Biologic Significance of the Short Tandem Trinucleotide Repeats in the Androgen Receptor Gene

Androgen receptor (AR), a ligand-activated transcription factor, regulates the transcription of many androgen-responsive genes and mediates the genomic actions of testosterone and other androgens. AR consists of a ligand-binding domain, a highly-conserved DNA-binding domain, a nuclear localization domain, and a modulatory transactivation domain that contains the activation function-1, activation function-2, and activation function-5. The N-terminal domain of AR includes a polyglutamine tract encoded by a CAG repeat sequence in exon 1, and a polyglycine tract encoded by the repeated codon GGC (1, 2). Similar homopolymeric stretches of amino acids are found in many transcription factors.

The CAG tandem repeat sequence in AR begins at codon 58 and has an average of 23 repeats. Very long CAG sequences cause the AR protein to form toxic aggregates that lead to the loss of motor neurons in the spinal cord and brain stem, and damage to muscle fibers resulting in X-linked Spinal and Bulbar Muscular Atrophy, or Kennedy disease (3). Men with Kennedy disease exhibit signs of partial androgen insensitivity such as gynecomastia and infertility. Similar unstable expansions of homopolymeric microsatellites have been linked to more than 2 dozen human diseases such as Huntington disease, Fragile X syndrome, amyotrophic lateral sclerosis, and Friedreich ataxia. Most short tandem repeats remain stable and the mechanisms that cause short tandem repeats in some locations to undergo unstable premutation and mutation level expansion and human disease are incompletely understood. Sun et al (4) found that disease-associated short tandem repeats are generally located at boundaries demarcating chromatin domains with high CpG island density.

The length of polyglutamine tracts is associated negatively with the transcriptional activity of AR in in vitro assays but does not affect its binding affinity (5). Whether the length of the polyglutamine tract in the range of its distribution in the general population is associated with circulating testosterone levels or the risk of testosterone-related traits and diseases, remains controversial. A large study conducted by the National Cancer Institute Breast and Prostate Cancer Cohort Consortium (6) found a strong association between longer CAG repeats and circulating testosterone and estradiol levels; however, the amount of variance explained was small. Conflicting data have been reported on the association of polyglutamine tract length and androgen disorders likely because of the small sample sizes of the earlier studies and other methodological problems. It has remained unclear whether the CAG repeat length influences the relation between testosterone levels and androgen-related traits and diseases.

The paper by Sasako et al in the Journal of Clinical Endocrinology & Metabolism (2) is significant in several aspects. The analyses by Sasako et al included quantification of CAG tract length from the whole genome sequence data from 75 269 men and whole exome sequence data in 181 217 men in the UK Biobank cohort. The large sample size of the study provided greater statistical power than was possible with previous studies. In addition to analyzing the relation of CAG and GGC tract length with the risk of androgen-related trials and diseases, the authors also evaluated their influence on the relation between testosterone levels and androgen-related traits and diseases. The authors used ExpansionHunter for accurate determination of CAG and GGC repeat lengths; the traditional polymerase chain reaction-based methods are susceptible to unequal amplification of differently sized alleles and lower product yields at longer allele lengths (7).

The analyses by Sasako et al (2) confirmed that CAG repeat length is positively associated with circulating total testosterone levels and showed that GGC trinucleotide repeat length also is positively associated with total testosterone level, suggesting that the longer CAG and GGC tract lengths cause partial androgen resistance. Both CAG and GGC repeat lengths also were positively associated with estimated bone mineral density measured using heel ultrasound. The CAG repeat length was negatively associated with male-pattern baldness. The effect sizes of these associations were small, and the lengths of the CAG and GGC tracts did not influence the relation of testosterone levels with androgen-related traits or diseases. Importantly, neither the CAG nor the GGC tract length was associated with risk of prostate cancer, testicular cancer, infertility, or diabetes mellitus.

These important findings on the significance of the trinucleotide repeats should be viewed in the context of some inherent limitations of the UK Biobank database. Total testosterone levels in the UK Biobank were measured using an immunoassay that has limited accuracy and precision in the low range. Free testosterone levels were estimated using an equation that assumes fixed values for the dissociation constants that are inconsistent with the dynamic conformational changes in the SHBG monomers because of the intermonomeric allostery and disregards the prevalent concentrations of other SHBG ligands such as DHT and estradiol. The estimates of bone mineral density by ultrasound of the heel are not a measure of bone mineral content or osteoporosis. The UK Biobank cohort is not a representative sample of the general population and suffers from healthy volunteer bias, survival bias, and recruitment bias from the inclusion of related individuals (8). The analyses by Sasako et al were limited to people of European ancestry. The International Classification of Disease-based diagnoses in electronic medical record are susceptible to inaccuracies.

The study by Sasako et al, by virtue of its large sample size and more accurate quantification of the CAG and GGC repeats in the AR gene, has substantially advanced our understanding of the role of these short tandem repeats in modulating testosterone’s effects and the risk of androgen-related disorders. Their finding that greater length of the CAG as well as GGC tracts is associated with higher testosterone levels is consistent with reduced transcriptional activity of the AR with longer repeat length. However, the overall effect is very small. These analyses do not reveal a significant relation between CAG and GGC tract length and the risk of prostate or testicular cancer, or infertility. In view of the small magnitude of the effect and the lack of significant association with the risk of androgen-related disorders, these data do not justify the evaluation of CAG and GGC tract length in clinical practice at present.

Abbreviation

AR

androgen receptor

Funding

Dr. Bhasin is partially supported by the Boston Claude D. Pepper Older Americans Independence Center (National Institute on Aging grant 5P30AG31679).

Disclosures

Dr. Bhasin reports receiving research grants paid to his institution from the National Institute on Aging, the National Institute of Child Health and Human Development—National Center for Medical Rehabilitation Research, AbbVie, Metro International Biotech, FPT; receiving consulting fees from Besins and Versanis; and equity interest in FPT.