4 DISCUSSION
The SNP rs41423247 (GC) of theNR3C1 gene was found to be protective against HPAS, while no association with HPAS was found with rs242941 and rs1876828 of the CRHR1 gene. In contrast, in Greek asthmatic childrenbasal or stimulated F levels were differentfor all three SNPs.5 Thediscrepancy in the findings between results of the two studies can be explained by the use of different adrenal function tests. In the Greek study the low dose ACTH stimulation test (not the gold standard) was used without defining definitive cut-offs3,6for basal and stimulated F levels. Relying on statistical differences of these levels vis-a-vis their respective genotypes could have led to erroneous conclusions. Furthermore, serum F is a very crude test, which is influenced by many variables3 e.g. by Circadian rhythm. There is no record in Tsatsali et al’s paper whether the basal F levels were taken in the early morning (which is essential). The MTP test, utilized in this study, is a gold standard adrenal function test with clearly defined criteria for HPAS. Analyses based on this test can be expected to produce valid results.3
In a recently published genome-wide association study (GWAS) investigating susceptibility to HPAS in asthmatic children, screening was not performed for rs41423247 (GC).7Instead rs591118, an intronic variant of the PDGFD gene, was significantly associated with HPAS. This particular SNP was not included in the current study, because the study was already completed when the paper reporting rs591118 was published.
In the current study, CC of rs41423247 is associated with higher ACTH levels when compared to theGCand GG genotypes. In adults with depression on the other hand, a dexamethasone/corticotropin-releasing hormone (CRH) test revealed that a poor response to anti-depressive therapy was associated with marginally lower median peak ACTH levels in patients with the CC genotype.8The results of this study could however have been confounded by the use of a variable dose of triiodothyronine given to some patients,which could have inducedunrecognised adrenal insufficiency in individuals at risk.9Tsartsali et al5 and others10reported that the GC genotype was associated with a significantly higher basal F level than the CC genotype. The GG genotype converselywas associated with lower F levels after a dexamethasone suppression test in Dutch elderlyindividuals.11The authors interpreted this as “hypersensitivity” to glucocorticoids. However, on re-analysis of their paper, CC is associated with a significantly higher F level post-dexamethasone, suggesting glucocorticoid resistanceof the CC genotype.
Both the HPAS-protective effect and the ACTH response seem to be inherited in an additive and dominant fashion. A similar phenotype is therefore expected withthe heterozygous GCand homozygous CCgenotype, except that the phenotypic effect of the latter should be more pronounced. Although the overall sample size was probably adequate, the low minor allele frequency of 23% (compared to 38% in Europe12) and the low CC genotype frequency of 6% (compared to 40% in Greek children), limit the interpretation of the study. This resulted in overlapping CIs of the GC and CC genotypesand an incalculable lower bound of the CC CI.However, the additive genetic effectaccounted for a huge effect size and a highly significant p-value (<2 x 10-16), which even far exceeds the required significance threshold for a GWAS (5 x 10-8).13 Hence, pending confirmation in a larger study, the most suitable marker for HPAS protection would still be rs41423247(CC) of theNR3C1 gene.
The biochemical phenotype attributed to both CC and GC genotypes ofrs41423247is akin to the one described in the syndrome of primary generalized glucocorticoid resistance (PGGR).14-16In this syndrome there is dysregulation of F feedback at hypothalamic and pituitary level, resulting in elevated ACTH and possibly F levels. In the current study, early morning basal F levels of the children with rs41423247were not elevated (results not shown).Clinical features of this syndrome e.g. hypertension, hirsutism or premature pubarche,were not detected.Hence, the biochemical phenotype of rs41423247(GC and CC) couldrepresent tissue-specific glucocorticoid resistance.16More likely, it isjust a common variant in the population (40% in our sample)presenting with reduced glucocorticoid sensitivity.16
PGGR is associated with 26 mutations16and 8 pathogenic SNPS17in the exome of NR3C1 at 5q31.3,while rs41423247falls within the intronic regulatory region of the gene,12confirmed independently by variant browser analysis(result not shown). Mechanistically, impaired promoter function would lead to impaired DNA transcription, RNA translation and protein production.
In a recently published study on asthma children in Turkey, the FEV1ofchildren with the GG genotype improved more after a high dose of ICS than children with the CG+CC genotype.18In the current study, three of five children with the CC genotype had a suboptimal FEV1(result not shown). Unfortunately the number is too small to draw any valid conclusions from it.
In the same study population, it was previously shown that, as the BMI z-score rises, the ACTH level rises.1In the current study, an association between CC genotypeand BMI z-score could be demonstrated. No association of genotype was found with overweight or obesity (results not shown). In Dutch elderly the GG genotype was associated with a lower BMI11, in line with the current study. However, others have found the opposite.19Thehigher BMI z-score associated with rs41423247is probably mediated through periodically elevated F levels. However, as shown, the BMI z-score does not confound the SNP effect on the ACTH level.
As findings in Greek and South African children (and possibly Turkish children and Dutch adults) are similar, it is likely that the results of this study are generalizable across all populations and ages. Pathophysiological and molecular explanations of the observed phenotype should also be equally applicable to all humans. However, inadvertentlythe study islimited by the small sample of children with the CC genotype. Hence, alarger study should be performed to confirm the finding. In addition, a “proof of concept” study (screening asthmatic children for rs41423247 (CC), and when positive, testing them with the MTP testto confirm the phenotype)should be done. If confirmed, rs41423247 (CC), as a marker of HPAS protection, should be screened for in asthmatic children on glucocorticoids. This would allow clinicians to choose which children can be safely treated with higher doses of glucocorticoids – afirst step towards personalized medicine in asthma therapy.
