Dental caries remains one of the most prevalent chronic diseases of childhood, and its true nature has been radically revised over the past two decades. Early research blamed a single villain, Streptococcus mutans, but shotgun metagenomic sequencing has since revealed that tooth decay is fundamentally a disease of microbial dysbiosis, an ecological imbalance in which acid-producing genera such as Veillonella, Bifidobacterium, Lactobacillus, and Actinomyces become enriched in carious lesions. At the functional level, the caries-associated microbiome upregulates genes for mixed-acid fermentation, while healthy mouths tend to express genes involved in antimicrobial peptide synthesis. Because caries is ultimately driven by the net metabolic output of a community rather than by any one species, interventions that can steer that community’s metabolism, rather than sterilize it, have become a major public health goal.

The research team, based at institutions including Shenzhen Futian District Maternal and Child Health Hospital and Southern University of Science and Technology, recruited eleven pediatric patients diagnosed with dental caries under strict criteria: no oral cleaning on the sampling day, no systemic disease, no antibiotics within three months, and no history of fluoride use. Saliva samples were collected with parental consent and ethical approval, then incubated at body temperature with hesperetin or naringenin for twenty-four hours. The researchers then extracted total DNA using a magnetic bead-based kit and performed shotgun metagenomic sequencing on the DNBSEQ-T7 platform, generating high-quality data with Q20 percentages above ninety-eight percent and Q30 values predominantly above ninety-four percent.

The bioinformatics pipeline was exhaustive. Raw reads were trimmed and filtered with Trimmomatic, host sequences were removed with KneadData, and de novo assembly was performed with MEGAHIT. Gene prediction used Prodigal, and non-redundant gene catalogs were constructed with CD-HIT, yielding more than 853,000 sequences for one sample group and more than 626,000 for another. Functional annotation was carried out against seven databases, including GO for broad functional categories, KEGG for metabolic pathways, EggNOG for orthologous groups, CAZy for carbohydrate-active enzymes, CARD for antibiotic resistance, and VFDB for virulence factors. Taxonomic assignment relied on DIAMOND alignment followed by lowest common ancestor classification in MEGAN. Reads mapping rates exceeded ninety percent for most samples, confirming that the assembled contigs reliably represented the dominant fraction of each sequenced community.

The headline result was the sheer scale of the perturbation. Across all eleven sample groups, roughly two in five annotated genes changed in abundance by more than seventy percent following flavonoid exposure. In most groups, hesperetin affected a numerically larger share of genes than naringenin, for example forty-three percent versus thirty-nine percent in one sample and forty-two percent versus thirty-nine percent in another, hinting that hesperetin may exert the broader genomic footprint. The authors are careful, however, to frame this as a descriptive observation rather than proof of intrinsically different potency, given the small cohort and the absence of formal statistical testing. What matters conceptually is that a natural compound, at a single physiologically relevant concentration, engaged the microbial community at a systems level.

Among the most biologically significant findings were shifts in the carbohydrate-active enzyme landscape. Glycosyltransferases, which build the extracellular polysaccharides of dental biofilm, and glycoside hydrolases, which break down dietary sugars, showed increased relative abundance in multiple hesperetin-treated samples. This aligns with the known role of flavonoids as modulators of carbohydrate metabolism and biofilm matrix synthesis in oral streptococci. The researchers caution that metagenomic sequencing measures gene abundance, not transcriptional activity, so whether enzymatic output actually changed remains an open question requiring metatranscriptomic or metabolomic follow-up. KEGG and EggNOG annotations likewise revealed treatment-associated shifts in metabolic pathway representation, a perspective the authors consider critical because a taxonomic shift is only clinically meaningful insofar as it alters the community’s net metabolic behavior.

