| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
THE STEM CELL NICHE |
aResearch Group Pharmacology of Cancer Treatment, German Cancer Research Center, Heidelberg, Germany;
bDepartment of Experimental Surgery Mannheim Faculty, University of Heidelberg, and Molecular Oncology of Solid Tumors Unit, Deutsches Krebsforschungszentrum (DKFZ), Heidelberg, Germany;
cHematology Branch, National Heart, Lung and Blood Institute, NIH, Bethesda, Maryland, USA;
dInstitute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany;
eDepartment of Internal Medicine V, University of Heidelberg, Heidelberg, Germany;
fCenter for Tumor Diagnostics and Therapy, Paracelsus Klinik, Osnabrueck, Germany
Key Words. Gene therapy • Multidrug resistance 1 • CD34+ • Rhesus macaque
Correspondence: Stefan Fruehauf, M.D., Center for Tumor Diagnostics and Therapy, Paracelsus Klinik, Am Natruper Holz 69, 49076 Osnabrueck, Germany. Telephone: 49-541-9663040; Fax: 49-541-9663046; e-mail: prof.stefan.fruehauf{at}pk-mx.de
Received on January 25, 2007;
accepted for publication on June 18, 2007.
First published online in STEM CELLS EXPRESS July 5, 2007.
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
Disclosure of potential conflicts of interest is found at the end of this article.
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
Anxiety was raised regarding further use of MDR1 gene transfer in human clinical trials following the occurrence of a polyclonal myeloproliferative syndrome in mice after transplantation of marrow cells transduced with a Harvey murine sarcoma virus vector expressing MDR1 (HaMDR1) [21, 22]. In addition, more recent concerns have arisen regarding insertional mutagenesis, based on an interaction between the vector backbone, the transgene, the multiplicity of infection, and the host cells [23]. Therefore, additional studies addressing the safety and feasibility of MDR1 gene therapy with new transduction and conditioning protocols and in the relevant target cell population are required.
Here, we used a nonhuman primate model with direct relevance to human biology to investigate the safety of MDR1 gene transfer. A proliferative advantage of MDR1-transduced versus NeoR-transduced HSCs after 1 year of follow-up was not observed using semiquantitative polymerase chain reaction (PCR) methods [24]. In this work, we quantified for the first time using a fluorescence-based quantitative real-time PCR (QRT-PCR) [25] the contribution of MDR1-transduced HSCs to hematopoiesis after a 4-year follow-up not only on a bulk cell population but also on the level of single hematopoietic stem and progenitor cells.
With respect to the safety aspects of gene transfer, we analyzed the integration pattern of the G1Na and HaMDR1 vectors, regarding the occurrence of integrations in oncogenes. An additional aim was to investigate whether primate MDR1 gene transfer results in the dominance of transduced clones after prolonged periods of follow-up. Since we were not able to perform ligation-mediated PCR (LM-PCR) [26] on all samples because of limited DNA amounts, we established a new, highly sensitive PCR method for detection of vector integration sites using exponential fragment marking (EFRAM).
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
Blood samples were taken at different time points, from 8 weeks up to 4 years after transplantation (8, 46, and 50 weeks and 4 years after transplantation for monkey M120; 8, 12, 16, 23, 54, and 58 weeks and 4 years after transplantation for monkey M038) and separated into mononuclear cells (MNCs) and granulocytes [24]. The samples were analyzed with LM-PCR or EFRAM-PCR, depending on the DNA amount of the sample, and then quantified using QRT-PCR.
Detection of Vector Integration Sites
LM-PCR. LM-PCR and integration site analysis were performed as described previously [26]. Briefly, 2,500 ng of DNA was digested using restriction enzyme BsmAI (New England Biolabs, Ipswich, MA, http://www.neb.com); after purification of the digest a one-step amplification reaction was performed using the long terminal repeat (LTR)-specific biotinylated primer HaRV292bio (5'biotin-CAGATGCGGTCCAGCCCTCA-3'). After purification and enrichment for biotin-marked fragments with streptavidin-coated paramagnetic beads (Dynal, Oslo, Norway, http://www.dynalbiotech.com), adapter ligation was performed using a blunt-end adapter oligo-cassette (Clontech, Palo Alto, CA, http://www.clontech.com). Amplification of the fragments was performed using the primers HaRV292 (5'-CAGATGCGGTCCAGCCCTCA-3') and AP1 (5'-GTAATACGACTCACTATAGGGC-3'). Subsequently, nested PCR was performed using the primers HaRV111 (5'-TCCTGACCTTGATCTGAAC-3') and AP2 (5'-ACTATAGGGCACGCGTGGT-3'). All PCR products were cloned into the pCR4 plasmid vector (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) and then sequenced using an ABI Prism Genetic Analyzer 310 (PE Applied Biosystems, Weiterstadt, Germany, http://www.appliedbiosystems.com).
