There's also this:
Abstract:
Androgenetic alopecia (AGA) is a common heritable polygenic disorder whose genetics is not fully understood, even though it seems to be X-linked. We carried out an epidemiological survey for AGA on 9,000 people from 8 isolated villages of a secluded region of Sardinia (Ogliastra), and identified a large cohort of affected individuals. We genotyped 200 cases and 200 controls (mean kinship 0.001) with the 500k chip array and conducted case–control association analysis on the X chromosome. We identified Xq11-q12 as strongly associated with AGA. In particular, we found that rs1352015 located 8 kb from the EDA2R gene showed the best result (P=7.77e−7). This region also contains the AR gene, hence we tested both genes in 492 cases and 492 controls. We found that the non-synonymous SNP rs1385699 on EDA2R gave the best result (P=3.9e−19) whereas rs6152 on the AR gene is less significant (P=4.17e−12). Further statistical analysis carried out by conditioning each gene to the presence of the other showed that the association with EDA2R is independent while the association with AR seems to be the result of linkage disequilibrium. These results give insight into the pathways involved in AGA etiology.
Androgenetic alopecia (AGA) is a common heritable polygenic disorder whose genetics is not fully understood, even though it seems to be X-linked. We carried out an epidemiological survey for AGA on 9,000 people from 8 isolated villages of a secluded region of Sardinia (Ogliastra), and identified a large cohort of affected individuals. We genotyped 200 cases and 200 controls (mean kinship 0.001) with the 500k chip array and conducted case–control association analysis on the X chromosome. We identified Xq11-q12 as strongly associated with AGA. In particular, we found that rs1352015 located 8 kb from the EDA2R gene showed the best result (P=7.77e−7). This region also contains the AR gene, hence we tested both genes in 492 cases and 492 controls. We found that the non-synonymous SNP rs1385699 on EDA2R gave the best result (P=3.9e−19) whereas rs6152 on the AR gene is less significant (P=4.17e−12). Further statistical analysis carried out by conditioning each gene to the presence of the other showed that the association with EDA2R is independent while the association with AR seems to be the result of linkage disequilibrium. These results give insight into the pathways involved in AGA etiology.
Our study shows that AR and EDA2R are significantly associated with AGA. However, there is some LD between the two most associated markers for each gene (rs6152, rs1385699: D′=0.74, r2=0.43). To test if they are independently associated, we conditioned the analysis of each gene to the other one. We used the UNPHASED software (Dudbridge, 2003), which permits the association of a marker to be conditioned to the presence of another marker. The analysis of rs1385699 conditioned to the presence of rs6152 gave a very significant P-value of 6.136e−9, whereas when we conditioned the analysis of rs6152 to the presence of rs1385699 the P-value was 0.04. Again, rs1385699 conditioned to the presence of rs12558842 gave a very significant result (P-value 0.007), whereas rs12558842 conditioned to the presence of the EDA2R variant did not give a significant result (P-value 0.06). These results show that in our population, the EDA2R gene variation causes susceptibility to AGA. The conditioned analysis suggests that markers on the AR gene could be associated because of LD. However, we cannot exclude that other variants in LD with both genes (that is, regulatory elements of either or both genes) could be associated with AGA. Moreover, the functional importance of AR has already been proven by many means, and its involvement in this pathology cannot be excluded. Further functional and genetic studies are needed to clarify the role of these two genes and their possible interactions in the etiology of AGA.
Two receptors for EDA were found that are specific for the two isoforms EDA-A1 and EDA-A2: EDAR and EDA2R, respectively. EDA-A1 and its receptor EDAR are capable of activating the NF-κB pathway and are implicated in hair growth (Botchkarev and Fessing, 2005). EDA2R is capable of activating the NF-κB pathway and also through TRAF3,6, JNK (c-Jun N-terminal kinase) (Sinha et al., 2002), which activates c-Jun. Mutations in EDA and EDAR give rise to ectodermal dysplasia, a clinical syndrome characterized by loss of hair, sweat glands, and teeth, whereas mutations in EDA2R do not (Monreal et al., 1999; Naito et al., 2002; Newton et al., 2004). Recently, a preliminary report suggested that EDAR may influence hair thickness in Asians (A. Fujimoto, R. Kimura, J. Ohashi, U. Samakkarn, W. Settheetham-Ishida, T. Ishida, Y. Morishita, T. Furusawa, M. Nakazawa, R. Ohtsuka, R. Yuliwulandari, L. Batubara, M.S. Mustofa, K. Tokunaga, A scan for genetic determinants of human hair morphology: EDAR is associated with Asian hair thickness, ASHJ Meeting 2007). EDA2R could influence the onset of AGA through the activation of the NF-κB pathway or by c-Jun, which has been shown to be critical for AR transactivation (Bubulya et al., 1996). Moreover, in adult mice, EDA2R is also expressed in the hair bulb and in differentiating hair matrix (Botchkarev and Fessing, 2005). Looking at the human expression data from the UniGene database (http://www.ncbi.nlm.nih.gov/sites/entrez), we noticed that it is expressed during embryonic life and, especially, in the first weeks after birth. Expression then seems to be absent until the 17th year of age, when it recurs in different tissues, including skin. This expression pattern fits very well with the course of AGA, with its onset around puberty.
I actually have
rs6152 = A (the non-risk allele, present in only 1 of the 54 young balding men in the paper you cited)
rs1385699 = T (the risk allele on the EDA2R gene, associated with baldness)
Leave a comment: