The iodothyronine deiodinases are a family of selenoproteins that control serum and intracellular thyroid hormone levels [1,2]. Their importance in thyroid hormone homeostasis has become increasingly clear from studies of deiodinase-deficient animals . The classic clinical example in which altered deiodinase function is encountered is in critical illness wherein substantial alterations in thyroid hormone metabolism are observed . Given that even modest variation in thyroid function is associated with substantial phenotypic effects , the deiodinases have become an attractive target for further study.
There are three deiodinases identified. Type 1 deiodinase (DIO1) and type 2 deiodinase (DIO2) convert thyroxine (T4) into the biologically active form tri-iodothyronine (T3) and also enhance clearance of reverse T3 (rT3) [1,5]. In contrast, type 3 deiodinase (DIO3) regulates the conversion of T3 to T2 and T4 to rT3 and is the main thyroid hormone inactivating enzyme . Their complex interrelationship has been highlighted in that mice devoid of all three deiodinases have a milder phenotype than knockout mice only lacking DIO3.
The deiodinases are also expressed to varying degrees in different tissues. DIO1 is predominantly expressed in the pituitary, thyroid, liver and kidney, whereas DIO2 is expressed in the central nervous system, pituitary, thyroid, heart, bone and brown adipose tissue. DIO3 is expressed in numerous tissues in fetal life, but only in the central nervous system and placenta in adults. Recently, the deiodinases have also been shown to be expressed in endothelial cells . This observation might be very clinically relevant given the important role of the endothelium as the first barrier to the bloodstream.
In recent years, the genetic architecture of thyroid function has become more established [8,9,10▪▪]. Taken together there is now growing interest in how genetic variation and abnormalities in the deiodinases may influence health [11▪▪]. The deiodinases may influence health outcomes, not only by altering serum thyroid hormone levels but also through altered intracellular activity. The latter could have substantial implications for individuals on thyroid hormone replacement or those with borderline thyroid function . This article complements a recent systematic review of the implications of genetic variation in the deiodinases [13▪] by focusing on recent clinical studies in this area.
Effect of genetic variation in the deiodinases on serum thyroid function
Twin and family studies estimate the heritability of thyroid stimulating hormone (TSH) levels to be 65%, and FT4 up to 39% . In a recent HapMap-based genome-wide association study , only 2.3% of the variance in FT4 was, however, explained. The availability of whole genome sequencing (WGS) in the UK10K project (http://www.uk10k.org/) enabled a reassessment of the genetic architecture of TSH and FT4. Although this study confirmed previous reports that variation at rs2235544 in DIO1 was robustly associated with FT4[8,10▪▪,15–18], no other single nucleotide polymorphisms (SNPs) in DIO1 were associated with FT4 or TSH levels even after conditional analysis was performed on rs2235544. WGS data allowed sequence kernel-based association testing of rare aggregates in DIO1, but none was associated with either FT4 or TSH.
Genome-wide association studies to date have failed to identify SNPs in DIO2 or DIO3 associated with either TSH or FT4[8,10▪▪]. With regard to SNPs in DIO3 their impact may be hampered by genetic imprinting as effects of DIO3 polymorphisms on thyroid hormone homeostasis depend on the parental origin of the variant allele . Notably even with WGS data available no associations were observed with TSH or FT4 in DIO2 or DIO3 although crucially this study did not assess the FT3:FT4 ratio [10▪▪]. This result is in keeping with recent data from the Thyroid Origin of Psychomotor Retardation study which identified two rare variants in DIO2 c.11T>A (p.L4H) and c.305C>T (p.T102I) although these variants had no evidence of a deleterious impact on either thyroid function or intelligence quotient (IQ)  in affected individuals. Further analysis of the effects of rare genetic variants on thyroid function is likely to be informative as genome-wide complex trait analysis [10▪▪] identified that common variants (minor allele frequency >1%) explain only 20% of the variance in FT4. The deiodinase genes therefore remain an attractive target for further study of rare variants associated with thyroid status, although much larger studies are likely to be needed.