EFRAM-PCR. First, 100 ng of DNA was digested using 80 U of HhaI (New England Biolabs) for 1 hour at 37°C. After purification of the digest (Mini Elute reaction cleanup kit; Qiagen, Hilden, Germany, http://www1.qiagen.com), primers A1 (5'-AGCACTCTCCAGCCTCTCACGGCG-3') and A3 (5'-CCGTGAGA-3'), 0.45 nmol of each, were allowed to anneal at 55°C for 10 minutes, creating asymmetric adapter cassettes that were ligated to the digested DNA-fragments using 400 U of T4 DNA ligase under incubation for 16 hours at 16°C. T4 ligase was then inactivated at 65°C for 10 minutes. Next, PCR was performed on the ligated fragments using 15 pmol of the biotinylated LTR-specific primer HaRV292bio, 15 pmol of primer A1, 7.5 U of Taq polymerase, 64 nmol of dNTPs, and water to a final volume of 100 µl. Samples were placed in a preheated thermocycler, and a 29 cycle amplification reaction was started using following amplification conditions: 95°C, 1 minute; 63°C, 1 minute; 72°C, 3 minutes; final extension at 72°C for 10 minutes. After purification and paramagnetic enrichment of biotin-marked fragments (Kilobase binder kit; Dynal), PCR amplification of the enriched fragments was performed using the Expand High Fidelity PCR System (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) with 15 pmol of the LTR-specific primer HaRV292 and 15 pmol of A1 under the following conditions: 94°C, 2 minutes; 95°C, 15 seconds; 63°C, 1 minute; 72°C, 90 seconds; 68°C, 3 minutes; 30 cycles; final extension at 72°C for 7 minutes. The PCR-product was diluted 1:50, and then nested PCR was performed with 15 pmol of LTR-specific primer HaRV111 and 15 pmol of primer A1 under the conditions mentioned above, except with reduced cycle number (n = 28) and a primer annealing temperature of 55°C. All PCR products were cloned into the pCR4 plasmid vector (Invitrogen) and then sequenced using an ABI Prism Genetic Analyzer 310 (PE Applied Biosystems). A scheme showing the principle of the EFRAM-PCR is displayed in Figure 1A.
|
Assignment of Clone Identity. The first 110 bases of the LTR sequences of both vectors are identical, except for the base at position 68, which is thymidine in the HaMDR1-LTR and cytosine in the G1Na-LTR. To confirm that the likelihood of each vector being detected by LM-PCR and EFRAM-PCR was the same, we designed primers that aligned to this equal sequence on both vector LTRs and led to amplification of LTR sequences of equal length. Using the difference of the LTR sequences in position 68, we were able to definitely allocate each identified clone to one vector or the other after sequencing.
Integration Analysis. The sequences detected using LM-PCR and EFRAM-PCR were first viewed using Chromas 2.23 software (Technelysium Pty. Ltd., Helensville, Queensland, Australia, http://www.technelysium.com). An integration site was considered valid only when the sequence of interest was flanked by the LTR sequence on one side and the adapter sequence on the other side. Integration analysis was performed using the IntegrationMap task (available online at http://genius.embnet.dkfz-heidelberg.de/menu/biounit/open-husar or for all services after registration for an account at http://genome.dkfz-heidelberg.de), as described by Giordano et al. [27]. Briefly, IntegrationMap first runs a MEGABLAST analysis using the NCBI 35b assembly as described [28]. The location (i.e., information about chromosomal position, contig position, and strand) is then computed. Using an application program interface of Ensembl.org (http://www.ensembl.org/Homo_sapiens), the positions of the next gene (i.e., its transcription start site), the next repetitive elements (e.g., short interspersed nucleotide elements [SINE], long interspersed nucleotide elements [LINE], simple repeats, and LTR elements), and the next CpG island are computed, and their distances to the exact integration base are calculated. If the integration position is localized within a gene, the exon or intron number is also displayed.
