Shotgun metagenomic profiling reveals ecological and functional alterations of the oral microbiome in craniosynostosis

Abstract

Objective

To elucidate the microbial drivers underlying of craniosynostosis (CS) , which involves premature suture fusion and secondary dentofacial malformations likely to increase dental disease burden.

Methods

Shotgun metagenomic sequencing of supragingival plaque from 44 participants (22 CS patients and 22 matched healthy controls, aged 6–17 years) were performed, following by bioinformatics evaluation.

Results

Beta diversity demonstrated significant differences between groups ( p < 0.01), whereas alpha diversity trended lower in the CS cohort. Taxonomic profiling revealed a dysbiotic signature in CS with high caries burden, defined by the enrichment of saccharolytic and anaerobic taxa (Scardovia, Actinomyces sp. oral taxon 448, Selenomonas sp. F0473, and Treponema lecithinolyticum)) alongside reduced health-associated genera like Haemophilus and Neisseria. Functional pathway analysis indicated metabolic remodeling, with upregulated fructan biosynthesis and starch degradation III pathways, consistent with caries-active biofilms.

Conclusion

These findings demonstrate that orofacial anomalies in CS favor the assembly of an acidogenic, virulent plaque biofilm. The first shotgun metagenomic profile of the oral microbiome in CS establishes a foundation for future investigations. Furthermore, clinical management of CS should extend beyond structural correction to incorporate microbiological monitoring and preventive strategies, reducing the elevated risk of dental disease in this vulnerable population.

Introduction

Craniosynostosis (CS) is a developmental abnormality characterised by the premature fusion of one or more cranial sutures during infancy [1,2]. CS is broadly classified into non-syndromic (isolated) and syndromic forms. More than 70% of cases are non-syndromic, typically involving early suture fusion without a recognised syndromic diagnosis or major systemic abnormalities. In contrast, syndromic craniosynostosis is associated with multiple congenital anomalies affecting various organ systems and is linked to defined genetic syndromes [3,4].

Premature fusion of cranial sutures in CS results in abnormal craniofacial development, which frequently involves orofacial structures and leads to clinically significant oral health complications [5,6]. Our previous studies in patients with CS demonstrated substantial differences in orofacial characteristics and oral health outcomes compared with the general paediatric population, including cleft palate, orthodontic abnormalities, dental crowding, compromised oral hygiene, malocclusion, and increased prevalence and severity of dental caries [7]. These structural and functional oral abnormalities suggest the presence of a biological environment that may predispose CS patients to alterations in the oral microbial ecosystem.

The oral microbiome comprises diverse microbial communities that play a critical role in maintaining oral health homoeostasis. Disruption of this microbial balance, referred to as dysbiosis, has been associated with both oral and systemic diseases [8,9]. Over the past decade, microbiome research has increasingly focused on elucidating the relationship between microbial communities and human health conditions. In patients with CS, prior studies using 16S rRNA sequencing have reported significant alterations in the gut microbiome, including changes in microbial diversity and increased relative abundance of Enterococcus and Staphylococcus genera [10,11]. However, to date, no studies have characterised the oral microbiome in CS patients, resulting in a lack of microbiome profiles and a limited understanding of their potential relationship with oral health abnormalities in this population.

To address this knowledge gap, we investigated the dental biofilm microbiome of 22 patients with CS and 22 sex- and age-matched healthy controls. Shotgun metagenomic sequencing was employed to provide a comprehensive assessment of microbial taxonomic composition and functional potential, offering greater resolution than conventional 16S rRNA sequencing approaches. This study aimed to characterise oral microbiome alterations associated with CS and to explore their relationship with oral health status.

Material and methods

Subject recruitment and sample collection

This study was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University (IRB No. 513-63) and conducted in accordance with the principles of the 1964 Declaration of Helsinki. All participants with CS were recruited from the Princess Sirindhorn Craniofacial Centre at King Chulalongkorn Memorial Hospital. Written informed consent was obtained from all participants or their legal guardians prior to enrolment. Healthy control (HC) participants were selected based on the absence of dental caries, periodontitis, or systemic health conditions. Exclusion criteria included current smoking, alcohol or drug abuse, and antibiotic use within the preceding three months. A total of 44 participants were enroled in this study, comprising 22 individuals with CS and 22 sex- and age-matched HC, for oral microbiome sampling and analysis. To assess clinical severity, the 22 CS participants were further classified into two categories: non-syndromic CS and syndromic CS. These classifications were subsequently utilised as an ordinal variable to evaluate associations between microbial shifts and disease severity.