Variation in DIO2 function appears to adversely affect an individual's TSH response. Luongo et al. recently showed that the type 2 iodothyronine deiodinase (D2) is essential for feedback regulation of TSH by T4. In this experiment, DIO2 activity was reduced 90% in Cga-cre D2KO mice compared with wild type. Basal TSH levels were increased almost two-fold, but serum T4 and T3 were not different from the controls in adult mice. In hypothyroid adult mice, suppression of TSH by T4, but not T3, was impaired in Cga-cre DIO2 mice. Despite the mild basal TSH elevation, the TSH increase in response to hypothyroidism was reduced four-fold in the Cga-cre D2KO compared with control mice despite an identical level of pituitary TSH α- and β-subunit mRNAs. Taken together this study suggests that impaired DIO2 function in thyrotrophs would be deleterious to the TSH compensation in individuals with lower thyroid function and/or iodine deficiency. In line with this, the -258G/x DIO2 polymorphism (rs12885300) variant has been associated with a decreased rate of acute TSH-stimulated FT4 secretion with a normal FT3 release from the thyroid gland . These results are consistent with the finding that the Thr92Ala substitution in DIO2 is associated with a decreased rate of acute TSH-stimulated T3 release from the thyroid consistent with a decrease in intrathyroidal deiodination . Although it should be noted that these studies were performed in small populations (n = 45). Nevertheless, they are proof of concept that common polymorphisms in DIO2 can subtly affect the circulating levels of thyroid hormone and might modulate intracellular thyroid hormone homeostasis and the response to environmental insults.
Effect of genetic variation in the deiodinases on phenotypic outcomes
Most work continues to focus on the consequences of variation in DIO2 and in particular the common variant, the Thr92Ala substitution (rs225014) which is not infrequent in all populations. Homozygosity for Thr92Ala was initially associated with insulin resistance, osteo-arthritis and several other key phenotypes [13▪]. Its mechanism of action, however, remains unclear. Initially, it had been thought to reduce DIO2 enzyme velocity ; however, this has since been disputed . More recent data have indicated another potential mechanism that the Thr92Ala substitution results in an accumulation in the Golgi apparatus and disruption of cellular functions, leading to an 81 gene fingerprint of transcriptional alterations including mitochondrial dysfunction, inflammation, apoptosis, DNA repair and growth factor signaling [25▪]. This may explain why this substitution results in such diverse phenotypic outcomes, but it does not explain the effects on the response to TRH stimulation  or the data suggesting that hypothyroid individuals on levothyroxine therapy appear to prefer combination therapy with levothyroxine and liothyronine than levothyroxine therapy alone . Speculation also continues that the Thr92Ala substitution may influence ubiquitination of the DIO2 enzyme which has a central role in determining its cellular activity . Another possible explanation for why the Thr92Ala variant is associated with preference for combination thyroid hormone replacement is that it may be tagging another rarer functional variant. The latter might also explain why laboratory studies of an effect on intracellular FT3 levels have been divergent, but population studies have been more convincing of an impact in the thyroid hormone pathway. Replication of these data in large populations is required to clarify the situation.
It might be expected that the effects of genetic variation in deiodinases would be amplified under conditions in which the supply of thyroid hormones is compromised by other factors. Evidence for an interaction between lower thyroid states and the Thr92Ala substitution has been demonstrated recently  in unpublished data from the Avon Longitudinal Study of Parents and Children which identified that Children with FT4 in the lowest quartile and who were homozygous for the Thr92Ala substitution had higher odds of an IQ less than 85 (odds ratio = 3.03, 95% confidence interval 1.38, 6.67, P = 0.006). Crucially, there was also evidence of interaction between FT4 in the lowest quartile and the Thr92AlaD2 substitution in their relationship with IQ (P = 0.006). Although intriguing these results do still require replication in an independent cohort.
An inadequate TSH response arising from this variant to borderline thyroid function may, however, not be part of the mechanism of action as recent data indicated that the Thr92Ala substitution and variation at rs12885300 were not associated with higher T4 doses to achieve TSH suppression in thyroid cancer in a modestly powered study (N = 285) . This is not in keeping with earlier studies (N = 191) which showed that individuals homozygous for the Thr92Ala substitution required higher doses of levothyroxine to achieve target TSH levels post-thyroidectomy for thyroid cancer .
In addition, the fact that serum FT4 and FT3 levels were not different between the genotype groups in this study did not fully support an altered pituitary setpoint in these patients either . Hence, it seems more likely that any additive effect of low thyroid hormone levels with D2 deiodinase genotype occurs at the level of local T3 generation in parts of the brain separate from the HPT axis control. In another small study , the Thr92Ala substitution was associated with an increased susceptibility to schizophrenia.