Real-Time Quantitative PCR
We used a fluorescence-based real-time quantitative PCR [25]. Absolute quantitation of input DNA was accomplished by generating a standard curve with serial dilutions of the plasmid-cloned target DNA and then comparing the signals obtained from samples with unknown amounts of vector or clone-specific DNA to the standard curve (User's Manual, ABI Prism 7700 Sequence Detection System; PE Applied Biosystems). Concentration of each plasmid solution was determined as a mean of three A260 photometric measurements and converted to the number of copies using the molecular weight of the plasmid. We made serial dilutions of the plasmids in the range of 100–105 copies per microliter. To minimize the amount of plasmid DNA absorbed to the wall of the tube, a constant amount of carrier DNA (salmon sperm DNA, 10 ng/µl) was added to the diluted plasmids. Dilutions were performed once, and sufficient numbers of aliquots of each dilution of the chosen plasmid standard curve were stored at –20°C. All real-time reactions were performed using new aliquots to minimize degradation of DNA because of repeated freeze/thaw cycles. A dilution series was only considered as acceptable if the correlation coefficient of the standard curve was more than 0.98 and the Ct values increased by approximately 3 for each 10-fold dilution, as recommended by PE Applied Biosystems (User's Manual, ABI Prism 7700 Sequence Detection System). The Ct value is defined as the cycle number at which a significant increase in the fluorescence signal is first detected. The PCR efficiency was deduced from each standard curve by ABI Prism 7700 software and expressed as E = 10–1/s – 1, where s is the curve slope.
To assess the total copy number of G1Na and HaMDR1 vector DNA in a sample and the total amount of DNA in the sample, three reference standard curves were generated using the NeoR gene, the MDR1 gene [29], and the rhesus macaque β-2-microglobulin gene (rhB2M). The pcDNA3.1(–) plasmid (Invitrogen) and the SFβm1 plasmid [30] were used for the NeoR and the MDR1 dilution series, respectively. For the rhB2M standard curve, a part of exon 2 of the gene was PCR-amplified using the Expand High Fidelity PCR System (Roche Diagnostics) together with 15 pmol of each primer, 5'-GTACTCCAAAGATTCAGGTT-3' and 5'-CACTTAACTGTC-CTGGGC-3' (based on the rhB2M sequence, GenBank accession no. AY349163) [31], by using the following PCR regimen: 94°C, 15 seconds; 52°C, 30 seconds; 72°C, 90 seconds; 30 cycles.
To assess the copy number of a single clone, the standard curves were constructed by diluting plasmids containing the LM-PCR or EFRAM-PCR product of this clone [29]. This copy number is the absolute copy number of the clone, which can be displayed as the percentage of the clone to the total DNA and as the percentage of the clone to the overall gene modified population (supplemental online Fig. 1).
Primers and Probes. Primers and TaqMan 5' Fam/3' TAMRA probes were designed using Primer Express software (version 1.5; PE Applied Biosystems). Primers were as follows: NeoR, 5'-CTTGGGTGGAGAGGCTATTCG-3' and 5'-CCTCGTCCTGCAGTTCATTCA-3'; rhB2M, 5'-CAGGTTTACTCACGCCATCCA-3' and 5'-ACTTTTCCCATTTTCTCTCCATTC T-3'; and LTR, 5'-TCCTGACCTTGATCTGAACTTCTCT-3'. The following sequences were used as probes: NeoR probe, 5'-CAAGACCGACCTGTCCGGTGCC-3'; rhB2M probe, 5'-TCCAGACACATAGCAATT-CAGGAAATTTGGCT-3'; G1Na-LTR probe, 5'-CATGCCTTG-CAAAATGGCGTTACTTAAGCTAGC-3'; and HaMDR1-LTR probe, 5'-CATGCCTTGTAAAATGGCGTTACTTAAGCTAGC-3'. MDR1 primers and probe were designed as described previously [3]. Sequences of unique primers binding to the flanking rhesus DNA were designed for 18 integration site sequences.
PCR Amplification. QRT-PCR was performed on an ABI Prism 7700 Sequence Detection System instrument (PE Applied Biosystems). PCR amplification was performed as described before [29]. The reactions were performed two times in triplicate.