Before sampling, participants abstained from food, drink, and oral hygiene for 12 hours. Supragingival biofilm was collected from all tooth surfaces using a sterile explorer and pooled into 1.5 ml tubes. Following collection, samples were immediately transported on dry ice to be cryopreserved at −80 °C.

DNA extraction, library preparation, and shotgun metagenome sequencing

The DNA samples were extracted by using the QIAamp DNA Microbiome Kit (Qiagen, Hilden, Germany) following the manufacturer’s protocols [12]. The extracted DNA was subsequently sequenced using the TruSeq Nano DNA Library Prep Kit and the Illumina NovaSeq system with paired-end reads (Macrogen, Seoul, South Korea), utilising sequencing-by-synthesis technology.

For quality control and host decontamination, raw sequencing reads were processed using KneadData version 0.12.3, which integrates Trimmomatic version 0.39 for adaptor trimming and quality filtering, and Bowtie2 version 2.5.2 for host read removal. Human genomic sequences were identified and removed by mapping reads against the Homo sapiens T2T-CHM13 version 2.0 reference genome using default parameters. This preprocessing workflow ensured the retention of high-quality, non-host reads for downstream metagenomic analysis.

Taxonomic and functional analysis

Oral microbiome taxonomic annotation was conducted by using MetaPhlAn version 4.1.1 with default parameters [13]. Taxonomic profiles were categorised by taxon level, including kingdom, phylum, class, order, family, genus, and species. For microbiome diversity, species-level was used to measure both alpha and beta diversity. Shannon, Simpson, and Observed species indices were selected and visualised for richness and evenness alpha diversity. Beta diversity was assessed using Aitchison distance metrics with multiplicative replacement to handle zero-inflation. Non-metric multidimensional scaling (NMDS) was used to visualise differences in community composition between samples based on the beta-diversity distance matrix. All diversity calculations were conducted with Scikit-bio library version 0.7.0.

For taxonomic differential analysis, taxa that passed the prevalence criteria, defined as being present in more than 25% of samples in at least one group, were included for downstream analysis. Phylum and genus levels were visualised as compositional and relative abundance of the significant taxon. At the species level, the highlighted taxa were selected by statistical testing and a higher effect size with the Cliff’s Delta test by using an in-house script with the NumPy library version 2.3.4. Spearman’s rank correlation analysis was performed to assess associations between the relative abundance of significantly different species and CS-related clinical characteristics and oral health indices.

Functional annotation was performed with HUMAnN version 3.9 to generate the pathway abundance profiles with associated species, which were mapped with the MetaCyc database [14]. The highlighted pathways were selected by post-hoc analysis of statistical significance with false discovery rate correction and their effect size. The effect size of the functional pathways was calculated with a log₂ fold change analysis.

Statistical analysis and visualisation

The statistical analysis and visualisation of this study were performed using Python version 3.12.11. Characteristics and oral health status between groups were compared using an independent t-test and Fisher’s exact test. Alpha diversity significance was calculated with a two-tailed Wilcoxon rank-sum test. Permutational Multivariate Analysis of Variance (PERMANOVA) was performed for beta diversity analysis with 9999 permutations. For differential abundance analysis of taxonomic and functional profiles, a two-tailed Wilcoxon rank-sum test was applied, followed by false discovery rate correction using the Benjamini-Hochberg method (FDR-BH). Statistical significance was defined as p < 0.05 and FDR < 0.25. Statistical analyses were conducted using SciPy library version 1.16.3 and Statsmodels library version 0.14.5. For Cliff’s delta, the effect size was defined as |Cliff’s Delta| $\ge$ 0.147. All data visualisation was performed using the Seaborn library version 0.13.2 and Matplotlib version 3.10.6.