It appears that other factors may also affect the effects of deiodinease variation on disease. The Thr92Ala substitution has been previously identified to be associated with osteoarthritis . Intriguingly, further analysis here has revealed that in osteo-arthritis, related to changes in articular cartilage, loss of epigenetic silencing results in upregulation of DIO2 expression amongst individuals carrying the risk allele at rs225014 in DIO2 in contrast to the reduced deiodinase activity predicted from this under other conditions [33▪▪]. This genetic upregulation of DIO2 expression appears to result in reduced ability of chondrocytes to deposit extracellular matrix components. Additional analysis here implied that the detrimental effects of DIO2 upregulation were the result of increased T3 synthesis, as reflected by the identical results when adding T3 to the culture medium used in the study [33▪▪]. Although these results are in contrast to earlier studies showing a decreased D2 activity in tissue homogenates of homozygous variant carriers , this again indicates that there is likely to be some interaction occurring between the local environment and the Thr92Ala substitution. A positive correlation between methylation at CpG-2031 and DIO2 expression in articular cartilage among carriers of the rs225014 risk allele was also observed. Possibly, the rs225014 SNP influences three-dimensional chromatin conformations underlying the relationship between the rs225014 tagged allelic imbalance and methylation-dependent upregulation of DIO2 among rs225014 risk allele carriers [33▪▪]. It appears likely therefore that individuals with the rs225014 risk allele are genetically predisposed to osteoarthritis, and potentially inhibitors of deiodinase functions could lead to novel treatments for osteoarthritis. Further studies in mice have indicated that lack of DIO2 protects against cartilage damage further supporting DIO2 activity as a therapeutic target in osteoarthritis .
A small study (N = 294) has also revealed that the Thr92Ala genotype is associated with reduced placental D2 activity (0.35 ± 0.15 vs. 1.96 ± 1.02 fmol/mg/min), but it was not associated with dysglycemia, increased insulin resistance or worse gestational outcomes . The lack of disease association here is, however, likely to be because of lack of power and may be more apparent in individuals with other predisposing factors as seen with osteoarthritis.
Genetic variation in the deiodinases is associated with altered thyroid function and adverse phenotypic outcomes [11▪▪]. To date, the effects have, however, been modest and not consistently replicated. Only one variant in DIO1 has been consistently associated with thyroid function in genome-wide association studies, although other variants in DIO1 and DIO2 have been identified in smaller candidate gene studies , which would not have passed the genome-wide association threshold. Genetic association studies in this area may need to be an order of magnitude larger than previous studies and will require WGS or similar high quality genetic data to enable identification of new SNPs associated with thyroid function. These are likely to be a combination of common variants with modest effects and rare variants with more substantial effects [10▪▪]. It should be highlighted that large collaborative studies even in selected populations at potentially higher risk of genetic mutations are required. For instance, a study of 100 individuals with high bone mass found no evidence of TRa or DIO1 mutations . A large meta-analysis of genetic variants associated with thyroid function is currently underway and results are expected in the near future.
A multitude of phenotypic associations have been identified with genetic variants in the deiodinases particularly the Thr92Ala substitution in DIO2 ranging from osteoarthritis to insulin resistance [13▪,27]. There has, however, been no substantial impact of deiodinase variants on some potential thyroid related phenotypes such as lipoprotein levels or obesity [13▪] and results have been inconsistently replicated. Furthermore, several potential mechanisms for the action of the Thr92Ala substitution on DIO2 and intracellular exist [20,33▪▪]. Taken together, these inconsistencies appear to be best explained by multiple pathway mechanisms, some of which are tissue specific (e.g., an effect on Golgi function in neurons, or gene methylation in chondrocytes) and others dependent on compounding environmental factors (such as iodine status  and thyroid status  in their relationship with IQ). A further possible mechanism is that the Thr92Ala substitution in some cases is not the functional variant and merely tags a rare variant, or other factors necessary such as gene–gene interactions.
Until further studies are done to explore these potential interactions, the significance of deiodinase polymorphisms in clinical practice remains uncertain. In particular, further studies are needed to explore their potential role, particularly in individuals on levothyroxine for hypothyroidism. There is a steadily rising number of patients on levothyroxine  and variation in the deiodinases may influence their treatment preference  and potentially treatment outcomes. Further trials in this area and treatment of subclinical thyroid disease which incorporate genetic analyses are therefore urgently needed.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING
Papers of particular interest, published within the annual period of review, have been highlighted as:
- ▪ of special interest
- ▪▪ of outstanding interest
1. Kohrle J. Local activation and inactivation of thyroid
hormones: the deiodinase family. Mol Cellular Endocrinol 1999; 151:103–119.