Cross-Reactions. To measure the specificity of each clone-specific reverse primer, a plasmid mix was generated for each clone using 450 ng of untransduced rhesus macaque DNA as a background and 10 copies of each plasmid containing an integration site except the one that served as template for the primer design. Then, QRT-PCR was performed on the amplification mix as described above, with the specific reverse primer for the clone missing in the plasmid mix. Untransduced rhesus macaque DNA served as a "no amplification" control (NAC), and H2O served as a "no template" control (NTC).
Statistical Analysis.
PCRs were performed two times in triplicate for a clone (A) and the vector (B). The means of the triplicates (A1 and B1) and the appropriate SEM (a1 and b1) in the first PCR run as well as the means of the triplicates (A2 and B2) and the appropriate SEM (a2 and b2) in the second PCR run on each sample were calculated. These means represent the absolute copy numbers of a clone or a vector, respectively. The relative proportion of a clone within the overall population of gene modified cells was estimated by the ratio of the absolute copy numbers of the clone and the vector for each PCR performance, that is, D1 = A1/B1 and D2 = A2/B2. From these two ratios, the average percentage of clone contribution D = 0.5 x (D1 + D2) was determined. The SE (d) for this average percentage D was calculated using the following formula: d = 0.5 x 
d12 + d22 with d1 = 1/B12 x A12 x b12 + B12 x a12 and d2 = 1/B22 x A22 x b22 + B22 x a22 [32].
The percentage of a clone's contribution to the total DNA, which represents the contribution of a clone to total primate hematopoiesis, and the percentage of the vector to the total DNA, which represents the overall gene modified population, were estimated similarly. We determined two different CV values: the intraexperimental CV, showing the degree of variation between the same samples (triplicates) in one experiment, and the interexperimental CV, showing the extent of variation in repeated experiments.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
Using monkeys M038 and M120, 47 x 106 and 34 x 106 CD34-enriched cells from mobilized PBCs, respectively, were obtained, transduced, and then expanded as described by Sellers et al. [24]. At the end of culture, the total cell number had expanded 40- and 37-fold for the HaMDR1-transduced cells and 47- and 67-fold for the G1Na-transduced cells [24]. These cells were then transplanted after total body irradiation [24].
Detection of Vector Integration Sites
LM-PCR is a suitable method to amplify multiple retroviral integration sites from at least 2,500 ng of genomic DNA [26]. Because of the limitation set by small amounts of DNA in our samples, we modified this method. After optimization of the EFRAM-PCR method on transduced cell line clones, we could detect 1 ng of transduced DNA in a background of 99 ng of untransduced DNA (Fig. 1B). The HT1080 cell line used has two vector integration sites per cell, which are located at chromosomes 1q41 and 3q21 (GenBank accession nos. AC092015 and AC083798, respectively). The bands bigger than 600 kb appearing in the samples with a low content of transduced DNA were nonspecific. This explains why with higher amounts of marked DNA they became less prominent: there were more specific binding sites available for the primers in the samples containing 10% and 100% transduced DNA, so nonspecific binding of the primers occurred less frequently.
Using LM-PCR and EFRAM-PCR, we detected a total of 122 unique contributing clones in granulocytes and mononuclear cells from 8 weeks up to 4 years after transplantation. Twenty samples were analyzed, 6 with LM-PCR and 14 with EFRAM-PCR. We used EFRAM-PCR only in samples with limited DNA availability. One hundred two clones contained the G1Na vector, whereas 20 contained the HaMDR1 vector. In monkey M120, the ratio of HaMDR1 clones to G1Na clones was 1/7.4 (8/59); in monkey M038, the ratio was 1/3.6 (12/43). The total number of identified clones for each monkey, vector, and PCR method is shown in Table 1. and supplemental online Table 1. The number of clones found at different time points for each animal is shown in Figure 2A (M120) and 2B (M038), which also shows the number of clones in granulocytes and in mononuclear cells (primary T-lymphocytes), respectively. The number of identified G1Na clones was at each time point higher than the number of identified HaMDR1 clones, regardless of the PCR method used.
|
|
The absolute quantitation assay was chosen to determine the copy number of the samples. Our standard curves showed a correlation coefficient of more than 0.98 and an increase of Ct values by approximately 3 for each 10-fold dilution (supplemental online Figs. 2, 3), as recommended by PE Applied Biosystems (User's Manual, ABI Prism 7700 Sequence Detection System). The mean intraexperimental/interexperimental coefficients of variation of the assay for the quantitation of the rhB2M, NeoR, and MDR1 genes were 0.19/0.22, 0.12/0.15, and 0.18/0.27, respectively.
