Multiple displacement amplification enables large-scale clonal analysis following retroviral gene therapy

Abstract

Analysis of the fate of retrovirally transduced cells after transplantation is often hampered by the scarcity of available DNA. We evaluated a promising method for whole-genome amplification, called multiple displacement amplification (MDA), with respect to even and accurate representation of retrovirally transduced genomic DNA. We proved that MDA is a suitable method to subsequently quantify engraftment efficiencies by quantitative real-time PCR by analyzing retrovirally transduced DNA in a background of untransduced DNA and retroviral integrations found in primary material from a retroviral transplantation model. The portion of these retroviral integrations in the amplified samples was 1.02-fold (range 0.2, to 2.1-fold) the portion determined in the original genomic DNA. Integration site analysis by ligation-mediated PCR (LM-PCR) is essential for the detection of retroviral integrations. The combination of MDA and LM-PCR showed an increase in the sensitivity of integration site analysis, as a specific integration site could be detected in a background of untransduced DNA, while the transduced DNA made up only 0.001%. These results show for the first time that MDA enables large-scale sensitive detection and reliable quantification of retrovirally transduced human genomic DNA and therefore facilitates follow-up analysis in gene therapy studies even from the smallest amounts of starting material.

FIG. 1.

Experimental design. LTR, long terminal repeat of retroviral vector; gDNA, genomic DNA; +MDA, with MDA; −MDA, without MDA.

FIG. 2.

Proportion of MDR1 transgene for comparison of fold changes of amplified to unamplified retrovirally transduced DNA. The ratio of proviral MDR1 to the endogenous hEpoR gene was determined in unamplified and amplified genomic DNA by QRT-PCR. The hEpoR gene was used to measure the total amount of DNA in the samples of the dilution series. Each gray bar represents the relative proportion of MDR1 in amplified DNA in comparison to the unamplified DNA template (black bars). This value was normalized to 1. The percentages of proviral MDR1 in relation to the hEpoR gene are tabulated for the same unamplified and amplified genomic DNA samples. For dilution series, transduced genomic DNA of cell line clone 2 was mixed with different amounts of untransduced HT1080 DNA from 50% to 0.1% (e.g., 25/75 means 25% transduced DNA and 75% untransduced background DNA), followed by amplification using the Repli-g kit. For all reactions, water served as a “no-template” control and untransduced HT1080 DNA as a “no-amplification” control. Error bars indicate standard deviations.

FIG. 3.

Detection of highly diluted transduced DNA after amplification followed by LM-PCR. The upper arrow indicates the specific integration site (external band [EB]; cell line clone 1). Mock, LM-PCR of untransduced DNA; NC (negative control), LM-PCR performed with water as template. The internal band (IB) originates from the 3′ long terminal repeat serving as a general control in LM-PCRs, which is the same for all SF1m vector-transduced cells. The disproportion between dilution factor and band intensity is not due to insufficient MDA but must be explained by specific characteristics of the LM-PCR method.

FIG. 4.

Improvement of sensitivity of integration analyses by MDA in the NOD/SCID mouse. (A) As shown in the scheme, 100 ng of genomic DNA of BM cells from a NOD/SCID mouse was used for MDA (right side) followed by three LM-PCRs. If the LM-PCR were performed by using 100 ng, no integration sites would be detectable, since this amount is not sufficient for performing LM-PCR with the NOD/SCID mouse BM samples used. Therefore, 2.5 μg starting material in the reaction without MDA was compared with 100 ng starting material in the reaction with MDA. The table shows the numbers of insertion sites and unique insertion sites without MDA and with MDA. (B) LM-PCR was performed with 2.5 μg of unamplified material (lane 1) and 2.5 μg of amplified material (starting material used for MDA, 100 ng) (three different reactions in lanes 2 to 4). The five distinct insertion sites of the LM-PCR detected in the unamplified material are marked by arrowheads, and the additional insertion sites after LM-PCR with amplified material are marked by diamonds (only once per band even if can be detected in different lanes). The internal band (IB) is indicated. In lanes 5 and 6 controls were loaded with water as template. The diffuse bands seen here are the primers. The data demonstrate that a multiplicity of LM-PCRs is superior to only one LM-PCR in terms of integration sites found.

FIG. 5.

Individual clone sizes of NOD/SCID repopulating retrovirally transduced CD34⁺ mobilized peripheral blood progenitor cells quantified in unamplified and amplified genomic DNA. Clone sizes relate to the amount of transduced cells in the samples, as measured by the proviral MDR1 cDNA. The error bars, representing standard deviations, were generated from the mean values from two QRT-PCRs performed in triplicate.