Results

Demographic and clinical characteristics of the study population

A total of 44 participants were enroled in this study, and full-mouth supragingival biofilm samples were collected. The CS group (n = 22) had a mean age of 8.91 ± 2.89 years, while the HC group (n = 22) had a mean age of 9.95 ± 3.43 years (p = 0.2918). Each group consisted of eleven males and eleven females. In the CS group, 12 participants had syndromic CS and ten had non-syndromic CS. The CS group exhibited a variety of craniofacial and dental anomalies, including cleft palate, malocclusion, anterior open bite, and posterior open bite (Figure 1). The mean oral hygiene status for the CS group, assessed using the Simplified Oral Hygiene Index (OHI-S), was 1.89 ± 0.80. Caries risk assessment revealed that half of the CS participants (n = 11, 50.00%) were classified as high risk, while eight (36.36%) and three (13.64%) were classified as moderate and low risk, respectively. Regarding caries prevalence, most CS participants exhibited active caries across multiple indices: DMFT (n = 13, 59.09%), DMFS (n = 13, 59.09%), deft (n = 16, 72.73%), and defs (n = 16, 72.73%), while all participants in the HC group had scores of 0 across all caries prevalence indices. Demographic data of CS and HC groups were shown in Table 1 and Supplementary Table S1.

Figure 1.

Orofacial manifestations in craniosynostosis (CS) and study design. (A) CS is associated with significant craniofacial and dental abnormalities. Structural orofacial anomalies in CS include a prevalence of dental crowding, tooth agenesis, and eruption failure. In addition, individuals with CS exhibit a higher prevalence of dental caries and poorer oral hygiene status compared with matched healthy controls (HC). (B–G) Representative intraoral images of CS participants demonstrating structural abnormalities, including severe dental crowding, dental caries, and eruption failure. Whole mouth supragingival plaque samples were collected from all participants (CS and matched HC) for metagenomic analysis.

Table 1.

Characteristics and oral features of the craniosynostosis patients and healthy controls.

Characteristics	Total	CS (n = 22)	HC (n = 22)	p-value
Age (year), mean + SD	9.43 ± 3.25	8.91 ± 2.89	9.95 ± 3.43	p = 0.2918 a
Gender, n (%)	44 (100%)
Male	22 (50%)	11 (50%)	11 (50%)	p = 1 b
Females	22 (50%)	11 (50%)	11 (50%)
Craniosynostosis types, n (%)	22 (100%)
Syndromic		12 (54.55%)	N/A
Non-syndromic		10 (45.45%)	N/A
Oral hygiene status: OHI-S	22 (100%)	1.89 ± 0.80	N/A
Caries risk assessment, n (%)	22 (100%)
Low risk		3 (13.64%)	N/A
Moderate risk		8 (36.36%)	N/A
High risk		11 (50.00%)	N/A
Carries prevalence	44 (100%)
All caries indices = 0		2 (9.09%)	22 (100%)	p < 0.001 b
All caries indices >0		20 (90.91%)	0 (0%)
DMFT = 0		9 (40.91%)	22 (100%)	p < 0.001 b
DMFT > 0		13 (59.09%)	0 (0%)
DMFS = 0		9 (40.91%)	22 (100%)	p < 0.001 b
DMFS > 0		13 (59.09%)	0 (0%)
deft = 0		6 (27.27%)	22 (100%)	p < 0.001 b
deft > 0		16 (72.73%)	0 (0%)
defs = 0		6 (27.27%)	22 (100%)	p < 0.001 b
deft > 0		16 (72.73%)	0 (0%)

dmft: decayed, missing, and filled primary teeth.

DMFT: decayed, missing, and filled permanent teeth.

dmfs: decayed, missing, and filled surface in primary teeth.

DMFS: decayed, missing, and filled surface in permanent teeth.

CS: Craniosynostosis subjects, HC: Healthy control subjects.

