AFLP-AFLP in silico-NGS approach reveals polymorphisms in repetitive elements in the malignant genome

Abstract

The increasing interest in exploring the human genome and identifying genetic risk factors contributing to the susceptibility to and outcome of diseases has supported the rapid development of genome-wide techniques. However, the large amount of obtained data requires extensive bioinformatics analysis. In this work, we established an approach combining amplified fragment length polymorphism (AFLP), AFLP in silico and next generation sequencing (NGS) methods to map the malignant genome of patients with chronic myeloid leukemia. We compared the unique DNA fingerprints of patients generated by the AFLP technique approach with those of healthy donors to identify AFLP markers associated with the disease and/or the response to treatment with imatinib, a tyrosine kinase inhibitor. Among the statistically significant AFLP markers selected for NGS analysis and virtual fingerprinting, we identified the sequences of three fragments in the region of DNA repeat element OldhAT1, LINE L1M7, LTR MER90, and satellite ALR/Alpha among repetitive elements, which may indicate a role of these non-coding repetitive sequences in hematological malignancy. SNPs leading to the presence/absence of these fragments were confirmed by Sanger sequencing. When evaluating the results of AFLP analysis for some fragments, we faced the frequently discussed size homoplasy, resulting in co-migration of non-identical AFLP fragments that may originate from an insertion/deletion, SNP, somatic mutation anywhere in the genome, or combination thereof. The AFLP-AFLP in silico-NGS procedure represents a smart alternative to microarrays and relatively expensive and bioinformatically challenging whole-genome sequencing to detect the association of variable regions of the human genome with diseases.

Fig 1. Workflow of newly established AFLP/NGS procedure for AFLP markers identification.

Fig 2. Hierarchical cluster analysis of all healthy controls and CML patients.

Dendrogram illustrating two distant clusters of healthy donors (N = 30; green) and patients with CML (N = 65; red) was constructed based on the distance matrix calculated using the unweighted neighbor-joining method in DARwin software v6.0 [24]. The calculation was done with all 3912 AFLP markers detected after fragmentation analysis in all samples of healthy donors and CML patients.

Fig 3. High-scoring segment pair distribution of the CTT_ACA_57 sequence in the human genome.

The sequence of CTT_ACA_57 is a part of the variable satellite sequence ALR/Alpha, occurring in the centromeric and pericentromeric regions of several chromosomes. A red box on chromosome 5 represents the best hit.

Fig 4. Three different situations regarding the presence/absence of rs11428156 (-/A/G/GG) in the AFLP marker CAC_AAG_063.

The presence/absence of rs11428156 (-/A/G/GG) was analyzed in the spectra obtained from fragment analysis and the sequence obtained through Sanger sequencing. Panels A and B show the presence of a 195-bp fragment resulting from the homozygous or heterozygous deletion of adenine, respectively, whereas panel C shows the wild-type sample. Panels D, E, and F show a portion of the sequences with the rs11428156 SNP, evaluated using the software Mutation Surveyor v3.10 (Softgenetics, State College, PA, USA), corresponding to the fragment analysis spectra. RFU—relative fluorescence unit.

Fig 5. The (G/T) substitution of SNP rs7906704 resulted in the introduction of an MseI restriction site, responsible for the formation of the CAT_ACC_55 fragment (175 bp).

The fragment is present (A, B) if the substitution is homozygous (D) or heterozygous (E). Panels C and F illustrate the absence of the fragment, as SNP rs7906704 was not detected. RFU—relative fluorescence unit.