were investigated using the HOMIM technology (Lourenco et al., 2014). HOMIM analysis of salivary microbiota revealed that a total of 16 bacterial species showed differences between pancreatic cancer samples and controls (Farrell et al., 2012). Advances in NGS have resulted in the upgrading of the HOMIM platform by HOMINGS,² covering approximately 600 oral bacterial taxa to date.

More recently, a new phylogenetic DNA microarray (OralArray) containing 22 probe sets targeting bacteria was designed for the systematic analysis of bacterial communities in different microbial ecosystems, including the oral microbiota (Parolin et al., 2017). Unlike other DNA microarrays, whose discriminatory power is based on DNA hybridization, OralArray is based on Ligation Detection Reaction technology associated with Universal Arrays, which possesses excellent discriminative power and good performance in terms of sensitivity. Validation of OralArray on oral samples demonstrated its ability to efficiently detect the most representative bacterial populations colonizing the oral cavity under physiological and pathological conditions.

Continuous improvements in DNA sequencing technology over the past few decades have dramatically improved molecular diagnostics. In contrast to first-generation sequencing, NGS technologies are characterized by “sequencing during synthesis” and “massive parallel sequencing” with the ability to simultaneously and independently sequence millions of nucleic acid fragments. NGS enables direct and comprehensive investigation of microbial communities in a single test and provides deeper insights into the microorganisms involved in human diseases (Jo et al., 2016). In recent decades, NGS has gained wide usage in many scenarios owing to its high-throughput capacity, fast turnaround time, and high cost-effectiveness. For example, high-throughput sequencing of the V3-V4 hypervariable regions of the 16S rRNA gene on the Illumina Miseq PE300 platform revealed clear oral microbial dysbiosis among individuals with severe dental caries of primary Sjögren’s syndrome (pSS). Moreover, Veillonella has been proposed as a potential biomarker in patients with pSS (Zhou et al., 2018). These findings may facilitate our understanding of the link between microbial communities and oral homeostasis and human diseases. To study and clarify whether saliva represents a suitable sample for monitoring supragingival microbiota in healthy children, the 16S rRNA gene amplicon was subjected to NGS using the Illumina Miseq system. Sequence reads were clustered into operational taxonomic units (OTUs) after mapping to bacterial reference databases and representative OTU sequences were used for taxonomic identification and abundance estimation. Compared with supragingival plaque, saliva had a less even and less diverse community. Rothia and Streptococcus were found to discriminate between saliva and plaques. In contrast, supragingival microbiota also exhibited positive associations with salivary microbiota (Shi et al., 2018). It would be interesting to examine the impact of different delivery modes on oral microbiota in healthy infants. 16S rRNA gene-based high-throughput sequencing revealed that newborns delivered via cesarean section and individuals delivered vaginally had distinct oral bacteria. The dominant bacteria in the cesarean section group included Petrimonas, Bacteroides, Desulfovibrio, Pseudomonas, and Staphylococcus. The most abundant genera in the vaginal delivery group were Lactobacillus, Prevotella, and Gardnerella spp. Moreover, the results suggest that repeated sterilization of the vulva might influence infants’ oral microbiota, and attention should be paid to clinical practice (Li et al., 2018b). Similar studies using this technique have been reported previously (Ren et al., 2017; Richards et al., 2017).

Because of the exceptional increase in the number of sequence reads and the reasonable cost, NGS has been widely used to assess the diversity and composition of bacterial ecosystems. However, NGS-based 16S rRNA gene sequencing still focuses only on bacteria and does not cover fungi and viruses. NGS technologies are inferior to Sanger sequencing in terms of the read length and accuracy. In addition, chimera generation during PCR and the intrinsic error rate of sequencing are major concerns. Optimization of the algorithms and experimental procedures is required to solve these problems.

