Limits...
A method to sequence and quantify DNA integration for monitoring outcome in gene therapy.

Brady T, Roth SL, Malani N, Wang GP, Berry CC, Leboulch P, Hacein-Bey-Abina S, Cavazzana-Calvo M, Papapetrou EP, Sadelain M, Savilahti H, Bushman FD - Nucleic Acids Res. (2011)

Bottom Line: Human genetic diseases have been successfully corrected by integration of functional copies of the defective genes into human cells, but in some cases integration of therapeutic vectors has activated proto-oncogenes and contributed to leukemia.Here, we show that a new method based on phage Mu transposition in vitro allows convenient and consistent recovery of integration site sequences in a form that can be analyzed directly using DNA barcoding and pyrosequencing.The method also allows simple estimation of the relative abundance of gene-modified cells from human gene therapy subjects, which has previously been lacking but is crucial for detecting expansion of cell clones that may be a prelude to adverse events.

View Article: PubMed Central - PubMed

Affiliation: Department of Microbiology, University of Pennsylvania School of Medicine, 3610 Hamilton Walk, Philadelphia, PA 19104-6076, USA.

ABSTRACT
Human genetic diseases have been successfully corrected by integration of functional copies of the defective genes into human cells, but in some cases integration of therapeutic vectors has activated proto-oncogenes and contributed to leukemia. For this reason, extensive efforts have focused on analyzing integration site populations from patient samples, but the most commonly used methods for recovering newly integrated DNA suffer from severe recovery biases. Here, we show that a new method based on phage Mu transposition in vitro allows convenient and consistent recovery of integration site sequences in a form that can be analyzed directly using DNA barcoding and pyrosequencing. The method also allows simple estimation of the relative abundance of gene-modified cells from human gene therapy subjects, which has previously been lacking but is crucial for detecting expansion of cell clones that may be a prelude to adverse events.

Show MeSH

Related in: MedlinePlus

Mu-mediated integration site recovery. (A) Integration frequency near genomic land marks. (B) Integration frequency near sites of histone methylation, acetylation or bound chromosomal proteins. Data sets compared are indicated by the column headings, features analyzed by the rows. Heat maps were constructed using the receiver operating characteristic (ROC) area method to compare the observed distributions to random distributions of integration sites (23,45). ROC areas of 0.5 indicated integration sites are present near the indicated genomic features as often as expected by chance. ROC areas >0.5 indicate positive association and areas <0.5 negative association. Associations are colour coded as indicated by the key at the bottom of each map. In (A), for the Gene Density, Expression Density and GC content measures, several different length genomic intervals were used for comparisons, which are indicated by numbers to the right of the black bar. In (B), ChIP-seq analysis was used to map the genome-wide distributions of sites of histone post-translational methylation or acetylation, or bound DNA binding proteins (41–44), and the results compared with integration site distributions. Statistical analysis shows that most associations where discernable colour can be seen in a heat-map tile achieve statistical significance. (C) Comparisons of the numbers and positions of Mu integration sites that allowed recovery of the vector integration sites at HMGA2, VPS13B and POLA2 generated during gene therapy for β-thalassemia.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3113588&req=5

Figure 4: Mu-mediated integration site recovery. (A) Integration frequency near genomic land marks. (B) Integration frequency near sites of histone methylation, acetylation or bound chromosomal proteins. Data sets compared are indicated by the column headings, features analyzed by the rows. Heat maps were constructed using the receiver operating characteristic (ROC) area method to compare the observed distributions to random distributions of integration sites (23,45). ROC areas of 0.5 indicated integration sites are present near the indicated genomic features as often as expected by chance. ROC areas >0.5 indicate positive association and areas <0.5 negative association. Associations are colour coded as indicated by the key at the bottom of each map. In (A), for the Gene Density, Expression Density and GC content measures, several different length genomic intervals were used for comparisons, which are indicated by numbers to the right of the black bar. In (B), ChIP-seq analysis was used to map the genome-wide distributions of sites of histone post-translational methylation or acetylation, or bound DNA binding proteins (41–44), and the results compared with integration site distributions. Statistical analysis shows that most associations where discernable colour can be seen in a heat-map tile achieve statistical significance. (C) Comparisons of the numbers and positions of Mu integration sites that allowed recovery of the vector integration sites at HMGA2, VPS13B and POLA2 generated during gene therapy for β-thalassemia.

