rs198968 — KLK4

Tooth enamel begins life as a soft, protein-rich matrix laid down by specialized cells called ameloblasts. The hard mineral — hydroxyapatite — crystallizes within this scaffold, but the scaffold itself must be completely removed before the crystals can interlock and form the hardest biological tissue in the human body. The enzyme responsible for dismantling that scaffold in the final stage of enamel development is KLK4¹¹ KLK4
Kallikrein-related peptidase 4, also known historically as enamel matrix serine proteinase-1 (EMSP1), is a chymotrypsin-like serine protease expressed exclusively during the maturation stage of amelogenesis.

Without KLK4, enamel forms with apparently normal thickness but the protein scaffold remains trapped between the mineral crystallites. The enamel looks normal at first, but the residual protein physically blocks crystal interlocking — the teeth are soft, yellow-brown, and prone to chipping. This severe phenotype occurs when both copies of KLK4 carry loss-of-function mutations (autosomal recessive amelogenesis imperfecta, OMIM 603767). The rs198968 variant is a common intronic polymorphism that does not destroy KLK4 function but appears to modulate how much enzyme is produced during the narrow developmental window when enamel hardens.

rs198968 lies within intron 1 of KLK4 at chromosome 19q13.33 (GRCh38 position 50,910,072), in a region annotated as a weak promoter in embryonic stem cell chromatin mapping. The variant is non-coding, so it does not change the KLK4 protein sequence. Instead, its effect appears to operate through regulatory mechanisms: functional assays from a molecular evolution study found that variants at and near this position reduce KLK4 transcriptional output.

The rs198968 locus sits within a larger 70-kb region showing signatures of positive selection in East Asian populations²² positive selection in East Asian populations
High FST (population differentiation) values and extended haplotype homozygosity indicate that a derived haplotype at this locus spread rapidly in East Asian ancestry groups, suggesting an adaptive advantage. The selective sweep encompasses rs198968 and two neighboring variants (rs1654556 in the 3′-UTR and rs17800874 in a nearby enhancer). Together, this haplotype reduces KLK4 expression synergistically. The evolutionary significance is not fully understood but may relate to enamel crystallite morphology or epidermal phenotypes also regulated by KLK4.

For the A/G polymorphism studied in pediatric dental caries research, the functional significance of the specific G versus A allele has not been independently determined at the molecular level. The association with caries susceptibility is therefore observational — a signal from population studies — rather than mechanistically characterized from first principles.

The largest study of rs198968 in European children comes from the Czech ELSPAC cohort³³ ELSPAC cohort
European Longitudinal Study of Pregnancy and Childhood, a multi-country birth cohort established in the early 1990s to track child health outcomes (Bečvářová et al., Clin Oral Investig, 2022), which enrolled 761 children with genotyping across 15 SNPs in enamel formation genes. For primary dentition (baby teeth), children with the GG genotype were substantially more likely to have severe dental caries (dmft ≥ 10): OR 2.28 (95% CI 1.11–4.69, p=0.019). The GG genotype appeared in 70.5% of high-caries cases compared to 51.1% of caries-free controls. No statistically significant association was found for permanent dentition, suggesting the window of susceptibility is primary tooth formation.

A contrasting result was reported from a Turkish early childhood caries study (Çolak et al., 2015; n=259 children aged 2–5), which found the opposite pattern: in multivariate analysis controlling for diet and hygiene, AG and GG genotypes were protective⁴⁴ AG and GG genotypes were protective
OR 0.15 and 0.17 respectively compared to AA as reference; the wide confidence intervals (0.03–0.89 and 0.03–0.92) and small sample size indicate substantial uncertainty in this finding. The discrepancy between these studies is notable: different reference genotypes, different environmental contexts (fluoridation, diet), and different populations likely all contribute. The Czech ELSPAC study's larger size and European-ancestry focus is more directly relevant to a European population-based recommendation.

A Polish children study (Boratyński et al., 2017; n=96 aged 20–42 months) also found rs198968 significantly associated with caries incidence (p=0.0069), corroborating the Czech direction. Taken together, the balance of evidence from European pediatric cohorts supports G as the risk-associated allele for primary dentition caries, with moderate overall confidence.

KLK4 activity is limited to the enamel maturation window — a period entirely completed before the primary teeth erupt into the mouth. By the time a child is born, the enamel quality of their primary teeth is already set. This means the actionable window for primary teeth is essentially prenatal: adequate maternal calcium and vitamin D intake during pregnancy support optimal enamel mineralization in the developing fetus.

Once enamel is formed — whether it is structurally optimal or subtly compromised — the protective strategies shift to maximizing remineralization and minimizing demineralization. Topical fluoride is the most evidence-backed intervention: fluoride substitutes for hydroxyl groups in hydroxyapatite to form fluorapatite, which is substantially more acid-resistant. For a child with genetically susceptible enamel, fluoride varnish applications in the first years of life, combined with appropriate-strength fluoride toothpaste from the first tooth, provide the strongest available protection during the critical primary dentition period.

rs198968 has been co-studied with rs17878486 (AMELX), rs2242670, rs2235091, and rs2978642 (also in KLK4) in enamel gene panels. The Czech ELSPAC study found that AMELX rs17878486 and KLK4 rs198968 were the primary drivers in primary dentition caries risk, with rs2242670 also showing association in permanent dentition. This convergent signal across AMELX (enamel scaffold protein) and KLK4 (enamel scaffold degrading enzyme) supports a polygenic model in which both the secretory and maturation phases of enamel formation carry independent heritable risk.

No formal gene-gene compound action has been characterized for rs198968 × rs17878486, but the joint biology is coherent: subtly reduced KLK4 expression (rs198968) combined with altered amelogenin isoform ratios (rs17878486) would both impair the coordination between protein scaffold assembly and removal — compounding the enamel quality deficit.