MLPA is a practical and complementary alternative to CMA for diagnostic testing in patients with autism spectrum disorders and identifying new candidate CNVs associated with autism

Background: Autism spectrum disorder (ASD) is a complex heterogeneous developmental disease with a significant genetic background that is frequently caused by rare copy number variants (CNVs). Microarray-based whole-genome approaches for CNV detection are widely accepted. However, the clinical significance of most CNV is poorly understood, so results obtained using such methods are sometimes ambiguous. We therefore evaluated a targeted approach based on multiplex ligation-dependent probe amplification (MLPA) using selected probemixes to detect clinically relevant variants for diagnostic testing of ASD patients. We compare the reliability and efficiency of this test to those of chromosomal microarray analysis (CMA) and other tests available to our laboratory. In addition, we identify new candidate genes for ASD identified in a cohort of ASD-diagnosed patients.

Method: We describe the use of MLPA, CMA, and karyotyping to detect CNV in 92 ASD patients and evaluate their clinical significance.

Result: Pathogenic and likely pathogenic mutations were identified by CMA in eight (8.07% of the studied cohort) and 12 (13.04%) ASD patients, respectively, and in eight (8.07%) and four (4.35%) patients, respectively, by MLPA. The detected mutations include the 22q13.3 deletion, which was attributed to ring chromosome 22 formation based on karyotyping. CMA revealed a total of 91 rare CNV in 55 patients: eight pathogenic, 15 designated variants of unknown significance (VOUS)-likely pathogenic, 10 VOUS-uncertain, and 58 VOUS-likely benign or benign. MLPA revealed 18 CNV in 18 individuals: eight pathogenic, four designated as VOUS-likely pathogenic, and six designated as VOUS-likely benign/benign. Rare CNVs were detected in 17 (58.62%) out of 29 females and 38 (60.32%) out of 63 males in the cohort. Two genes, DOCK8 and PARK2, were found to be overlapped by CNV designated pathogenic, VOUS-likely pathogenic, or VOUS-uncertain in multiple patients. Moreover, the studied ASD cohort exhibited significant (p < 0.05) enrichment of duplications encompassing DOCK8.

Conclusion: Multiplex ligation-dependent probe amplification and CMA yielded concordant results for 12 patients bearing CNV designated pathogenic or VOUS-likely pathogenic. Unambiguous diagnoses were achieved for eight patients (corresponding to 8.7% of the total studied population) by both MLPA and CMA, for one (1.09%) patient by karyotyping, and for one (1.09%) patient by FRAXA testing. MLPA and CMA thus achieved identical reliability with respect to clinically relevant findings. As such, MLPA could be useful as a fast and inexpensive test in patients with syndromic autism. The detection rate of potentially pathogenic variants (VOUS-likely pathogenic) achieved by CMA was higher than that for MLPA (13.04% vs. 4.35%). However, there was no corresponding difference in the rate of unambiguous diagnoses of ASD patients. In addition, the results obtained suggest that DOCK8 may play a role in the etiology of ASD.