Metagenomic NGS is defined as the process of profiling the entire set of nucleic acids from a given sample that may contain completely different kingdoms of microbes. It sequences and assigns millions of reads to corresponding reference genomes for taxonomic analysis. This powerful technique allows for an unbiased approach to assess the presence of microorganisms and their proportions in a sample. Unlike targeted methods (closed-ended analysis), unbiased hypothesis-free detection (open-ended analysis) is a major advantage of NGS. This technique is sometimes called shotgun sequencing; therefore, it can cover both DNA and RNA derived from bacteria, archaea, viruses, fungi, and parasites, making the diagnosis of mixed infections/co-infections possible. mNGS is akin to “searching for a needle in a haystack” screening technique and theoretically it can detect almost any pathogens in a single run that can never be achieved by traditional approaches (Simner et al., 2018; Gu et al., 2019). During the incipient stage of the COVID-19 pandemic, mNGS offers an accurate and rapid diagnosis compared to previous methods (Chen et al., 2020).

Because mNGS transforms our understanding of the microbial makeup of a mixed sample and the correlation between microbial inhabitants and human diseases, it has been extensively applied to various clinical scenarios (Palacios et al., 2008; Li et al., 2018a; Wang et al., 2019, 2020; Qian et al., 2020; Gu et al., 2021). In addition to depicting microbiota DNA sequences, mNGS can also offer metatranscriptomic data, providing higher-dimensional information. Thus, it can address several issues, such as the microbes present, their functions, and how they interact. Undoubtedly, the integration of the metagenome and metatranscriptome will allow researchers to systematically characterize the complex functions of human ecosystems and vastly expand our awareness of microbe-microbe and microbe-host interactions.

Relatively limited data is available regarding its application in oral microbiota analysis (Zhang Y. et al., 2021). Song et al. (2022) used mNGS to investigate potential pathogens in patients with complex oral mucosal infections and oral leukoplakia. mNGS of oral mucosal tissues identified Candida albicans, human gamma herpesvirus 4, and many other pathogens within 36 h (Song et al., 2022). These findings demonstrate that mNGS is an efficient and sensitive tool capable of simultaneously detecting almost all bacteria, viruses, and fungi potentially associated with disease progression. In addition, mNGS is less affected by antibiotic use and has a higher positive detection rate than conventional detection methods. Metagenomic analysis of dental swabs and plaques from patients with periodontal disease has revealed a core community of disease-associated microbes. Some functional genes and metabolic pathways involved in glycan biosynthesis and bacterial chemotaxis were over-represented in the periodontal disease microbiota. This work enriches the understanding of the oral microbial community structure and metabolic variation that is closely associated with periodontal health (Wang et al., 2013).

Although mNGS can produce in-depth, unbiased information and uncover novel organisms, some issues and obstacles need to be solved before high-complexity technology can become the mainstream and route approach in real settings (Han et al., 2019). First, given the excellent analytical sensitivity of mNGS, cross-contamination during specimen collection, library preparation, assay runs, and bioinformatics classification is a major concern. Quality controls are needed to check “splashome” and “kitome” contamination batch by batch, otherwise, its specificity remains the proverbial elephant in the room. Second, mNGS did not detect all pathogens equally. For example, mycobacteria may be more difficult to detect because routine lysis and extraction are difficult to release from mycobacteria. Consequently, even a few reads of mycobacteria are often considered positive, and other methods are required for confirmation. The clinical interpretation of mNGS reports, distinguishing carrier state/colonization from infection, remains challenging. In fact, mNGS lacks methodological standardization and consensus criteria for positive/negative results, making the interpretation of the findings somewhat subjective. Bioinformatics analysis is sophisticated, even for professionals. Sometimes, mNGS reports need to be reevaluated. It is essential to highlight that mNGS alone does not establish pathogenicity, and that close cooperation between clinicians and technicians/bioinformaticians is required to discuss the positive, inconclusive, and negative results of mNGS. Currently, no mNGS tests have been approved by the Food and Drug Administration. However, it may be worthwhile because some diagnoses are made solely by mNGS, whereas no other tests provide useful information. Taken together, mNGS has potential as a front-line diagnostic tool in the near future; however, more effort and research are needed to improve its clinical utility.

Metagenomics involves WGS where the sequences of entire genomes are randomly sheared and determined by a “shotgun” strategy. WGS can potentially offer species-or even strain-level taxonomic resolution, whereas high-throughput sequencing of the 16S rRNA gene normally profiles the taxonomic composition at the genus or species level in microbiota analyses. Usually, this is preferred term when this technique is used for human whole-genomic analysis.