2. Bianco AC, Kim BW. Deiodinases
: implications of the local control of thyroid
hormone action. J Clin Investig 2006; 116:2571–2579.
3. Peeters RP, Debaveye Y, Fliers E, Visser TJ. Changes within the thyroid
axis during critical illness. Crit Care Clin 2006; 22:41–55.
4. Taylor PN, Razvi S, Pearce SH, Dayan CM. Clinical review: a review of the clinical consequences of variation in thyroid
function within the reference range. J Clin Endocrinol Metab 2013; 98:3562–3571.
5. Bianco AC, Salvatore D, Gereben B, et al. Biochemistry, cellular and molecular biology, and physiological roles of the iodothyronine selenodeiodinases. Endocr Rev 2002; 23:38–89.
6. Galton VA, de Waard E, Parlow AF, et al. Life without the iodothyronine deiodinases
. Endocrinology 2014; 155:4081–4087.
7. Sabatino L, Lubrano V, Balzan S, et al. Thyroid
D1, D2, and D3 are expressed in human endothelial dermal microvascular line: effects of thyroid
hormones. Mol Cellular Biochem 2015; 399:87–94.
8. Porcu E, Medici M, Pistis G, et al. A meta-analysis of thyroid
-related traits reveals novel loci and gender-specific differences in the regulation of thyroid
function. PLoS Genet 2013; 9:e1003266.
9. Medici M, Porcu E, Pistis G, et al. Identification of novel genetic Loci associated with thyroid
peroxidase antibodies and clinical thyroid
disease. PLoS Genet 2014; 10:e1004123.
10▪▪. Taylor PN, Porcu E, Chew S, et al. Whole-genome sequence-based analysis of thyroid
function. Nat Commun 2015; 6:5681.
First whole genome-based analysis of the genetic architecture of thyroid function.
11▪▪. Medici M, Visser WE, Visser TJ, Peeters RP. Genetic determination of the hypothalamic-pituitary-thyroid
axis: where do we stand? Endocr Rev 2015; 36:214–244.
Comprehensive overview of the genetic architecture of thyroid function.
12. Dayan CM, Saravanan P, Bayly G. Whose normal thyroid
function is better: yours or mine? Lancet 2002; 360:353–354.
13▪. Verloop H, Dekkers OM, Peeters RP, et al. Genetics in endocrinology: genetic variation in deiodinases
: a systematic review of potential clinical effects in humans. Eur J Endocrinol 2014; 171:R123–R135.
Good overview of the clinical effects of variation in the deiodinases.
14. Panicker V, Wilson SG, Spector TD, et al. Heritability of serum TSH, free T4 and free T3 concentrations: a study of a large UK twin cohort. Clin Endocrinol (Oxf) 2008; 68:652–659.
15. Medici M, van der Deure WM, Verbiest M, et al. A large-scale association analysis of 68 thyroid
hormone pathway genes with serum TSH and FT4 levels. Eur J Endocrinol 2011; 164:781–788.
16. van der Deure WM, Hansen PS, Peeters RP, et al. The effect of genetic variation in the type 1 deiodinase gene on the inter-individual variation in serum thyroid
hormone levels: an investigation in healthy Danish twins. Clin Endocrinol (Oxf) 2009; 70:954–960.
17. Peeters RP, van Toor H, Klootwijk W, et al. Polymorphisms in thyroid
hormone pathway genes are associated with plasma TSH and iodothyronine levels in healthy subjects. J Clin Endocrinol Metab 2003; 88:2880–2888.
18. Panicker V, Cluett C, Shields B, et al. A common variation in deiodinase 1 gene DIO1
is associated with the relative levels of free thyroxine and triiodothyronine. J Clin Endocrinol Metab 2008; 93:3075–3081.
19. Peeters RP, van der Deure WM, Visser TJ. Genetic variation in thyroid
hormone pathway genes; polymorphisms in the TSH receptor and the iodothyronine deiodinases
. Eur J Endocrinol 2006; 155:655–662.
20. Zevenbergen C, Klootwijk W, Peeters RP, et al. Functional analysis of novel genetic variation in the thyroid
hormone activating type 2 deiodinase. J Clin Endocrinol Metab 2014; 99:E2429–E2436.
21. Luongo C, Martin C, Vella K, et al. The selective loss of the type 2 iodothyronine deiodinase in mouse thyrotrophs increases basal TSH but blunts the thyrotropin response to hypothyroidism. Endocrinology 2015; 156:745–754.