In conclusion, rs41423247(GC) of theNR3C1 gene was found to be protective against HPAS. The CC genotype was associated with the highest ACTH levels. TheC allele is inherited as an additive and dominant trait. The GCgenotype was independently associated with BMI z-score. If confirmed by alarger and a “proof of concept” study,rs41423247 (CC) could well be a marker of HPAS protection, thereby allowing for personalized treatment decisions to be madein asthma care.
ACKNOWLEDGEMENTS
We are grateful to Carl Lombard from the Medical Research Council for calculating the sample size. Sisters Poggenpoel and Van Wyk are thanked for collecting the salivary samples of the patients. We express our gratitude toAlvera Vorsterfrom the Central Analytical Facility for the DNA extractions and genotyping of the participants. We thank Maia Lesosky from the School of Public Health, University of Cape Town, for statistical advice. For their advice on genetic modelling we are indebted to Heather Cordell from the International Centre for Life, Newcastle University and Ahmed Rabie, Centre of Biotechnology of Sfax, Tunisia.
REFERENCES
  1. Zöllner EW, Lombard CJ, Galal U, Hough FS, Irusen EM, Weinberg E. Hypothalamic-pituitary-adrenal axis suppression in asthmatic school children. Pediatrics 2012; doi: 10.1542/peds.2012-1147.
  2. Zöllner EW, Lombard C, Galal U, Hough S, Irusen E, Weinberg E. Hypothalamic-pituitary-adrenal axis suppression in asthmatic children on inhaled and nasal corticosteroids – more common than expected? J PediatrEndocrinolMetab 2011; 24 (7-8): 529-534.
  3. Zöllner EW. Hypothalamic-piuitary-adrenal axis suppression in asthmatic children on inhaled corticosteroids: Part 1. Which test should be used ?Pediatr Allergy Immunol 2007;18:401 – 9.
  4. Zöllner EW, Lombard CJ, Galal U, Hough FS, Irusen EM, Weinberg E. Screening for hypothalamic-pituitary-adrenal axis suppression in asthmatic children remains problematic: a cross- sectional study. BMJ Open 2013; pii: e002935.doi: 10.1136/bmjopen-2013-002935.
  5. Tsartsali L, Papadopoulos M, Lagona E, Papadimitriou A, Kanaka-Gantenbein C, Louizou E, Kastania A, Priftis KN, Chrousos G. Association of hypothalamic-pituitary-adrenal axis-related polymorphisms with stress in asthmatic children on inhaled corticosteroids. Neuroimmunomodulation 2012; 19: 88-95.
  6. Zöllner EW. Asthma treatment in children: A guide to screening for and management of hypothalamic-pituitary-adrenal axis suppression. S Afr Med J 2019; 109(5):306-309. DOI:10.7196/SAMJ.2019.v109i5.13932.
  7. Hawcutt DB, Francis B, Carr DF et al. Susceptibility to corticosteroid-induced adrenal suppression: a genome-wide association study. Lancet Respir Med 2018;6: 442-450.
  8. Brower JP, Appelhof BC, Van Rossum EFC et al. Prediction of treatment response by HPA-axis and glucocorticoid receptor polymorphisms in major depression. Psychoneuroendocrinology 2006; 31: 1154-1163.
  9. Garber JR, Cobin RH, Gharib H et al for the American Association of Clinical Endocrinologists and American Thyroid Association Taskforce on Hypothyroidism in Adults. Thyroid, 2012; 22 (12): 1200-1235.
  10. DeRijk RH, Single nucleotide polymorphisms related to HPA axis reactivity. Neuroimmunomodulation 2009; 16: 340-352.
  11. Huizenga NATM, Koper JW, De Lange P, Pols HAPP, Stolk RP, Burger H, Grobbee DE, Brinkmann AO, De Jong FH, Lamberts SWJ. A polymorphism in the glucocorticoid receptor gene may be associated with an increased sensitivity to glucocorticoids in vivo. J ClinEndocrinolMetab 1998; 83: 144-151.
  12. National Library of Medicine. National Center for Biotechnology Information. dbSNP. Short Genetic Variations. Reference SNP (rs) Report. rs41423247. https://www.ncbi.nlm.nih.gov/snp/rs41423247#frequency_tab. Accessed April 7, 2020.
  13. Dudbridge F, Gusnanto A. Estimation of Significance Thresholds for Genomewide Association Scans. Genet Epidemiol 2008; 32: 227-234.
  14. NCBI Resources. OMIM. https://omim.org/entry/615962. Accessed January 17, 2020.
  15. Jiméne-Panizo A, Pérez P, Rojas AM et al. Nonc-canonical dimerization of the androgen receptor and other nuclear receptors: implications for human disease. Endocrine-Related Cancer 2019; 26: R479-R497.
  16. Van Rossum EFC, Van den Akker ELT. Glucocorticoid Resistance. In: Ghizzoni L, Cappa M, Chrousos G, Loche S, Maghinie M (eds): Pediatric Adrenal Diseases. Endocr Dev. Basel: Karger; 2011; 20. 127-136 pp.
  17. https://www-ncbi-nlm-nih-gov.ez.sun.ac.za/clinvar/?term=NR3C1%5Bgene%5D. Accessed May 11, 2020.
  18. Keskin O, UlucaŰ, Birben E, Coșkun Y et al. Genetic associations of the response to inhaled corticosteroids in children during an asthma exacerbation. Pediatr Allergy Immunol 2016: 27: 507-513.
  19. Manenschijn L, Van den Akker ELT, Lamberts SWJ, Van Rossum EFC. Clinical Features Associated with Glucocorticoid Receptor Polymorphisms. An Overview. Ann NY AcadSci 2009; 1179: 179-198.