Owing to technical restrictions, targeting sub-region sequencing for short fragments represents a historical compromise. For example, most of the 16 s rRNA gene sequencing based on NGS only amplifies a portion of the marker gene, providing relatively limited information for bacterial community characterization. The advent of innovative long-read sequencing technologies, including the PacBio and Oxford nanopore platforms, has changed this and opened new avenues for microbiota studies (Jiao et al., 2013; Li et al., 2016). In addition to maintaining the features of high-throughput and in-depth sequencing, the read lengths are up to 1.0–1.5 × 10⁴ and 2–5 × 10³ for PacBio and Oxford nanopore single-molecule sequencing platforms, respectively (Mosher et al., 2014; Laver et al., 2015; van Dijk et al., 2018). However, TGS is capable of routinely generating reads in excess of 1,500 bp and completely covering the full 16S rRNA gene (Franzen et al., 2015; Schloss et al., 2016; Wagner et al., 2016). Previous studies have shown that TGS of the full-length (V1-V9) 16S rRNA gene using the PacBio technique generated sufficient accuracy to resolve subtle nucleotide variations for taxonomic analysis at species or even strain levels (Earl et al., 2018; Johnson et al., 2019). Moreover, TGS has additional advantages over NGS platforms for deciphering complex microbial ecosystems. For example, nanopore sequencing is mediated by the translation of an electrical signal into a sequence of nucleotides and does not involve nucleic acid amplification. In addition, nanopore sequencing can directly identify native base modifications and sequence viral RNA genomes in native status (Quick et al., 2016; Rand et al., 2017; Simpson et al., 2017; Garalde et al., 2018; Liu et al., 2019; Freed et al., 2020). Consequently, this substantially eliminates the biases introduced by PCR or reverse transcription. Many critical human diseases are caused by emerging RNA viruses such as COVID-19, severe acute respiratory syndrome, and Ebola.

To assess the performance of different high-throughput sequencing platforms in analyzing oral microbial communities and their relation to health, saliva samples from five healthy preschool children were collected and subjected to sequencing on the MiSeq (targeting V3-V4) and PacBio RS II (targeting V1-V9) platforms. Although the PacBio system provided a lower amount of clean data, it had a longer read length and higher resolution of nucleotide sequences at the species or strain level than those of NGS (Wang Y. et al., 2017; Zhang et al., 2020). Long-read sequencing of two independent oral S. epidermidis isolates was conducted using the Oxford Nanopore MinION sequencing platform. TGS results revealed that both isolates harbored a novel structural organization. The identification of independent genetic evolution further illustrates the diversity of arginine catabolic mobile element elements in S. epidermidis, which may confer a selective advantage to oral S. epidermidis in dental plaque (McManus et al., 2019).

The main drawback of TGS is its high error rate, which has been significantly improved (Rang et al., 2018). The state-of-the-art methodologies open up more possibility of identifying microorganisms as a second wave of technical advances in the recent decade (Ciuffreda et al., 2021), it should have a wider application in medical practice.

Dental caries, one of the most prevalent oral diseases, usually leads to tooth pain and pulp or periapical infections. This process involves three reversible stages. In the first stage, acidification is mild and infrequent, resulting from a microbial biofilm containing mainly non-mutans Streptococcus and Actinomyces (Takahashi and Nyvad, 2008). This is consistent with the balance of the demineralization/remineralization equilibrium or the conversion of the mineral balance toward net mineral gain. Once a continuous sugar supply is established, acidification changes moderately and frequently, which may improve the capability of non-mutans bacteria to produce acid and resist PH reduction. Therefore, non-mutans streptococci and other aciduric strains may selectively increase (Aas et al., 2005, 2008). These acidogenic and aciduric processes of the microorganism may shift the demineralization/remineralization equilibrium toward net mineral loss, resulting in the progression of dental caries (Becker et al., 2002; Martin et al., 2002; Munson et al., 2004; Mantzourani et al., 2009). Under continuous and prolonged acidic conditions, aciduric bacteria such as mutans streptococci, Lactobacillus, and Actinomyces may become dominant through acid-induced selection (Takahashi and Nyvad, 2011). Environmental acidification is a major factor in phenotypic and genotypic changes in microorganisms during the initiation and development of caries.

