Allele-specific and multiplex PCR based tools for cost-effective and comprehensive genetic testing in Congenital Adrenal Hyperplasia

Congenital Adrenal Hyperplasia (CAH) is an autosomal recessive disorder due to enzyme defects in adrenal steroidogenesis. Several genes code for these enzymes, out of which mutations in the CYP21A2 gene resulting in 21 hydroxylase deficiency, contribute to the most common form of CAH. However, pseudogene imposed challenges complicate genotyping CYP21A2 gene, and there is also a lack of comprehensive molecular investigations in other genetic forms of CAH in India. Here, we describe a cost-effective, highly specific, and sensitive Allele Specific PCR (ASPCR) assay designed and optimized in-house to screen eight common pathogenic mutations in the CYP21A2 gene. We have also established and utilized a multiplex PCR assay for target enrichment and Next-generation sequencing (NGS) of CYP11B1, CYP17A1, POR, and CYP19A1 genes. Following preliminary amplification of the functional gene CYP21A2, ASPCR based genotyping of eight common mutations - P30L, I2G, 8BPdel, I172N, E6CLUS (I235N, V236E, M238K) V281L, Q318X, and R356W was carried out. These results were further validated using Sanger and Next-generation sequencing. Once optimized to be specific and sensitive, the advantage of ASPCR in CYP21A2 genotyping extends to provide genetic screening for both adult and paediatric subjects and carrier testing at a low cost and less time. Furthermore, multiplex PCR coupled NGS has shown to be cost-effective and robust for parallel multigene sequencing in CAH.


DNA extraction and long-range PCR
DNA extraction was carried out with 2 ml EDTA whole blood using Gentra Puregene kit from QIAGEN® (Hilden, Germany) and quantified using NanoDrop TM spectrophotometer. Long-range PCRs were utilized for locus-specific amplification of the functional gene CYP21A2 (6.2 kbp) and pseudogene CYP21A1P (6.1 kbp) with TaKaRa LA PCR TM Kit (ver.2.1) using previously published protocols [1] . In addition, the results were validated with TaqI restriction digestion [1] . Based on these results, samples suspected for large 30 kbp deletion and large gene conversion were validated with MLPA and additional long-range PCRs with specific primers for these rearrangements described previously by Greene et al [2] . Figs. 1 and 2 show results of locus-specific amplification and restriction digestion of the above genes and their interpretation in identifying rearrangements

Allele Specific PCR (ASPCR) for screening eight hotspot mutations in CYP21A2 gene
The long-range PCR product of the CYP21A2 gene was utilized as a template for Allele Specific PCR (ASPCR) to genotype eight common hotspot mutations in the CYP21A2 gene. ASPCR, a modified application of conventional PCR technique, is a strategy to detect point mutations and small deletions by deliberately introducing mismatches in the primers. Primer designing is crucial in ASPCR to generate detectable amplicons from the mutation target while minimizing false priming at the nontarget allele. A wild-type (WT) primer complementary to the normal sequence is designed for each target sequence harboring the hotspot mutation. A mutant (MT) primer complementary to the 3' terminal base of the mutation under study is also designed for the same target. A reverse primer common for both the WT and MT is designed to maintain the same size for both WT and MT products. The wild type primer will provide amplification only with the wild type allele and there is no amplification when the allele is mutated.
Similarly, a mutant primer can amplify only the DNA sequence that carries the mutation. This enables the identification of the hotspot mutation under simple PCR conditions. Mismatches at the penultimate bases are often intentionally added to increase the specificity of the ASPCR [3] . If the terminal destabilization is weak, a strong destabilizing mismatch is added at the penultimate base and vice versa with a strong destabilization at the terminal base. Two WT forward primers were designed for I2G splice site mutation, including two WT alleles A and C. For all the eight hotspot mutations, common internal control primers were designed in such a way, it is amplified with both WT and MT alleles.   [1] results with TaqI in large gene conversion on 1% agarose gel electrophoresis -Lane 1 and 2 shows normal restriction digested fragments of functional and pseudogene amplified with long range PCR in the negative control. In one subject, there was no amplification with functional gene primers CYP779f/Tena36F2, and so there were no digested products as seen in lane 3. However, there was amplification with the pseudogene primers (CYP779f/ XA-36F) with a restriction digestion pattern similar to the functional gene, as shown in lane 4. This suggests a homozygous large gene conversion involving the proximal end of CYP21A2 and the distal end of CYP21A1P genes. b) Restriction digestion results in large 30 kbp deletion on 1% agarose gel electrophoresis: Lane 4 and 5 show normal restriction digested fragments of the functional and pseudogene in the negative control. Lane 1 shows a restriction digestion pattern of a sample with homozygous 30 kbp deletion. Since the deletion involves forming a chimeric (fusion) gene with the proximal end of CYP21A1P and the distal end of CYP21A2 genes, there is no amplification with pseudogene primers (CYP779f/ XA-36F). However, the product amplified with functional gene primers (CYP779f/Tena36F2) gives a restriction digestion pattern similar to that of the pseudogene. A heterozygous 30 kbp deletion on one allele results in amplification with both the primer sets, but the product from functional gene primers produces three restriction digestion bands resulting in a combination of functional and pseudogene, as seen in lane 2.

Pre-clean up
The long-range PCR product of the CYP21A2 gene is purified using Agencourt AMPure XP (Beckman Coulter Life Sciences, USA) magnetic beads with the following protocol. This cleaned up product is used as a template for ASPCR

Standardization of ASPCR conditions
The ASPCR was in house standardized with Emerald Amp® Max PCR master mix (Takara Bio Inc, Japan) in 15 μl reaction volume. The primer sequences are given below in table 1 . Optimal annealing temperature and template concentration were utilized with appropriate positive and negative controls, and the below conditions were finalized to achieve optimal results. P30L hotspot mutation required primer redesigning to overcome false-positive results. Change in DNA extraction techniques can also affect the specificity of ASPCR and might require further standardization of the template concentration used. Details of the ASPCR reaction mix and program are mentioned in Tables 2a and 2b Following this, samples were screened for all the hotspot mutations with mutant primers, and the results were also validated with Sanger and NGS sequencing ( Fig. 3 ).

MLPA and ASPCR in identifying chimeric genes
Large 30 kbp deletion in 21 -hydroxylase deficiency results in the formation of chimeric genes involving the proximal end of CYP21A1P and the distal end of CYP21A2 genes. MLPA (Multiplex Ligation-dependent Probe Amplification) is the most common technique employed in molecular analysis of large deletions and duplications in routine clinical practice. In this study, we utilized MLPA to validate large 30 kbp deletion suspected from the results of long-range PCR and restriction digestion using SALSA MLPA CAH Probemix P050 C1 from MRC-Holland [4] . Simultaneously allelespecific PCR was also carried out. Results of some of these samples are discussed below in Fig. 4 .
The junction sites to classify classical and attenuated chimeras depend on the series of deleterious pseudogene mutations present in the extent of rearrangement. However, CYP21A2 probes in the        [5] . c) A sample positive for heterozygous 30 kbp deletion with a copy number of 0.5 in several CYP21A2 probes. The black arrowheads indicate the copy number of MLPA probes for Intron 2 splice site -each for wild type allele C and A to be zero with DQ of 0 and 0.09, respectively. With ASPCR, this subject was heterozygous for P30L, 8BPdel, I172N, E6CLUS and V281L and homozygous for I2G mutations. The parental screening revealed that the mother was a carrier for 30 kbp deletion and the father for the I2G splice variant. These results indicate that the subject is heterozygous for 30 kbp deletion with chimeric gene CH5 [5] on one allele and I2G splice mutation on the other allele.
utilized MLPA assay span only till exon 7 out of 10 (probe: CYP21A2-7(WT) wt F306 + T). But Q318X and R356W probes are also required to identify chimeras CH3 and CH8. Therefore, the ASPCR, including these mutations, is advantageous to identify the above chimeras.

Multiplex PCR based target enrichment for NGS testing in CAH
A multiplex PCR program was designed to comprehensively screen for CYP21A2, CYP11B1, CYP17A1 and POR genes in CAH along with the CYP19A1 gene that causes aromatase deficiency mimicking CAH.   Table 4 The coding and splice site regions of four genes -CYP11B1, CYP17A1, POR and CYP19A1 were amplified in 28 amplicons in 6 groups. Primers for the CYP11B1 gene were adapted from white et al. [6] . The primers were pooled into six groups based on the amplicon sizes ( Table 3 and 4 ). The multiplex PCR was carried out using QIAGEN® Multiplex PCR kit. The PCR reaction mix and the conditions are described in table 5 and 6 respectively. The concentration of primers used was 10 pmol/μl. The PCR products were visualized on 2% agarose gel electrophoresis ( Fig. 5 .). Multiplex PCR products were    pooled along with the long range PCR product of CYP21A2 gene and sequencing with Ion Torrent PGM TM was performed following methods from published protocols [7] . Multiplex PCR coupled NGS sequencing achieved a uniform coverage across five genes with an average base coverage depth of 700X and with > 99% of the target having 20X coverage ( Table 7 and Fig. 6 .).
With the above comprehensive strategy, clinically significant variants were identified in CYP21A2, CYP11B1 and CYP19A1 genes in 97.2% of the study subjects (n = 72) suspected for 21 hydroxylase and 11 beta hydroxylase deficiency. No disease-causing variants were identified in CYP17A1 and POR genes. However, several polymorphisms were identified in the above two genes ( table 8 ) indicating effective use of this CAH -NGS panel in clinical settings. Table 7 Target coverage summary generated from Ion torrent coverage analysis plugin for CAH -5 gene panel with 29 amplicons. The coverage of amplicons 17&18 and 19&20 are merged. Fig. 6. a. Coverage analysis report of a representative sample sequenced for CYP21A2 gene with 100% of the target having a minimum coverage of 20X reads. b . Coverage analysis report of a representative sample sequenced for five genes CAH panel in 29 amplicons with 99.72% of the target having a minimum coverage of 20X reads and 99.44% of the target with 100X reads.

Conclusion
The ASPCR assay was found to be highly specific and sensitive to detect all eight hotspot mutations in CYP21A2 gene that were also identified by NGS and Sanger sequencing, validating its sensitivity and specificity. This assay is a simple cost-effective technique to genotype point mutations in CYP21A2 gene and to identify junction sites in chimeric genes of CYP21A2 -CYP21A1P rearrangement that contributes to more than 90% of mutations in 21 -hydroxylase deficiency. Careful standardization enabled accurate and precise results that can provide a genetic diagnosis to a significant proportion of the CAH cohort in a clinical setting. The multiplex PCR assay enables a cost-effective step in NGS processing of CAH genes achieving uniform coverage matrices across the genes.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.