^a independent t-test,

^b Fisher’s exact test

Alterations in oral microbiome diversity and community structure in CS patients

After quality control with KneadData, the average sequence of 37,823,750 reads per samples were used for downstream analysis (Supplementary Table S2). Taxonomic profiling from MetaPhlAn of the oral microbiome identified 13 phyla, 40 classes, 53 orders, 62 families, 118 genera, and 338 species. Following the application of a prevalence filter (presence in ≥25% of samples within at least one group), 276 species were found to be shared between both groups, while 23 and 39 species were unique to the CS and HC groups, respectively (Figure 2A; Supplementary Table S3). Beta diversity analysis revealed a significantly different microbial community structure between the groups (PERMANOVA, p < 0.01; Figure 2B). Regarding alpha diversity, the CS group exhibited lower mean diversity and greater variance across all indices compared to the HC group; however, these differences did not reach statistical significance (Shannon index: 3.66 ± 0.69 vs. 3.89 ± 0.36, p = 0.518; Simpson index: 0.93 ± 0.06 vs. 0.96 ± 0.01, p = 0.245; observed species: 193.68 ± 54.96 vs. 211.59 ± 46.71, p = 0.372; Figure 2C−E). Despite the lack of statistical significance, the lower Shannon and Simpson indices in the CS group suggest a trend toward decreased microbial evenness and increased dominance by specific taxa.

Figure 2.

Species diversity analysis of CS and HC cohorts. (A) Overlap of unique and shared species across groups. (B) Non-metric multidimensional scaling (NMDS) plot of beta diversity using Aitchison distance, showing significant clustering by group (PERMANOVA, p < 0.001). (C-E) Alpha diversity metrics showed no significant differences between groups despite a slight downward trend in the CS group.

Taxonomic shifts in the oral microbiome: enrichment of fusobacteriota and depletion of pseudomonadota in craniosynostosis

At the phylum level, both CS and HC groups were dominated by Firmicutes and Actinomycota, which together comprised a substantial proportion of the oral microbiome in both groups, albeit their relative abundance differed between groups (Figure 3A). In the CS group, Firmicutes (25.77 ± 9.60%) and Actinomycota (18.13 ± 14.20%) were the two most abundant phyla, followed by Candidatus Saccharibacteria (11.50 ± 8.47%). In contrast, Actinomycota (19.97 ± 10.04%) and Firmicutes (18.90 ± 5.79%) were similarly dominant in the HC group, with Pseudomonadota (18.62 ± 10.22%) ranking third.

Figure 3.

Taxonomic distribution and differential abundance at the phylum and genus levels. (A) Relative abundance profiles of dominant phyla (mean >1%). (B) Heatmap representing significantly enriched phyla based on mean relative abundance. (C) Relative abundance profiles of dominant genera (mean > 1%). (D) Heatmap representing significantly enriched genera based on mean relative abundance. Heatmap data were normalised and log10-transformed.

Differential abundance analysis identified three phyla with statistically significant differences between groups (Figure 3B; Supplementary Table S4). Among these, Pseudomonadota was significantly depleted in CS group (p = 0.0124, FDR-BH = 0.0538), whereas Fusobacteriota was significant enriched (p = 0.0013, FDR-BH = 0.0176), both of which represent moderately to highly abundant phyla. Although, Candidatus Gracilibacteria also reached statistical significance (p = 0.0035, FDR = 0.0226), its relative abundance was minimal in both groups, indicating a limited contribution overall community structure.

At the genus level, Actinomyces was the most abundant taxon in both CS (9.59 ± 8.01%) and HC (11.83 ± 7.64%) groups, followed by unclassified Candidatus Saccharibacteria (9.15 ± 7.05% in CS and 8.88 ± 6.31% in HC) The third most abundant genus was Veillonella (7.45 ± 5.50%) in the CS group and Neisseria (8.38 ± 6.31%) in the HC group, as shown in Figure 3C. Differential abundance screening identified thirteen genera with nominal significance between groups (p < 0.05; Figure 3D and Supplementary Table S5). After Benjamini–Hochberg correction for multiple testing, five genera remained significant (FDR-BH < 0.25). Lautropia (FDR-BH = 0.0282, Cliff's Delta = -0.6405), Porphyromonas (FDR-BH = 0.0410, Cliff's Delta = −0.6860), Abiotrophia (FDR-BH = 0.1007, Cliff's Delta = −0.5537), and Haemophilus (FDR-BH = 0.1025, Cliff's Delta = −0.5661) were significantly depleted in the CS group. In contrast, an unclassified genus within the phylum Bacteroidota (GGB1025; FDR-BH = 0.2092, Cliff's Delta = 0.2314) was enriched in CS.

It was also notable that the CS group was characterised by a significant enrichment of the phylum Fusobacteriota, driven primarily by the genera Leptotrichia and Pseudoleptotrichia. Conversely, the HC group exhibited a significantly higher prevalence of the phylum Pseudomonadota, encompassing several constituent genera, including Aggregatibacter, Lautropia, Neisseria, and Haemophilus.