|
The contribution of each vector transgene at different time points is shown in Figure 3A (M120) and 3B (M038). Eight weeks after transplantation, the percentages of G1Na-positive DNA in the granulocyte and MNC fractions for monkey M120 were 4.05% ± 0.63% and 7.18% ± 0.78%, respectively, descending to contribution levels at approximately 1% 4 years post-transplantation (Fig. 3A). In monkey M038, the percentage of G1Na-positive granulocyte DNA declined from 2.82% ± 0.24% 8 weeks after transplantation to 1.62% ± 0.12% 16 weeks after transplantation and was virtually constant at a level around 1.5% after week 16 (Fig. 3B). Although the data of MNCs show more fluctuations for monkey M038, the tendency is the same as in monkey M120: a decline of the percentages of G1Na-positive DNA from 4.57% ± 0.24% after 12 weeks to 0.93% ± 0.14% after 4 years. The percentage of HaMDR1-positive granulocyte DNA was at a constant level of less than 1% for 4 years in both monkeys (range, 0.18% ± 0.02% to 0.27% ± 0.05% in monkey M120 and 0.43% ± 0.05% to 0.19% ± 0.04% in monkey M038); the percentage of HaMDR1-transduced MNCs declined from an initial value of approximately 1.5% and converged thereafter to the granulocyte level in both monkeys (range, 1.26% ± 0.16% to 0.34% ± 0.04% in monkey M120 and 1.42% ± 0.09% to 0.40% ± 0.06% in monkey M038) (Fig. 3A, 3B).
Quantitation of Contributions of Single Clones
To measure the contribution of individual transgene-marked CD34+ cells to hematopoiesis, we performed QRT-PCR on single clones. The validation of this method for the approach of quantification of single clones was performed by quantifying an individual plasmid in a plasmid mixture [29]. Three plasmids containing different LM-PCR product were mixed in certain ratios, and then the proportion of an individual plasmid was quantified. Nagy et al. [29] were able to detect the proportion of an individual plasmid down to 10 copies in a total plasmid number of 105. The mean intraexperimental and interexperimental coefficients of variation of our assay for the quantitation of the single clones were 0.41 and 0.35, respectively. This loss of precision in the quantitation of such low values is explained by the actual principle of the PCR methodology, with a variation as low as 1 Ct, leading to 100% variation in the final quantification. Considering this, it is important to emphasize that the use of QRT-PCR at such low quantities is only suitable to measure a range of values for a clone in which it most likely appears. Furthermore, since in some cases the Ct values of the samples approached the no-amplification Ct value of 45, we included only reactions in which the NTC and the NAC were absolutely negative with a Ct value of 45.
The mean PCR efficiency for the clones was 83.76% ± 1.06% (range, 79.54% to 87.69%); the deviation of approximately 10% remains in the scope of the SEM. Thus, the PCR efficiencies allow a direct comparison of the copy numbers. Repression of the PCR efficiency due to nonspecific retroviral common primer annealing was excluded using a clone-specific primer and probe set for clone 1. The PCR efficiency for this individual set was 84.37% ± 2.11%; the correlation coefficient of the standard curve was 0.99. The clone could not be detected with this specific set.
We analyzed 18 (16 G1Na and 2 HaMDR1 clones) of the 24 clones identified in both monkeys 4 years after transplantation. The remaining six clones could not be quantified because of difficulty identifying non-cross-reacting primers. The quantitation of clones found at earlier time points was not possible because of limited DNA availability.
Ten of the 16 G1Na clones analyzed were not detectable and thus contributed at levels less than the QRT-PCR detection limit of 0.0001% [29]. The remaining six clones had copy numbers between 1.43 ± 0.52 and 7.1 ± 0.34, equivalent to 0.09% ± 0.03% and 0.42% ± 0.03% of G1Na-positive DNA. Two HaMDR1 clones were analyzed, showing copy numbers of 3.52 ± 1.11 and 2.87 ± 0.56, respectively, equivalent to 0.71% ± 0.23% and 0.84% ± 0.2% of HaMDR1-positive DNA. An overview of the proportion of all quantified clones is given in Table 2. The individual clones contributed at levels of 0.0008% ± 0.0003% to 0.0041% ± 0.00032% to the total primate hematopoiesis, calculated on the basis of rhB2M gene copies (Table 2; Fig. 4).
|
|
We found five integrations in gene loci involved in malignant transformation of cells, three times with the G1Na vector and two times with the HaMDR1 vector. One integration of the G1Na vector occurred in the TIAM1 (T-cell lymphoma invasion and metastasis one) oncogene, one in the SKI (v-ski sarcoma viral oncogene homolog) transcript and one in the Evi1 (ecotropic virus integration site one) gene. One integration of the HaMDR1 vector in the MECT1 (mucoepidermoid carcinoma translocated 1 isoform) gene and one in the transcription regulating gene SIN3A (transcriptional corepressor Sin3A) were found.
