In a Caucasian population residing in New Zealand we evaluated 29 SNPs in seven genes that might play a direct or indirect role in CD pathogenesis. Some of the SNPs associated with CD in this cohort have been reported elsewhere [23
], whilst others do not appear to have been reported. Here we focus on the gene-serum selenium interaction on CD.
Low serum selenium was found to be significantly associated with CD. No significant difference was found between medium and high serum selenium cases and controls ( and ), suggesting that there is a level of serum selenium that is adequate. There have been previous reports on low selenium status in CD patients, especially in young patients [29
]. Levels as low as those found amongst nearly half of the CD patients in this study have been shown to have negative implications for the general maintenance of genomic stability [31
]. Selenium is known to be depleted in CD patients on enteral nutrition [32
], but this was not the form of nutritional intake for most of this group. Thus, an important question that is unanswered by this study is whether this association is a cause or an effect of the disease.
The failure of some large selenium human intervention studies to improve human health, e.g., the SELECT trial [34
], has raised questions as to what serum levels of selenium should be attained for optimal human health, and whether supplementation is beneficial. A recent review by Rayman suggests 122 ng/mL is desirable, and levels both below and above this threshold may be undesirable [35
]. On the current analysis, most of the control subjects had levels close to this optimal level, whereas many of the CD patients were significantly deficient in this mineral. It may be appropriate to consider supplementation for such individuals.
With respect to smoking, in those with a low or medium serum selenium level, there was a significant association with CD. It appears that high serum selenium levels may be protective against smoking-related risks of CD and the therapeutic effects amongst smokers with CD could be important to explore further. Cigarette smoking is known to have a detrimental effect with respect to the risk of CD and course of disease [36
]. Surprisingly, smoking is a protective factor in patients with UC [38
In the present study, SNPs in prioritized genes have been statistically analyzed to investigate whether, at a given serum selenium level, the SNPs were associated with a different risk of CD. Interestingly, one SNP in GPX1 (rs1800668), one SNP in GPX3 (rs3792797), three SNPs in DIO1 (rs12095080, rs2294511 and rs2294512), two SNPs in DIO2 (rs225011 and rs225012), two SNPs in SEPHS1 (rs17529609 and rs7901303), three SNPs in SEPSECS (rs11937742, rs1553153 and rs2302565) and one SNP in TXNRD2 (rs1548357) () were significantly associated with modified selenium levels indicating that these genes might increase the risk of developing CD.
In IBD, oxidative stress plays a major role in disease pathogenesis. GPX 1–4 and 6 are selenium dependent [39
] and some of the most important antioxidant enzymes in humans and therefore it is possible that SNPs in GPX1
may be involved in modulating several diseases. GPX1 is a ubiquitous enzyme with antioxidant properties, and plays an important role in antioxidative defense [40
] and therefore a dysfunction of GPX1
might be involved in CD pathogenesis. In this study we found that rs1800668 (C/T) in GPX1
is significantly associated with a gene-serum selenium interaction on CD (p
> 0.027) but did not remain significant after adjustment for multiple testing. The rs1800668 polymorphism has not been widely studied or found to be significantly associated with disease. One of the difficulties associated with GPX activity is that it can change over time depending on lifestyle factors, and vary from one cell type to another [42
]. Hansen et al
. found that only 5% of the total variation in GPX (in erythrocytes) could be accounted for by genotype and lifestyle factors such as smoking, alcohol and fruit and vegetable intake [42
Interestingly, previous studies found that altered expression and activity of DIOs is present in several models of heart failure, brain injury, tumor tissues and local inflammation [43
]. These findings support the hypothesis that SNPs in DIOs may play an important role in the development of CD. However, in this study none of the ten SNPs tested in the DIO1
genes were found to significantly interact with gene-serum selenium with CD after multiple corrections was applied. The role of selenium in modifying thyroid hormone levels is controversial [47
]. Patients with autoimmune thyroiditis and a baseline serum selenium levels around 83 ng/mL supplemented with 200 µg/day selenomethionine for 3 or 6 months have shown no changes in free thyroxin or free tri iodothyronine levels [50
], suggesting that DIO activity is saturated at these serum selenium levels. The current study group has a serum selenium level of 101 ng/mL which is higher than the levels recorded in Anastasilakis et al.
]. Therefore, it is unlikely that selenium level has a direct influence on DIO1
activity or thyroid function in this group.
The effect of selenium on the thyroid hormone metabolism might be due to lower cytokine release that results in a higher activity of DIO1
. Therefore, selenium deficiency might lower the DIO1
activity by lower availability of selenium for DIO1
synthesis. It is unlikely that selenium supplementation has a direct influence on DIO1
activity or thyroid function, but selenium indirectly improves clinical conditions which lead to normalization of thyroid functions. Furthermore, the effect of selenium depends on the tissue. For example, the thyroid and pituitary are more resistant to dietary selenium compared to the liver [51
]. In addition, interaction effects between selenium and other dietary nutrients such as iodine exists, and some of these interactions that modify thyroid hormone metabolisms and antioxidant status have been described elsewhere [52
]. It was not possible to consider these other dietary nutrients in this study.
SNPs in both SEPSECS
were found to significantly interact with serum selenium levels and CD, even after correction for multiple testing. The three SNPs that have been shown to be significant have not been reported elsewhere. SEPHS2
has been shown to be essential for Selenocysteine synthesis, but this is not so for SEPHS1
]. It is possible that SEPHS1
is involved in the regulation of oxidative stress and cell growth [56
]. The interaction of SEPHS1
with p53 has been reported by Chung et al
]. Gender specific p53 variants have been associated with CD [58
]. The p53 mutations in UC related cancers have also been shown [59
]. This implies a possible interaction of SEPHS1
with the mutated p53 pathway causing a malabsorption of selenium. The gut inflammation in IBD could be affecting the function of p53 [60
] and causing downstream influence on other pathways such as the functionality of SEPHS1
It is possible that the SEPSECS is involved in CD development due to their involvement in the synthesis of selenocysteine. However, the functional consequences or secondary effects are more likely to be directly related to CD pathogenesis. Genetic variants in SEPSECS may lead to insufficient function resulting in an inappropriate synthesis of selenoproteins which in turn lead to decreased selenium levels. However, the exact function of SEPSECS and SEPHS1 is still unclear and further studies are needed to investigate their role in CD development. Further studies are required to link genetic variants in selenoproteins and associated entities with modified selenium levels and altered selenium metabolism.