Species-level shifts: depletion of neisseria and haemophilus species in craniosynostosis

Of the 338 species detected, 49 exhibited nominal differences in relative abundance between groups (p < 0.05; Supplementary Table S6). A volcano plot based on effect size (Cliff’s delta) identified 48 differentially abundant species, with 20 enriched in the CS group and 28 enriched in the HC group (Figure 4A). Following correction for multiple testing, 11 species remained statistically significant at a relaxed false discovery rate threshold (FDR-BH < 0.25), a threshold selected to accommodate the exploratory nature of this analysis.

Figure 4.

Species-level differential taxonomic abundance and effect size analysis. (A) Volcano plot illustrating species-level biomarkers identified between groups. The x-axis represents the effect size (Cliff’s delta, | ρ | $\ge$ 0.147) and the y-axis represents statistical significance (-log₁₀(p-value), p < 0.05). (B) Heatmap representing species exhibiting significant differential on average relative abundance based on a False Discovery Rate (FDR-BH) < 0.25. Heatmap data were normalised and log10-transformed.

Among these 11 significant species, eight—including Porphyromonas bobii, Lautropia mirabilis, Streptococcus australis, Haemophilus parainfluenzae, Candidatus Saccharibacteria bacterium oral taxon 955, Streptococcus infantis, Neisseria oralis, and Neisseria mucosa—were significantly depleted in the CS group. Conversely, three species were enriched in the CS group: Actinomyces sp. oral taxon 448, Selenomonas sp. F0473, and Treponema lecithinolyticum (Figure 4B).

Functional metabolic alterations: upregulation of carbohydrate and nitrate related pathways

Functional profiling using HUMAnN annotation identified 404 metabolic pathways based on the MetaCyc database. Of these, 69 pathways exhibited nominal differences in abundance (raw p < 0.05), all of which exhibited FDR-BH values below 0.25 after correction (Supplementary Table S7). Applying more stringent fold-change criteria (| log₂(fold change) | >1), volcano plot analysis highlighted twelve pathways with substantial alterations (Figure 5A), with seven enriched in the CS group and five in the HC group (Figure 5B). In the CS group, the enriched pathways comprised four degradation/utilisation/assimilation pathways (nitrate reduction VI (assimilatory), starch degradation III, L-ascorbate degradation I (bacterial, anaerobic), and chitin deacetylation) and three biosynthesis pathways (fructan biosynthesis, super pathway of pyrimidine deoxyribonucleoside salvage, and starch biosynthesis). In the HC group, the five enriched pathways were functionally diverse: two involved biosynthesis (NAD salvage pathway II (PNC IV cycle) and norspermidine biosynthesis), and one each involved generation of precursor metabolites and energy (glyoxylate cycle), degradation/utilisation/assimilation (assimilatory sulphate reduction I), and macromolecule modification (5-oxo-L-proline metabolism).

Figure 5.

Functional pathway differential abundance and effect size analysis. (A) Volcano plot illustrating species-level biomarkers identified between groups. The x-axis represents the effect size (|Log₂ (Fold change) |>1), and the y-axis represents statistical significance (-log₁₀(FDR-BH), FDR-BH < 0.25). (B) Heatmap representing pathways exhibiting significant differential expression based on FDR-BH < 0.05. The functional pathway was normalised as copy per million (CPM) units. Heatmap data were normalised and log10-transformed.

Taxonomic and functional metabolic association with clinical characteristics in craniosynostosis

To better understand the relationship between the observed microbial shifts, clinical severity, and oral health status, Spearman's rank correlation analysis was performed. Within the CS cohort, we evaluated associations between significantly differentially abundant microbial features (species and functional pathways) and CS-specific clinical characteristics, including disease severity (syndromic vs. non-syndromic type), OHI-S index, and caries risk assessment. Crucially, to help disentangle the microbiological associations related to the craniofacial anomaly from those driven by the secondary dental disease burden, we also assessed correlations between these microbial features and continuous clinical caries indices across all participants (both CS and HC groups).

