SeqURE – a new copy-capture based method for sequencing of unknown Retroposition events

Abstract

Background Retroelements (REs) occupy a significant part of all eukaryotic genomes including humans. The majority of retroelements in the human genome are inactive and unable to retrotranspose. Dozens of active copies are repressed in most normal tissues by various cellular mechanisms. These copies can become active in normal germline and brain tissues or in cancer, leading to new retroposition events. The consequences of such events and their role in normal cell functioning and carcinogenesis are not yet fully understood. If new insertions occur in a small portion of cells they can be found only with the use of specific methods based on RE enrichment and high-throughput sequencing. The downside of the high sensitivity of such methods is the presence of various artifacts imitating real insertions, which in many cases cannot be validated due to lack of the initial template DNA. For this reason, adequate assessment of rare (< 1%) subclonal cancer specific RE insertions is complicated. Results Here we describe a new copy-capture technique which we implemented in a method called SeqURE for Sequencing Unknown of Retroposition Events that allows for efficient and reliable identification of new genomic RE insertions. The method is based on the capture of copies of target molecules (copy-capture), selective amplification and sequencing of genomic regions adjacent to active RE insertions from both sides. Importantly, the template genomic DNA remains intact and can be used for validation experiments. In addition, we applied a novel system for testing method sensitivity and precisely showed the ability of the developed method to reliably detect insertions present in 1 out of 100 cells and a substantial portion of insertions present in 1 out of 1000 cells. Using advantages of the method we showed the absence of somatic Alu insertions in colorectal cancer samples bearing tumor-specific L1HS insertions. Conclusions This study presents the first description and implementation of the copy-capture technique and provides the first methodological basis for the quantitative assessment of RE insertions present in a small portion of cells.

The Absence of Retroelement Activity Is Characteristic for Childhood Acute Leukemias and Adult Acute Lymphoblastic Leukemia

“…However, the exact sensitivity of the commonly used methods in this respect is unknown. Recently, we directly measured the sensitivity of our selective amplification-based approach for somatic Alu insertions identification [24]. For this purpose, we mixed cells from four healthy individuals in different proportions (ranging from 1:100 to 1:1000), prepared libraries containing Alu flanks, sequenced them, and used polymorphic Alu insertions that were present in individual genomes to evaluate the potential of the method to detect insertions that are present in 0.1-1% of cells in a sample.…”

“…Peripheral blood mononuclear cells (PBMC) of two healthy donors were isolated from peripheral blood by a standard Ficoll-Paque (PanEco, Moscow, Russia) centrifugation protocol. Polymorphic L1 insertions that were specific for the donors were identified as described previously [24]. Briefly, we isolated the gDNA from PBMC of individual 1 and individual 2, prepared L1 flank libraries separately without mixing the cells, and sequenced them (MiSeq 2 × 150 233,390 reads for individual 1 and 170,120 for individual 2).…”

“…Tumor genomic DNA (gDNA) was isolated from bone marrow or peripheral blood with 23-99% (median 82%, all the samples except for three contained more than 50% of tumor cells) blasts in the disease onset, whereas normal gDNA of the same patient was obtained from either buccal swabs (in adult ALL cases) or bone marrow in complete (molecular MRD negative) remission (childhood ALL and AML). For library preparation, we used a modification of a previously published method [24,25]. The design of the method is shown in Figure 1.…”

“…We also used RNA-Seq data that was generated in our lab for 15 childhood ALL and 9 colorectal cancer samples. In 3 out of 9 colorectal cancer samples, we previously identified tumor-specific L1 insertions using our method [24]. Additionally, we downloaded 7 normal bone marrow [28] and 6 chronic lymphocytic leukemia [29] whole-transcriptome data that were generated using the same approach (i.e., RNA-Seq of polyA fraction; the accession numbers GSE162427 and E-MTAB-1733, respectively).…”

Mobile retroelements induced by hypomethylating agents are restricted to transpose in myeloid malignancies

“…However, as we expected that novel L1 insertions could be present in minor subpopulations and the whole genome sequencing methods would have limited sensitivity to detect them, we applied a targeted-NGS-based approach with the ability to capture retrotransposition events in 1% of cells (26). This approach was previously used to demonstrate somatic retrotranspositions in brain tissue (37) and colorectal cancer (29) and the absence of RE activity in leukemia (26). Thus, we assume that the sensitivity of the NGS method was not the reason for the negative result.…”