Gingivitis is a reversible inflammation caused by the accumulation of bacterial plaque in the gingival tissue. The microbiota involved are commensal microbiota, such as Actinomyces species, F. nucleatum, Prevotella intermedia, Bacteroides, Capnocytophaga, and Eikenella (Huang et al., 2011; Igic et al., 2012).

Gingivitis may develop into periodontitis, resulting from loss of control. Periodontal disease is a chronic irreversible inflammation caused by the destruction of gum tissue, alveolar bone, and tooth loss. P. intermedia, F. nucleatum, Peptostreptococcus micros and Prevotella nigrescens, defined as microbiota of the “orange complex,” as well as the “red complex,” composed of Treponema denticola (T. denticola), Porphyromonas gingivalis (P. gingivalis), and Tannerella forsythia (T. forsythia) are the most commonly associated with periodontitis (Socransky et al., 1998). Although these microbial species are present in low numbers in healthy individuals, they are thought to be responsible for the occurrence and progression of the disease (Kumar et al., 2003). These complex microorganisms fall or disappear (below the detection limit) after efficient treatment.

Moreover, the presence of fungi, protozoa, viruses, and methanogenic archaea is highly correlated with the severity of chronic periodontitis (Horz et al., 2015; Zhu et al., 2015; Lauritano et al., 2016; Jabri et al., 2021).

Oral potentially malignant disorders (OPMDs) are a series of lesions or conditions that may undergo malignant transformation to oral cancer. These diseases are most commonly represented as clinically white patches (oral lichen planus, leukoplakia), red patches (erythroplakia), or red and white patches (erythroleukoplakia). Although the etiopathogenesis of OPMDs is intricate and unclear, contemporary studies have suggested that microbial dysbiosis influences disease development through molecular mechanisms. Significant evidence supports that Candida colonization is regularly correlated with oral leukoplakia (OL), which is called as “Candida leukoplakia,” with infiltration of hyphae in the superficial epithelium (Abdulrahim et al., 2013; Gupta et al., 2019; Weerasekera et al., 2021). C. albicans is the most common Candida species associated with leukoplakia, and its genotype A strains are more frequently associated with OL transformation (Alnuaimi et al., 2015). Over the last decade, increased attention has been paid to microbial infections in the pathogenesis of oral lichen planus. Recent evidence indicates that bacteria are abundant throughout the epithelium and lamina propria of OLP tissues and are positively associated with the levels of infiltrated T cells and CD3⁺, CD4⁺, and CD8⁺ cells. Therefore, microorganisms may play a key role in OLP onset (Choi et al., 2016; Bombeccari et al., 2017; Baek and Choi, 2018; Yu et al., 2020). The microbiota varies among different OLP types, with higher levels of bacteria and lower levels of fungi. Higher enrichment of the fungi Candida and Aspergillus was identified in patients with reticular OLP than in healthy controls. In addition, Alternaria and Sclerotiniaceaeunidentified were significantly more abundant in patients with erosive OLP (Li et al., 2019).

