Flanking-Sequence Exponential Anchored (FLEA) PCR - a Sensitive and Highly Specific Method for Detecting Retroviral Integrant-Host-Junction Sequences.

Retroviral vectors integrate in the host cell genome and are widely used for gene therapy. Since integration is largely random, each site represents a unique integration event and can be used to study clonal evolution of engrafted cells. Integration can however be deleterious, and has been associated with oncogenesis. To date, PCR methods for studying the integrant-host junction (IHJ) have been laborious, unspecific and difficult to perform on complex samples. We have developed a method that is simpler and highly specific. We generate libraries of integrant-host junctions which can then be resolved by sequencing. The approach works even for analyses of highly complex samples. 1) Linear PCR is performed on sample DNA with biotinylated primer annealing to the 5′ end of the LTR. The subsequent longer biotinylated linear PCR fragments which contain IHJ or internal junctions are separated from genomic DNA with streptavidin paramagnetic beads. 2) We add an anchor primer with a known 5′ sequence and several degenerate 3′ bases. 3) A complimentary strand from this primer to the LTR is generated with T7 DNA polymerase. 4) A blocking oligo is added to the polymerase mixture, which anneals just before the 3′LTR of the internal junction and prevents the complementary strand from the internal sequence being generated, so that >90% of the anchored DNA contains external flanking genomic sequence. 5) We PCR amplify the material from a nested primer in the LTR to a primer, which anneals to the known sequence of the anchor primer. 6) Finally, a smear is generated which contains randomly anchored IHJs. FLEA-PCR was designed so that the most 5′ 33bp of the LTR (which contain an NheI site) are unprimed allowing easy validation of PCR product. Sequencing of these smears reveals that >99% of these DNA fragments contain this 33bp of 5′ extremity and are preceded by genomic sequences. With blocking, 90% of sequences are external IHJs. To validate the technique, 6 HeLa cell clones containing a single integrant were studied. The 5′ IHJ was characterized by FLEA-PCR amplification. The 3′ IHJ was successfully amplified by PCR with a primer designed on prediction from the human genome sequence and the 4bp repeat of junctional sequences was evident in all HeLa clones. Clonable smears can be generated with samples containing <0.1% transduced cells. The mean length of flanking genomic sequence is 97bp (ranging from 0 to 674bp). Digestion of smears amplified with a 6FAM-labelled LTR primer with CviJI and capillary electropheresis analysis allows precise counting of sample complexity. More accurately, random sequencing of smear can be assembled by custom written software. It is possible to study samples with over 80 different integrants in a single experiment. There are several advantages of using FLEA-PCR over restriction digest based methods. There is less PCR artifact. Since integrants yield varying fragments of random length, there is no PCR bias caused by differing amplicon lengths. This is of particular benefit when a short internal band can "swamp" the reaction. Secondly, no bias is accrued due to the frequency of occurrence of a particular restriction site. Finally, the technique is highly compatible with automation and high-throughput sequencing. This approach should facilitate IHJ analysis for assessing both retroviral safety and the biologic fate of transduced cells.