Overall, 38 species and 53 functional pathways demonstrated at least one significant correlation with these clinical variables (p < 0.05) (Figure 6, Supplementary Table S8-S11). Among the significant bacterial species analysed, Selenomonas sp. F0473 exhibited a very strong positive correlation with syndromic CS (ρ = 0.85, p < 0.001). For OHI-S scores, Veillonella atypica showed a strong positive correlation (ρ = 0.68, p < 0.001), whereas Capnocytophaga ochracea demonstrated a strong negative correlation (ρ = −0.62, p < 0.01).

Figure 6.

Craniosynostosis characteristics correlated with significant species and significant pathways (p < 0.05). All correlations were calculated with Spearman’s correlation with craniosynostosis characteristics, including syndromic types, simplified oral hygiene index (OHI-S), caries risk assessment, and caries indices (DMFT, DMFS, deft, defs).

With respect to caries risk assessment, Streptococcus mutans exhibited a strong positive correlation (ρ = 0.76, p < 0.001), while Eikenella corrodens showed a strong negative correlation (ρ = −0.62, p < 0.01).

Regarding permanent dentition caries indices (DMFT, DMFS, deft, defs) Indices, —Actinomyces sp. oral taxon 448, Veillonella nakazawae, Leptotrichia wadei, and Veillonella sp. 3627—displayed moderate positive correlations with DMFT and DMFS indices (0.4 < ρ < 0.6, p < 0.01).

For primary dentition indices (deft and defs), five species—Actinomyces sp. oral taxon 448, Veillonella nakazawae, Streptococcus mutans, Selenomonas sp. F0473, and Segatella salivae—exhibited moderate positive correlations (0.4 < ρ < 0.6, p < 0.01).

Functional pathways analysis revealed a strong association with CS-related characteristics. With respect to OHI-S scores, the purine ribonucleoside degradation pathway demonstrated a strong positive correlation (ρ = 0.61, p < 0.01). In contrast, four pathways—pyridoxal 5′-phosphate biosynthesis I, pyruvate fermentation to propanoate I, the super pathway of sulphate assimilation and cysteine biosynthesis, and assimilatory sulphate reduction I—exhibited strong negative correlations (−0.6 < ρ < −0.8, p < 0.001).

Caries risk assessment similarly showed a strong positive correlation with the purine ribonucleoside degradation pathway (ρ = 0.61, p < 0.01), along with additional strong positive correlations with pyrimidine deoxyribonucleoside salvage, L-lysine biosynthesis II, and tRNA charging pathways (0.6 < ρ < 0.8, p < 0.01). Conversely, three degradation pathways—L-tryptophan degradation X (mammalian, via tryptamine), serotonin degradation, and aromatic biogenic amine degradation (bacterial)—displayed strong negative correlations (−0.6 < ρ < −0.8, p < 0.001).

For permanent dentition caries indices (DMFT and DMFS), the super pathway of pyrimidine deoxyribonucleoside salvage showed a strong positive correlation, while the super pathway of sulphate assimilation and cysteine biosynthesis and the super pathway of L-methionine biosynthesis (by sulfhydrylation) exhibited strong negative correlations (0.6 < |ρ| < 0.8, p < 0.01) with both indices.

For primary dentition indices (deft and defs), the purine ribonucleoside degradation pathway exhibited a strong positive correlation, whereas serotonin degradation, aromatic biogenic amine degradation (bacterial), and pyruvate fermentation to propanoate I showed strong negative correlations (0.6 < |ρ| < 0.8, p < 0.001).

Discussion

The dentofacial malformations secondary to craniosynostosis, including severe arch crowding and altered respiratory mechanics, create a distinct oral environment that may predispose to a dysbiotic microbiome [7,15]. These structural anomalies may disrupt the homoeostatic flow of saliva and alter oxygen tension, potentially contributing to shifts in the resident microbiota.

This study represents the first comprehensive metagenomic characterisation in the literature of the CS oral ecosystem. By leveraging shotgun sequencing, we circumvent the limitations of 16S rRNA gene profiling to delineate species-level shifts and functional metabolic perturbations [16]. Our findings offer unprecedented insights into microbial architecture associated with CS, establishing a foundational description of microbial composition and function in CS, establishing a baseline for future investigations.

