The main findings in the present study were the significant association between the G-allele of the IL-18 +183 A/G polymorphism and lower circulating IL-18 levels, which specifically was present in patients with T2DM and MetS. The frequency of the +183 AA genotype was higher in CAD patients with HT, but without any relation to CAD itself.
The +183 G-alleles' influence on IL-18 levels shown in our relatively large study of stable CAD patients is in accordance with results from one previous study in a similar population . The polymorphisms' functional mechanism has not been clarified. As the location of the +183 A/G polymorphism is in the 3'UTR of the gene, an interference with mRNA stability or the translational process is suggested, as well as an interaction with the 5' end in regulating the transcriptional activity. Although no difference in allelic degradation of mRNA and no difference in expression of an upstream reporter gene were found in functional experiments , these possibilities cannot be excluded. The +183 A/G polymorphism is reported to be in complete linkage disequilibrium (LD) with the 5'UTR -105C/T polymorphism, both being parts of 6 major haplotypes defined by 5 polymorphisms in the IL-18 gene . It was shown that the haplotype carrying the +183 G-allele was the one associated with the lowest IL-18 levels , supporting the polymorphisms' functional role in lowering IL-18 levels.
The +183 polymorphism modified slightly the expression of the IL-18 gene in circulating leukocytes, with higher mRNA levels related to the G-allele. This result seem to present a discrepancy to the lower IL-18 circulating levels, and also to previous gene-expression results from lymphoblastoid cell-lines . The observed 1.13-fold increase may not be of biological significance. Further, circulating leukocytes may not be the main source contributing to circulating IL-18, although IL-18 mRNA was shown to be constitutively expressed in human peripheral blood mononuclear cells due to lack of a destabilizing sequence in the 3'UTR of the IL-18 gene . A post-translational impact of the inflammasome and Caspase-1, in the cleavage of pro IL-18 to its active secreted form, may contribute to the inconsistency. The results may, nevertheless, also indicate a compensatory reaction from the circulating inflammatory cells.
We observed no influence of the -137 G/C and -607 C/A polymorphisms on circulating IL-18 levels or on IL-18 gene expression. These two promoter polymorphisms have previously been associated with lower transcriptional activity in in vitro studies [18, 23, 24]. However, in a human population no association was observed for the -137 G/C variant .
Significantly higher IL-18 levels were found in male patients as compared to females. This gender difference is to our knowledge novel information. In a recent meta-analysis, in which nine of twelve studies included both gender, the association between IL-18 and CVD events was generalized to both gender . However, an increased IL-18 related risk of CV events in men warrant further investigation. Interestingly, an experimental study performed in mice indicated that increased endogenous IL-18 production reduced survival only in the male animals .
Increased IL-18 levels in patients withT2DM and MetS have previously been reported, and our observations support these results [9, 11, 27–29]. The importance of IL-18 in MetS was also underlined in a review by Trøseid et al. . It was also recently shown increased IL-18 mRNA expression in adipose tissue from MetS subjects compared to non-MetS individuals . We could further demonstrate that the influence of the +183 A/G polymorphism was especially apparent in the T2DM and Mets groups, suggesting a beneficial influence of the G-allele in these patients. G-allele carriers have also been reported to have reduced frequency of MetS and the subsequent IL-18 levels were suggested to be of genetic origin . Although the frequency of the +183 G-allele was not differently distributed in MetS and T2DM patients as compared to the unaffected patients in the present study, an association is still feasible.
The observed risk of having HT was significantly affected by the presence of the combined genotypes of the three polymorphisms investigated (+183 A/G, -137 G/C and -607 C/A). The +183 GG genotype (in combination with -137 GG and -607 CC) was associated with a 50% lower HT risk, whereas the +183 AA, in combination with -137 CC and -607 AA, was associated with a 70% higher risk. It should be emphasized that our stable CAD patients with HT did not present higher IL-18 levels, which might be due to their treatment regimen. Nevertheless, as the combined genotypes in the total population were associated with lower and higher circulating IL-18 levels, respectively, a related mechanism is plausible.
An inflammatory state accompanied by the A-allele of the +183 polymorphism may consequently induce a cumulative increased risk for HT. Genotypes are fixed during life-time and a pro-inflammatory state initiated early in life may consequently lead to development of HT over time in these genotype-specific patients. We have previously shown that arterial stiffness was associated with both elevated levels of IL-18 and elevated systolic blood pressure in MetS patients . Chronic elevated levels of IL-18 may lead to a persistently increased expression of IL-18 inducible cytokines downstream of the NF-κB/AP-1 signal pathway, i.e. IFNγ and MMPs [34, 35], which are of importance for an inflammatory state in atherosclerosis in general and for arterial remodeling, important in HT. HT has on the other hand been shown to induce an upregulation of IL-18 through β2-adrenergic receptor stimulation in endothelial cells. Thus, a positive feed-back mechanism has been suggested .
In the attempt to sort out the role of IL-18 BP in regulating IL-18 levels, our results were not clarifying. The levels of IL-18 and IL-18 BP were significantly correlated, and as the expression of IL-18 BP has been shown to be markedly up-regulated by IFNγ, a negative feedback mechanism on IL-18 activity has been suggested . The interrelated values may also be a consequence of the methodology used in measuring the two markers. The results indicate that the utilized IL-18 assay also measures IL-18 in complex with IL-18 BP and not only the free IL-18 molecule, also stated by others [38, 39]. As the manufacturer cannot clarify what the assay exactly measures, no valid conclusion can be drawn.
The control group is small and seven years younger than the patient population, which may limit the statistical value of these results. The controls were, however, only included in the gene-analysis. As the +183 A/G polymorphism deviates from HWE in controls, probably due to small sample size, the observed related frequencies should be carefully interpreted.
In summary, the +183 A/G polymorphism induced lower IL-18 levels consistently and significantly, and has conceivably a functional role in regulating the transcriptional and/or the translational process. The beneficial +183 GG genotype was especially apparent in relation to IL-18 levels in MetS and T2DM patients. The modified risk for HT according to the different genotypes indicates an importance of this gene in HT patients, possibly through IL-18 levels. As HT, MetS and T2DM are associated with worse prognosis in CVD, the regulatory mechanisms of IL-18, especially through the +183 A/G polymorphism should be further explored.