The results of the present study suggest that the 348T variant is associated with an increased risk of developing AP in African Americans. Zhang et al. (2003)13
previously reported a significant association between c.348T>C and AP in Brazilian subjects, but did not observe this relationship in African Americans. The significance of the 348T allele association with AP in African Americans was further validated by comparison of our study’s AP cases with randomly-selected African American subjects sampled by the HapMap project. The 348T/T allele exhibited a strong association with AP case status when our study controls were pooled with HapMap subjects and compared to our AP case group (P
= 0.0007, ). Since the prevalence of AP in African Americans is approximately 2%,19,20
this could explain why a small number of 348T homozygotes were present in the HapMap subject group. The allele frequencies of SNPs observed in our African American control group were reasonably similar to the report of Zhang et al. (2003).13
In disagreement with their findings, SNPs c.568A>T and c.576T>C>G were not significantly associated with AP cases in our African American study population. Four haplotypes (568A.576T, 568A.576C, 568A.576G and 568T.576C) reportedly have a statistically significant association with AP in African Americans. We did not find similar haplotype associations in our African American population, but haplotypes 348T.568A, 348T.576T and 348T.568A.576T were significantly increased in AP cases, whereas haplotypes 348C.568A, 348C.576T and 348C.568A.576T were significantly increased in controls. Disparities between our findings and the previous study13
could be due to differences in selection criteria of the study population. We recruited a relatively homogenous population of African Americans whose ancestral origins are in West Africa or the Bantu territory. Our findings are in agreement with a recent report,14
which found no differences in the frequencies of SNPs c.301G>C, c.546C>A and c.568A>T in African American AP cases and controls.
In our Turkish case population, frequencies of all SNPs were similar to those previously reported.13
The allele frequencies of SNP observed in our Turkish control group were also very similar with the exception of c.306T>C, which was more prevalent in our Turkish population group than previously reported.13
In contrast to previous work, c.306T>C was not associated with AP in our Turkish population.
The African American and Turkish populations recruited for the present study exhibited some differences with respect to frequencies of FPR1
SNPs. The SNP c.306T>C was detected only in Turkish subjects, while the 348T variant was much less common in Turks than in African Americans. These findings are consistent with previous work.13
A recent study of Japanese AP cases and controls detected eleven polymorphisms in the entire FPR1
ORF, including the five SNPs detected in all of our subjects as well as six other SNPs (c.32C>T, c.117C>T, c.289C>A, c.553A>G, c.634G>A and c.1037C>A).15
In Japanese subjects, SNPs c.301G>C and c.546C>A exhibited a significant association with AP, but c.568A>T, c.576T>C>G or c.348T>C were not significantly associated. In the populations screened to date, it is clear that the frequency of FPR1
SNPs can vary widely between races.
While it is unclear how a synonymous SNP like c.348T>C could influence FPR expression, the free energy of the predicted secondary structure of mRNA associated with haplotype 348T.568A was higher than that of 348C.568A (). This was associated with changes in mRNA structure. The significance of this change has not been evaluated, but higher free energy and associated changes in structure could potentially lower mRNA stability and decrease translational efficiency. It is also possible that 348T could be linked to one or more polymorphisms in the FPR promoter region that could influence FPR gene expression. In a recent study in Japanese subjects, a 21.1 kb region including the FPR1
promoter region, 5′ and 3′ flanking areas were screened for polymorphisms. Thirty SNPs were reported, eight of which occurred in the 5′ region. Three of these SNPs (-12915C>T, -10056T>C, -8430A>G) were significantly associated with AP.15
The -12915 T/T genotype was associated with a marked reduction in FPR1
mRNA levels of AP patients compared to those expressing the -12915 C/C genotype.15
A decrease in mRNA levels would be expected to result in reduced receptor expression which would impair PMN chemotaxis toward bacteria, making individuals more susceptible to development of AP. Linkages involving synonymous and non-synonymous coding region polymorphisms have been observed in genes of other receptors.21,22
Studies suggest that approximately 50% of genes may have an association with one or more regulatory SNPs (outside coding regions) and many of them occur in the promoters. Promoter SNPs may alter stability, curvature or flexibility of the promoter and this could affect transcription.23
A significant number of African Americans are affected by aggressive periodontitis.19,20,24
Our study suggests that individuals who are homozygous for 348T in the FPR1
gene could have an increased risk for developing AP. If untreated, AP patients are at a greater risk of premature tooth loss. Early diagnosis and intervention could mitigate the consequences of this disease and help reduce treatment costs for these patients.