Analysis regarding clone size, vector, and integration site did not reveal any correlations: 4 of the 18 clones quantified occurred in RefSeq genes. These genes were the STK10 (serine/threonine kinase 10) gene; the CYP7B1 (cytochrome P450 family seven subfamily B) gene; the LOC201895 (hypothetical protein LOC201895) gene, hit by the G1Na vector; and the SIN3A gene, hit by the HaMDR1 vector. Although the clones carrying the insertion sites in the STK10 and the CYP7B1 genes were too small to be detected, the contribution of the LOC201895 clone and the SIN3A clone to rhesus macaque hematopoiesis was 0.0041% ± 0.00054% and 0.0022% ± 0.00067%, respectively (Table 2). The contribution of the SIN3A clone was comparable to the contribution of the clones containing integrations other than in genes or oncogenes (Table 2).
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
Number of Clones Descending from G1Na and HaMDR1-Transduced Cells
We found a total of 122 different contributing clones in granulocytes and mononuclear cells from 8 weeks up to 4 years after reinfusion of gene-modified CD34+ cells. One hundred two clones descend from the G1Na vector and only 20 from the HaMDR1 vector (Table 1). Using EFRAM-PCR, we were able to detect integration sites using very small amounts of DNA (100 ng). In contrast to LM-PCR, in which fragments are biotin-linked in a one-step amplification reaction, or to linear amplification-mediated PCR (LAM-PCR) [38], which uses linear biotin marking of DNA fragments, the fragments are biotin-marked and amplified simultaneously in EFRAM-PCR. Using this new method, we could reduce the amount of starting DNA from 2,500 ng needed for LM-PCR to 100 ng in EFRAM-PCR. Thus, EFRAM-PCR is a highly sensitive method applicable for samples containing only few transduced cells and in situations in which only small amounts of DNA are available.
Since the efficiency of enrichment of fragments using paramagnetic beads depends on the fragment size (as provided by the manufacturer), we used primers that aligned to both vectors and led to amplification of LTR sequences of equal length, so that the chance for each vector to be detected by LM-PCR was equal. Thus, a methodical advantage for the G1Na vector during detection of the two vectors as the underlying reason for the higher G1Na clone count can be excluded.
Another reason for the difference in the number of identified clones could be a difference in the number of transduced cells. The ratio of HaMDR1-transduced to G1Na-transduced cells was 1/1.18 (40/47) for monkey M038 and 1/1.81 (37/67) for monkey M120 [24]. One would expect similar ratios for the number of clones, assuming similar engraftment levels for both vectors. However, the ratio of identified HaMDR1 clones to G1Na clones was much higher, 1/3.6 in monkey M038 and 1/7.4 in monkey M120; thus, four times as many G1Na clones were found as expected. The discrepancy between the ratio of transduced cells and the ratio of identified clones could be due to a smaller overall contribution of HaMDR1-transduced cells to hematopoiesis. To investigate this, we performed QRT-PCR to measure the contribution of all transduced cells and single transduced cells to hematopoiesis.
Contribution of the G1Na and HaMDR1 Vectors to Hematopoiesis
In both monkeys, the percentage of G1Na-positive DNA was initially high, between 4% and 7%, 8 weeks after transplantation and declined to contribution levels of approximately 1%, whereas the percentage of HaMDR1-positive DNA remained constantly less than 1% over the whole observation period (Fig. 3A, 3B).