Perhaps the most intriguing observation emerged from the beta diversity analyses. Principal component analysis after centered log-ratio transformation, principal coordinate analysis based on Bray-Curtis and Binary-Jaccard distances, and non-metric multidimensional scaling all showed that hesperetin-treated and naringenin-treated samples occupied distinct regions of compositional space, both taxonomically and functionally. On the functional axis, the first principal component explained between 50.74 and 61.11 percent of the variance, and NMDS stress values as low as 0.0653 confirmed reliable ordinations. The two flavonoids differ chemically by only a single hydroxyl group, yet they appeared to push the microbiome in subtly different ecological directions, a structure-activity relationship the authors plan to make the centerpiece of their follow-up mechanistic work.

Crucially, the restructuring did not resemble a scorched-earth antibiotic effect. Bacteria remained dominant in all samples, and the Shannon diversity index showed bidirectional, sample-specific changes, decreasing in one child’s community upon flavonoid exposure while increasing in another’s. This preservation of overall richness, combined with clear separation in beta diversity space, suggests the compounds may selectively restructure the community, potentially suppressing cariogenic taxa or disrupting the synergistic metabolic networks that drive enamel demineralization while sparing health-associated commensals. The authors stress that this interpretation remains speculative, but the ecological selectivity is precisely what caries prevention demands, since the goal is to restore microbial homeostasis rather than sterilize the oral cavity.

The pediatric focus adds public health weight to the findings. The developing oral microbiome of children is a dynamic, malleable ecosystem, and early childhood caries is increasingly recognized as a sentinel event that predicts long-term oral health trajectories and has been linked to systemic conditions including malnutrition, growth impairment, and poor academic performance. If dietary flavonoids can be shown in controlled studies to guide the developing microbiome toward a health-associated state, whether through dietary enrichment, functional foods, or topical formulations, the implications for early-life prevention could be substantial. The authors explicitly note that no gut microbiota data were generated, so no conclusions about the gut can be drawn from this work.

The limitations are equally clear and the authors enumerate them candidly. Eleven subjects cannot capture the enormous inter-individual variability of the oral microbiome, the ex vivo batch-culture system cannot reproduce salivary flow, diet, host immunity, or mature three-dimensional biofilm architecture, and a single concentration at a single time point precludes any statement about dose-response or temporal dynamics. Ex vivo incubation may also stochastically lose or retain low-abundance taxa, potentially confounding the observed shifts, and strain-level resolution for key cariogenic species was not achieved. The team frames the entire study as hypothesis-generating, with raw sequencing data deposited in the NCBI Sequence Read Archive under BioProject PRJNA1379526 to serve as a metagenomic baseline. Even within those constraints, the message is striking: hesperetin and naringenin, the everyday polyphenols of oranges and grapefruits, are not nutritionally passive. Under laboratory conditions they behave as ecological modulators, capable of broadly remodeling the taxonomic and functional architecture of the caries-associated oral metagenome, and they may represent the opening chapter of a genuinely ecological approach to preventing the world’s most common chronic childhood disease.

Subject of Research: Metagenomic analysis of the effects of the citrus flavonoids hesperetin and naringenin on the oral microbiome of children with dental caries

Article Title: Citrus flavonoids hesperetin and naringenin remodel the caries-associated oral microbiome: An exploratory metagenomic study

Article References: Zhu, L., Zhu, Y., Zhang, Z., Hu, S., Li, X., Zhou, Y., Shi, Z., Huang, D., Huang, R., Liu, Z., Wang, Y., Ma, F., & Tang, B. (2026). Citrus flavonoids hesperetin and naringenin remodel the caries-associated oral microbiome: An exploratory metagenomic study. Journal of Agriculture and Food Research, 31, Article 103320. https://doi.org/10.1016/j.jafr.2026.103320

Image Credits: AI Generated

DOI: 10.1016/j.jafr.2026.103320

Keywords: dental caries, oral microbiome, hesperetin, naringenin, citrus flavonoids, metagenomics, pediatric health, microbial dysbiosis, carbohydrate-active enzymes, beta diversity, natural compounds, biofilm