Oral squamous cell carcinoma (OSCC) is the most common malignant neoplasm of the oral cavity, accounting for 90% of head and neck cancers. The 5-year survival rate has remained at 50% within the last few decades due to asymptomatic presentation and poor prognosis. The driving factors vary significantly, including smoking, alcohol intake, and betel nut chewing. Recent studies have suggested that changes in the composition of the oral microbiota may play a role in the induction of oral cancer (Mager et al., 2005; Hooper et al., 2009; Schmidt et al., 2014; Srinivasprasad et al., 2015; Deo and Deshmukh, 2020; Stasiewicz and Karpinski, 2022). P. gingivalis, a critical pathogen in periodontal disease, is a significant trigger agent for oral cancer, and its colonization in tumor tissues is correlated with poor survival in patients with carcinoma (Karmakar et al., 2020; Nwizu et al., 2020; Kavarthapu and Gurumoorthy, 2021; Soder et al., 2021). A study has shown that the microbial communities in OSCC patients at stage 4 became more complicated compared to those in healthy controls. In addition, microbiological populations were significantly altered by the development of cancer from the early stage to the late stage. At the phylum level, the oral samples showed large changes in the abundances of Actinobacteria, Bacteroidetes, and Fusobacteria. The number of Fusobacteria increased significantly during cancer progression in the healthy controls. Although the abundance of dominant genera such as Fusobacteria increased with cancer development, the abundance of Haemophilus, Streptococcus, Actinomyces, and Porphyromonas decreased. Parvimona micra, Haemophilus influenzae, Fusobacterium periodonticum, Streptococcus constellatus, and Filifactor alocis progressively increased with oral cancer development (Yang et al., 2018). Another study revealed that the microbial composition of five genera, Enterococcus, Parvimonas, Peptostreptococcus, Bacillus, and Slackia, showed relatively significant differences between epithelial precursor lesions and cancer tissues. This shift in microbiota might serve as a potential biomarker for monitoring the occurrence and prognosis (Lee et al., 2017). The highly correlated organisms are summarized in Table 2 and Figure 2.

Atherosclerotic disease, which is characterized by the accumulation of lipids and recruitment of T cells into the arterial wall, is considered a chronic inflammatory disorder. Although a number of traditional risk factors, such as diabetes, obesity, high blood pressure, and smoking are widely accepted, oral infection is increasingly manifested in atherosclerotic plaque samples (Fernandes et al., 2014; Aarabi et al., 2015; Slocum et al., 2016; Atarbashi-Moghadam et al., 2018; Herrera et al., 2020). Chhibber-Goel et al. (2016) demonstrated the presence of 23 oral commensal bacteria, either individually or in coexistence, within carotid endarterectomy biopsies from patients undergoing surgical procedures. Of these 23 bacteria, five were unique to coronary plaques, including P. gingivalis, P. endodontalis, P. intermedia, C. rectus, and P. nigrescens. The remaining 18 genera, otherwise confirmed in non-cardiac organs, are responsible for over 30 non-cardiac lesions (Chhibber-Goel et al., 2016).

Adverse pregnancy outcomes are a broad term containing preeclampsia, low birth weight, preterm premature rupture of membranes, early-onset neonatal sepsis, miscarriage, stillbirth, and fetal growth retardation (Figuero et al., 2020). The relationship between specific pathogenic bacteria and APO has been known for a long time (Komine-Aizawa et al., 2019; Bobetsis et al., 2020; Xu and Han, 2022). The prevalent oral species are Filifactor alocis, Fusobacterium nucleatum, Porphyromonas gingivalis, Campylobacter rectus, T. Denticola, and T. Forsythia, among others (Cobb et al., 2017). Stimulated fetal inflammatory and immune responses may ultimately increase the potential for APO owing to the virulent properties assigned to microbiota in the intrauterine environment.

Diabetes mellitus is a metabolic syndrome characterized by hyperglycemia caused by a defect in insulin secretion, decreased insulin action or both (Harreiter and Roden, 2019). Multiple studies have verified a bidirectional relationship between diabetes and periodontitis. Diabetes leads to a shift in connective tissue metabolism and consequently reduces the capacity of inflammation resolution and remodeling, which results in aggravated periodontal impairment. However, the periodontal condition in the host may negatively affect glycemic control in diabetic subjects and exacerbate diabetes progression and its complications (Lalla and Papapanou, 2011; Preshaw et al., 2012; Bascones-Martinez et al., 2015; Nascimento et al., 2018; Sanz et al., 2018; Baeza et al., 2020; Zhang X. et al., 2021). Casarin et al. (2013) reported dramatic differences in the subgingival ecosystem between individuals with and without diabetes. Compared to the levels in non-diabetic individuals, diabetes patients exhibited a higher abundance of Neisseria, TM7, Gemella, Eikenella, Selenomonas, Actinomyces, Capnocytophaga, Fusobacterium, Aggregatibacter, Veillonella, and Streptococcus genera and a lower abundance of Filifactor, Synergistetes, Tannerella, Eubacterium, Porphyromonas, and Treponema genera (Casarin et al., 2013).