Since transduction efficiency was similar for both vectors in both monkeys [24], it remains unclear why the percentage of G1Na-positive DNA is constantly higher than the percentage of HaMDR1-positive DNA. One reason could be technical issues during the transduction process, such as toxicity of supernatants or reduced gene transfer rate because of high numbers of virus-like particles. Biological reasons for the reduced marking rates could be both a selective disadvantage of HaMDR1-transduced cells upon engraftment and reduced proliferation or release into the peripheral blood, depending on MDR1 expression. However, if MDR1 expression in rhesus hematopoiesis confers a selective disadvantage, the clones persisting long-term may have been selected to escape MDR1 overexpression. In addition to epigenetic silencing, ectopic MDR1 expression may also be abrogated by instability of the vector genome [39, 40]. Proof of MDR1 expression in vivo was obtained for monkey M038 at week 27 and for monkey M120 at week 20 [24]. Since only genomic DNA was available to us, we could not perform expression analysis at later time points. Therefore, the conclusions drawn in this study point out that the transduction itself did not lead to clonal dominance. A proof of MDR1 expression would be essential in the investigation of proliferative advantage after application of chemotherapy.
The decrease of G1Na marking rates over 4 years can be interpreted either as a negative selective pressure on G1Na vector-transduced cells or as a positive selective pressure on HaMDR1-transduced cells. Although the MDR1 vector increases, relative to the control vector, potentially showing a selective advantage of HaMDR1-transduced cells, a selective negative pressure on G1Na vector-transduced cells is rather likely, since earlier studies showed that positive selective pressure on MDR1-transduced cells results in an increase of marking rates [4] because of the fact that contribution levels of HaMDR1-transduced cells remain constant over 4 years. This suggestion is corroborated by a recent study in which integration of the NeoR gene-carrying vector did not remain silent but also altered the gene expression pattern of transduced cells [41], as shown for other vectors containing control transgenes such as green fluorescent protein (GFP) [42]. An approach for future trials to overcome this experimental source of error could be the use of a control vector containing the inverted transgene.
Thus, in both monkeys, there were no indications for a misregulated proliferation of HaMDR1-positive cells after 4 years, which would be apparent in an increase of the percentages. In contrast, the stable contribution of HaMDR1-positive HSCs to hematopoiesis implies normal cell behavior.
The higher contribution of G1Na-transduced cells explains the higher number of identified G1Na clones. To further investigate the contribution of these clones on a quantitative level, we performed QRT-PCR on 18 clones (16 G1Na and 2 HaMDR1 clones) identified in both monkeys 4 years after transplantation.
Contribution of Individual Clones to Hematopoiesis
Compared with earlier studies with a time frame of only 6–8 weeks [29] or the analysis of exclusively one integration site [37], this is the first time that the contribution of multiple distinct single transduced repopulating cells is quantified after a long-term follow-up. To rule out repression of the PCR efficiency due to nonspecific retroviral common primer annealing, we chose to analyze clone 1, one of the smallest clones, with an individual primer and probe set. Assuming that there is an efficiency problem due to nonspecific primer binding, a small clone should be measured with a bigger clone size if analyzed with a clone-specific primer and probe set. However, multiple QRT-PCR analyses using this specific set, including the use of different clone-specific reverse primers, could not detect this clone. Thus, the clone-specific set was less sensitive than the vector-specific set with a clone-specific reverse primer, ruling out a limitation factor due to nonspecific primer annealing in the measurement of clone sizes in this setting.
Ten of the clones analyzed were not detectable and thus were under the QRT-PCR detection limit of 0.0001% [29], pointing to extremely small copy numbers. Six G1Na clones could be quantified, showing a clonal size of 0.09% ± 0.03% to 0.42% ± 0.03% of G1Na-transduced PBCs; the two HaMDR1 clones showed a clonal size of 0.71% ± 0.23% and 0.84% ± 0.2% of HaMDR1-transduced PBCs (Table 2). The eight clones contributed to total primate hematopoiesis at levels of 0.0008% ± 0.0003% to 0.0041% ± 0.00032% (Table 2; Fig. 4), also showing that our assay had a high sensitivity for the detection of individual clone sizes. We found no dominant clone or clonal expansion of either G1Na- or HaMDR1-transduced cell clones. However, these data have to be regarded with caution, as only a small number of HaMDR1 clones could be identified and quantified, despite a maximum effort to retrieve from fresh and archival material available from these monkeys. Thus, these data are rather an indication of normal behavior of HaMDR1 clones, given the stable overall contribution of HaMDR1-transduced cells. Furthermore, we cannot rule out that some clones may have escaped detection for technical reasons (e.g., size or secondary structure of the PCR product). Additional biostatistical analyses of clonal dynamics were not performed, as this would require time course data of clonal contribution for at least 3–5 quantified clones per vector type. Also, a statistical confirmation of a difference of, for example, 10% in the frequency of dominant clones between these groups on the basis of Fisher's exact test with a given significance level of 5% and a power of 80% requires at least 170 quantified clones for each vector [32]. These numbers cannot be achieved with the DNA amounts available to us.
To assess and predict clonal behavior and potential clonal dominance (e.g., by the use of quantitative modeling approaches [43, 44]), it is of high interest to quantify the contribution of distinct clones over a prolonged period of time. Currently, unfortunately, a large amount of DNA for quantitation of a clone is needed. This large amount of DNA is often not available. During the experimental phase of this study, there was no possibility to quantify clones at other time points than 4 years after transplantation because of limited DNA amounts available from earlier time points.
Integration in Oncogenes Without Incidence of Leukemia
The G1Na vector showed preferential integrations in RefSeq genes, near transcription start sites of genes, SINEs, and LINEs. Furthermore, a significantly elevated RISC score [26] (RISC = 9, p < .05) for chromosome 19 was observed. These findings are in line with patterns observed in other MLV vector-based studies [45, 46]. Of the HaMDR1 integrations, 73% were found in RefSeq genes. Yet, because of the small number of mapped HaMDR1 integration sites, this finding cannot be judged as a statistically significant preference.
We found an integration of the G1Na vector in the Evi1 gene in monkey M038 58 weeks after transplantation, without the evidence of an abnormal increase of G1Na-positive DNA in the QRT-PCR analysis (Fig. 3B) [37], whereas earlier integration studies observed development of leukemia in mice after vector integration in this proto-oncogene [47].
We did not observe any correlation among clone size, vector, and integration site. The contribution of the SIN3A clone was comparable to the contribution of the clones containing integrations other than in genes or oncogenes. In addition, the overall contribution of HaMDR1-transduced cells to hematopoiesis remained less than 1%. Thus, to date there is no indication that HaMDR1 integrations in the SIN3A and MECT1 oncogenes caused uncontrolled proliferative behavior. Taken together, our data support the assumption that an integration in a single oncogene cannot be solely responsible for causing malignant transformation of a cell [37, 48, 49].
|
SUMMARY |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
The myeloproliferative disorder of HaMDR1-transduced cells observed in earlier studies in mice [21, 22] was most likely due to multiple integrations of the vector in one cell, as shown in a recent study [49]. Clinically effective transduction and transplantation protocols [18], as well as new vectors [50, 51], are now required to test the efficacy of MDR1 gene therapy and in vivo chemoselection in a clinical setting.
|
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionSummaryDisclosure of Potential...AcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Ahmed, N. Arseni, H. Glimm, W. Hiddemann, C. Buske, and M. Feuring-Buske Constitutive Expression of the ATP-Binding Cassette Transporter ABCG2 Enhances the Growth Potential of Early Human Hematopoietic Progenitors Stem Cells, March 1, 2008; 26(3): 810 - 818. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
) was performed. An adapter oligo-cassette was ligated to flanking DNA. Fragments containing LTR-genomic DNA junctions were marked with a biotinylated LTR-specific primer (HaRV292-bio) and enriched by streptavidin-coated paramagnetic beads. Other DNA fragments were flushed away. Nested PCR was performed (A1, adapter-specific primer; HaRV292/HaRV111, LTR-specific primers). (B): Agarose gel electrophoresis of PCR products obtained with EFRAM-PCR. EFRAM-PCR was performed on serial dilutions of the cell line HT1080 containing 0.1%–100% retrovirally transduced DNA in a background of untransduced DNA. The arrow indicates the internal band that represents the PCR product amplified from the proviral 3'LTR. With increasing amounts of vector-containing DNA, we observed increasing signal intensity of the internal band expected at 300 base pairs. To confirm that the band was specific, we cloned and sequenced the obtained band of the 1% dilution sample. Mock controls were negative. Abbreviations: EFRAM-PCR, exponential fragment-marking polymerase chain reaction; kbp, kilobase pairs; LTR, long terminal repeat.
, Grans; 
, MNCs). #, clones detected with ligation-mediated polymerase chain reaction (PCR). All other clones were detected with exponential fragment-marking PCR. Abbreviations: Grans, granulocytes; MNC, mononuclear cell; wk, week; yr, year.