22. Peltsverger MY, Butler PW, Alberobello AT, et al. The -258A/G (SNP rs12885300) polymorphism of the human type 2 deiodinase gene is associated with a shift in the pattern of secretion of thyroid
hormones following a TRH-induced acute rise in TSH. Eur J Endocrinol 2012; 166:839–845.
23. Butler PW, Smith SM, Linderman JD, et al. The Thr92Ala 5’ type 2 deiodinase gene polymorphism is associated with a delayed triiodothyronine secretion in response to the thyrotropin-releasing hormone-stimulation test: a pharmacogenomic study. Thyroid
24. Canani LH, Capp C, Dora JM, et al. The type 2 deiodinase A/G (Thr92Ala) polymorphism is associated with decreased enzyme velocity and increased insulin resistance in patients with type 2 diabetes mellitus. J Clin Endocrinol Metab 2005; 90:3472–3478.
25▪. McAninch EA, Jo S, Preite NZ, et al. Prevalent polymorphism in thyroid
hormone-activating enzyme leaves a genetic fingerprint that underlies associated clinical syndromes. J Clin Endocrinol Metab 2015; 100:920–933.
Novel insight into potential mechanism of action of Thr92Ala.
26. Panicker V, Saravanan P, Vaidya B, et al. Common variation in the DIO2
gene predicts baseline psychological well being and response to combination thyroxine plus triiodothyronine therapy in hypothyroid patients. J Clin Endocrinol Metab 2009; 94:1623–1629.
27. Bianco AC, Casula S. Thyroid
hormone replacement therapy: three ‘simple’ questions, complex answers. Eur Thyroid
J 2012; 1:88–98.
28. Taylor PN, Okosieme OE, Hales C, et al. 37th Annual Meeting of the European Thyroid
Association. Leiden, The Netherlands: Abstracts. Eur Thyroid
J 2013; 2 (Suppl 1):195.
29. Santoro AB, Vargens DD, Barros Filho Mde C, et al. Effect of UGT1A1, UGT1A3, DIO1
polymorphisms on L-thyroxine doses required for TSH suppression in patients with differentiated thyroid
cancer. Br J Clin Pharmacol 2014; 78:1067–1075.
30. Torlontano M, Durante C, Torrente I, et al. Type 2 deiodinase polymorphism (threonine 92 alanine) predicts L-thyroxine dose to achieve target thyrotropin levels in thyroidectomized patients. J Clin Endocrinol Metab 2008; 93:910–913.
31. Colak A, Akan G, Oncu F, et al. 1508 – association study of the dio2
gene as a susceptibility candidate for schizophrenia in the Turkish population: a case-control study. Eur Psychiatry 28:1.
32. Meulenbelt I, Min JL, Bos S, et al. Identification of DIO2
as a new susceptibility locus for symptomatic osteoarthritis. Hum Mol Genet 2008; 17:1867–1875.
33▪▪. Bomer N, den Hollander W, Ramos YFM, et al. Underlying molecular mechanisms of DIO2
susceptibility in symptomatic osteoarthritis. Ann Rheumatic Dis 2014; (in press).
Excellent overview of how variation in DIO2 including the Thr92Ala may cause osteoarthritis.
34. Bomer N, Cornelis FMF, Ramos YF, et al. The effect of forced exercise on knee joints in Dio2
−/− mice: type II iodothyronine deiodinase-deficient mice are less prone to develop OA-like cartilage damage upon excessive mechanical stress. Ann Rheumatic Dis 2014; (in press).
35. Dora JM, Wajner SM, Costa JD, et al. Type 2 deiodinase Thr92Ala polymorphism is associated with disrupted placental activity but not with dysglycemia or adverse gestational outcomes: a genetic association study. Fertility Sterility 2014; 101:833–839.
36. Gogakos A, Logan JG, Waung JA, et al. THRA and DIO2
mutations are unlikely to be a common cause of increased bone mineral density in euthyroid postmenopausal women. Eur J Endocrinol 2014; 170:637–644.
37. Guo TW, Zhang FC, Yang MS, et al. Positive association of the DIO2
(deiodinase type 2) gene with mental retardation in the iodine-deficient areas of China. J Med Genet 2004; 41:585–590.
38. Taylor PN, Iqbal A, Minassian C, et al. Falling threshold for treatment of borderline elevated thyrotropin levels-balancing benefits and risks: evidence from a large community-based study. JAMA Intern Med 2014; 174:32–39.