Multiplex ligation-dependent probe amplification analysis is useful for diagnosing congenital adrenal hyperplasia but requires a deep knowledge of CYP21A2 genetics.

We read with great interest the recent report in Clinical Chemistry by Cantürk et al. (1 ). These authors affirmed that the CYP21A1P (cytochrome P450, family 21, subfamily A, polypeptide 1 pseudogene) genotype interferes with quantitative multiplex ligation-dependent probe amplification (MLPA) analysis of the CYP21A2 (cytochrome P450, family 21, subfamily A, polypeptide 2) gene. They also reported that the p.I172N and p.Q318X mutations were absent in 3.6% and 8.5%, respectively, of the CYP21A1P alleles (200 unrelated individuals examined). Because CYP21A2 MLPAspecific probes contain the wild-type sequences for the Del8bp, p.I172N, Cluster E6, and p.Q318X mutations, Cantürk et al. stated that “without a parallel CYP21A1P [sequencing] analysis, one would falsely infer a heterozygous duplication of CYP21A2 with a risk of 3.6%” (p.I172N) with an MLPA method, or 8.5% for p.Q318X. From our experience, we believe that such assertions are incorrect. Indeed, it is not possible to establish the CYP21A2 copy number from the signal of only 1 specific gene probe (the exon 4 or exon 8 CYP21A2 probe in this case). When MLPA analysis is performed, it is necessary to consider the value of all specific gene probes (2 ). If CYP21A2 duplication were actually present, all 5 specific probes would show a ratio 1.3. That occurs when all CYP21A2 alleles are wild type for the following mutations: exon 3 8-bp deletion, p.I172N, clusterE6, and p.Q318X. On the contrary, if a duplicated CYP21A2 allele is present and carries the p.Q318X mutation, only the exon 8 probe would show a typical ratio (0.7–1.3) (2 ). Furthermore, if the p.I172N or p.Q318X mutation is absent in CYP21A1P, the CYP21A2specific probes for exon 4 or exon 8 would also bind the wild-type CYP21A1P pseudogene sequence, giving a ratio 1.3. For this reason and given the data of Cantürk et al., we think it more correct to affirm that exon 4 and exon 8 CYP21A2 probes may give a ratio 1.3 in 3.6% and 8.5% of the cases, respectively. Therefore, when that occurs, it is incorrect to immediately assume gene duplication, but it is necessary to consider the value of all specific CYP21A2 probes for a definitive interpretation of the MLPA analysis. In this case, pseudogene sequencing can also be performed to strengthen the diagnosis. Quantitative CYP21A2 diagnosis could provide an incorrect result when the CYP21A1P pseudogene lacks all of the following mutations: Del8bp, p.I172N, ClusterE6, and p.Q318X. In fact, in this case all specific CYP21A2 gene probes might also recognize the pseudogene sequence showing a ratio 1.3. To the best of our knowledge, that has never been reported. Our group first used MLPA analysis for the diagnosis of congenital adrenal hyperplasia (2). Other groups with considerable experience have used this technique successfully (3, 4), and it is currently used for routine analysis of congenital adrenal hyperplasia as a valid alternative to the Southern blot. As previously reported (2), we confirm that MLPA analysis is a very informative tool for the molecular diagnosis of congenital adrenal hyperplasia. Considering also the findings of Cantürk et al., however, we stress that the use of this methodology requires a deep knowledge of CYP21A2 genetics. The CYP21A2 gene, which encodes a 21hydroxylase, has a complex structure and is considered one of the most polymorphic of human genes.