Distinct beta diversity patterns indicate a difference in the overall structure of the oral microbiome in CS patients, with notable characterised by significant inter-individual variability. Although a reduction in alpha diversity was observed, the high intra-group variance suggests that microbial communities in CS are heterogeneous and influenced by multiple factors. Given the clinical spectrum of syndromic and non-syndromic etiologies, anatomical severity may influence the composition of the oral microenvironment. The enrichment of specific taxa suggests a shift in community balance; however, due to the high caries burden in our CS cohort, it is difficult to determine whether these differences are attributable to craniofacial structure or to dental disease severity.

The CS cohort presented a pervasive cariogenic phenotype, with 91% of patients (20/22) exhibiting dental caries. This clinical state was mirrored by a significant taxonomic enrichment of saccharolytic genera, including Leptotrichia, Prevotella, Scardovia, and Pseudoleptotrichia [17–20]. Notably, the dominance of Leptotrichia, a recognised sentinel taxon in cariogenesis, suggests a shift toward an acidogenic ecological state [21,22]. The metabolic profile of these taxa, characterised by the fermentation of dietary carbohydrates into lactic acid, facilitates a sustained proton gradient that drives enamel demineralisation [23,24]. These findings are consistent with the ecological plaque hypothesis, in which depletion of core commensals such as Corynebacterium, Haemophilus, and Neisseria accompanies progression toward a caries-active state [25–27]. Our observation of significant enrichment in Scardovia and Streptococcus mutans (p = 0.048, Log2FC = 2.62) further reinforces this model. However, given the markedly higher prevalence of active caries in the CS group, these microbial patterns likely reflect disease status and plaque accumulation rather than craniosynostosis itself.

A defining feature of the CS microbial landscape was the disproportionate enrichment of Actinomyces sp. oral taxon 448. This species was positively associated with multiple caries indices in our cohort. These data align with previous characterisations of Actinomyces as a pathognomonic marker for aggressive dental pathologies, including active root caries [28] and early childhood caries [29,30]. Its enrichment is likely to reflect plaque retention and high caries activity within the CS group. Selenomonas sp. F0473 emerged as a dominant member of the CS microbiome, showing a potent association with syndromic CS. While functionally under-characterised, its strong correlation with all caries indices suggests it may function as a syndromic-specific pathobiont. Given that related taxa, such as Selenomonas sputigena, are acid-tolerant anaerobes that synergise with Streptococcus mutans to accelerate biofilm maturation [31], Selenomonas sp. F0473 likely exploits the anaerobic stagnation zones created by severe orofacial defects related to syndromic conditions. We posit that the unique topographical constraints of the CS oral cavity facilitate the expansion of this taxon, which in turn amplifies the acidogenic potential of the biofilm, driving the high caries burden observed in these patients. This said whether its enrichment is driven by craniofacial morphology or by the high burden of dental disease remains uncertain.

Despite its high relative abundance in the CS cohort, Treponema lecithinolyticum presented a noteworthy divergence from the prevailing cariogenic signature, exhibiting no correlation with any dental caries index This lack of association with caries suggests a potentially different ecological role within the CS oral environment [32,33]. Given its established associations with progressive periodontitis and endodontic infections, we suggest that its prevalence in CS may reflect an underlying predisposition to periodontal dysbiosis or localised inflammatory states.

Metagenomic functional profiling revealed that the nitrate reduction VI pathway and fructan biosynthesis exhibited the most pronounced differential enrichment within the CS cohort. While nitrate reduction is traditionally viewed as a hallmark of oral eubiosis, its upregulation in the CS group, despite the depletion of classical nitrate-reducing commensals like Neisseria and Haemophilus [34–37], suggests a compensatory metabolic shift. This functional enrichment may be driven by the CS-associated taxa, such as Actinomyces, acting as a homoeostatic buffer against the dominant acidogenic pathways [36,37]. By facilitating the conversion of nitrate into alkaline nitrogenous byproducts, this pathway serves as a critical pH-stabilising module, potentially mitigating the enamel demineralisation driven by lactic acid accumulation. Consequently, the metabolic landscape of the CS microbiome appears to be defined by a delicate competition between saccharolytic acidification and nitrergic buffering responses [36,37].

Parallel to nitrate metabolism, the enrichment of fructan biosynthesis underscores a fundamental shift toward a virulent biofilm architecture in the CS cohort. These extracellular polysaccharides (EPS) serve as the primary polymeric scaffold, promoting microbial adhesion and stabilising the structural integrity of the cariogenic biofilm [38,39]. Species-resolved functional attribution identified Actinomyces sp. oral taxon 448, S. mutans, and S. salivarius as the primary drivers of this matrix assembly (Supplementary Table S12). Furthermore, the enrichment of the starch degradation III pathway, also linked to Actinomyces sp. oral taxon 448, suggests a metabolic convergence with S. mutans [40–42]. By mobilising complex carbohydrates into fermentable monomers, this pathway not only fuels acidogenesis but also provides the essential substrates for EPS synthesis and virulent gene modulation. Together, these functional findings are consistent with a cariogenic metabolic environment, though they likely mirror the high disease burden observed clinically.

Correlation analysis revealed a robust positive association between S. mutans abundance and high-risk scores in the caries risk assessment, underscoring the prognostic utility of these frameworks in identifying cariogenic trajectories within the CS population. Reflecting this heightened risk, the CS cohort exhibited significantly elevated OHI-S scores [7], indicating a substantial biofilm burden. Notably, these hygiene indices positively correlated with the prevalence of Veillonella atypica, a species recognised as a critical ecological bridge in dental plaque. V. atypica facilitates biofilm maturation by participating in metabolic cross-feeding, specifically by utilising lactic acid produced by primary colonisers like Streptococcus [43–45]. These findings further support that increased plaque accumulation in the CS group likely contributes substantially to the observed microbial configuration.

Several limitations of our study warrant consideration. First, recruitment from a single tertiary centre may limit generalisability. Second, the cross-sectional design precludes the establishment of causal inference between the microbial shifts and the progression of CS. Importantly, the markedly higher prevalence of dental caries and poorer oral hygiene in the CS cohort represents a major confounding factor. This makes it difficult to distinguish the primary effects of CS from secondary, disease-related microbial shifts; future studies utilising caries-matched designs are necessary to strengthen causal interpretations. Furthermore, OHI-S and caries risk assessment were performed only in the CS group, and correlation analyses with these characteristics were therefore limited to CS patients, which may introduce bias in the interpretation of these associations. Although our sample size is constrained, a perennial challenge in orphan disease research, this study represents a pioneering metagenomic baseline for this population. Because of this limited sample size, our analysis relied on robust, non-parametric rank-based statistical testing; future studies with larger cohorts would benefit from applying compositional data-aware models (such as MaAsLin2) to further validate these microbial shifts. Furthermore, the inherent clinical and environmental heterogeneity, encompassing varying syndromic phenotypes and dietary habits, likely contributed to the observed inter-individual taxonomic dispersion.

Conclusion

This study, for the first time reports, a significant association between craniosynostosis and a distinct oral microbiome. The CS with high caries burden cohort demonstrated enrichment of taxa such as Actinomyces sp. oral taxon 448 and Selenomonas sp. F0473, alongside functional pathways consistent with carbohydrate fermentation and biofilm maturation. Furthermore, high-resolution functional profiling reveals a pro-cariogenic metabolic architecture, providing a mechanistic basis for the heightened susceptibility to dental decay observed in this population. Collectively, these results emphasise the importance of proactive oral health management in CS patients and provide a framework for future longitudinal studies designed to disentangle craniofacial structural influences from secondary effects of dental disease.

Supplementary Material

Supplementary Material

Supplementary_Table_CS_Final.docx

Funding Statement

This research was funded by Thailand Science Research and Innovation Fund Chulalongkorn University (HEA_FF_69_036_3200_003). Health Systems Research Institute (69-143, 69-145). the Second Century Fund (C2F), Chulalongkorn University. the 90th Anniversary of Chulalongkorn University Scholarship under the Ratchadapisek Somphot Endowment Fund (GCUGR112569113D). Chulalongkorn University, second century high potential professoriate.

Disclosure statement

The authors declare no competing interests.

Data availability statement

All oral microbiome sequences were deposited in the National Centre for Biotechnology Information by Sequence Read Archive (SRA) in BioProject number PRJNA1394340 and linked to http://www.ncbi.nlm.nih.gov/bioproject/1394340.

Supplementary material

Supplemental data for this article can be accessed at https://doi.org/10.1080/20002297.2026.2687219.