Mentions: The genome-wide distribution patterns of integration target sites for HIV and γ-retroviruses have been studied extensively (13,14,23,39,40), allowing us to assess whether the Mu-mediated method reported similar trends. For studies of integration frequency near genomic landmarks, restriction enzyme-based methods usually provided an adequate overview, because the restriction biases are only weakly related to those landmarks. Figure 4A and B compares the distributions of integration sites isolated using the two methods for HIV and γ-retrovirus-based vectors. Integration sites are compared from infections of tissue culture cells and for one SCID-X1 patient (3,7). Figure 4A summarizes the relationship of integration sites to genomic features, using a heat map format to indicate increased or decreased integration frequency compared with random distributions. For both HIV and γ-retroviral vectors, the genome-wide trends were closely similar for the Mu-mediated or restriction enzyme-mediated methods. Both HIV-based and γ-retrovirus-based vectors favor integration in regions of high gene density and associated genomic landmarks. Gamma-retroviruses favor integration near gene 5′-ends. HIV and γ-retroviruses show a complex pattern of favored and disfavored integration sites near regions of histone methylation, acetylation and bound proteins, as indicated by comparison to data from ChIP-seq experiments (41–44) (Figure 4B), again showing favored integration near marks of active transcription. These patterns matched closely for the Mu-mediated and restriction-enzyme-based methods. The genome-wide patterns for SCID-X1 gene therapy closely paralleled those seen for γ-retroviral vectors, as reported previously (3,6,7). Thus, conclusions on genome-wide distributions of integration sites from the Mu-mediated method parallel those from extensive previous studies.Figure 4.


A method to sequence and quantify DNA integration for monitoring outcome in gene therapy.

Brady T, Roth SL, Malani N, Wang GP, Berry CC, Leboulch P, Hacein-Bey-Abina S, Cavazzana-Calvo M, Papapetrou EP, Sadelain M, Savilahti H, Bushman FD - Nucleic Acids Res. (2011)

Mu-mediated integration site recovery. (A) Integration frequency near genomic land marks. (B) Integration frequency near sites of histone methylation, acetylation or bound chromosomal proteins. Data sets compared are indicated by the column headings, features analyzed by the rows. Heat maps were constructed using the receiver operating characteristic (ROC) area method to compare the observed distributions to random distributions of integration sites (23,45). ROC areas of 0.5 indicated integration sites are present near the indicated genomic features as often as expected by chance. ROC areas >0.5 indicate positive association and areas <0.5 negative association. Associations are colour coded as indicated by the key at the bottom of each map. In (A), for the Gene Density, Expression Density and GC content measures, several different length genomic intervals were used for comparisons, which are indicated by numbers to the right of the black bar. In (B), ChIP-seq analysis was used to map the genome-wide distributions of sites of histone post-translational methylation or acetylation, or bound DNA binding proteins (41–44), and the results compared with integration site distributions. Statistical analysis shows that most associations where discernable colour can be seen in a heat-map tile achieve statistical significance. (C) Comparisons of the numbers and positions of Mu integration sites that allowed recovery of the vector integration sites at HMGA2, VPS13B and POLA2 generated during gene therapy for β-thalassemia.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3113588&req=5

Figure 4: Mu-mediated integration site recovery. (A) Integration frequency near genomic land marks. (B) Integration frequency near sites of histone methylation, acetylation or bound chromosomal proteins. Data sets compared are indicated by the column headings, features analyzed by the rows. Heat maps were constructed using the receiver operating characteristic (ROC) area method to compare the observed distributions to random distributions of integration sites (23,45). ROC areas of 0.5 indicated integration sites are present near the indicated genomic features as often as expected by chance. ROC areas >0.5 indicate positive association and areas <0.5 negative association. Associations are colour coded as indicated by the key at the bottom of each map. In (A), for the Gene Density, Expression Density and GC content measures, several different length genomic intervals were used for comparisons, which are indicated by numbers to the right of the black bar. In (B), ChIP-seq analysis was used to map the genome-wide distributions of sites of histone post-translational methylation or acetylation, or bound DNA binding proteins (41–44), and the results compared with integration site distributions. Statistical analysis shows that most associations where discernable colour can be seen in a heat-map tile achieve statistical significance. (C) Comparisons of the numbers and positions of Mu integration sites that allowed recovery of the vector integration sites at HMGA2, VPS13B and POLA2 generated during gene therapy for β-thalassemia.
Mentions: The genome-wide distribution patterns of integration target sites for HIV and γ-retroviruses have been studied extensively (13,14,23,39,40), allowing us to assess whether the Mu-mediated method reported similar trends. For studies of integration frequency near genomic landmarks, restriction enzyme-based methods usually provided an adequate overview, because the restriction biases are only weakly related to those landmarks. Figure 4A and B compares the distributions of integration sites isolated using the two methods for HIV and γ-retrovirus-based vectors. Integration sites are compared from infections of tissue culture cells and for one SCID-X1 patient (3,7). Figure 4A summarizes the relationship of integration sites to genomic features, using a heat map format to indicate increased or decreased integration frequency compared with random distributions. For both HIV and γ-retroviral vectors, the genome-wide trends were closely similar for the Mu-mediated or restriction enzyme-mediated methods. Both HIV-based and γ-retrovirus-based vectors favor integration in regions of high gene density and associated genomic landmarks. Gamma-retroviruses favor integration near gene 5′-ends. HIV and γ-retroviruses show a complex pattern of favored and disfavored integration sites near regions of histone methylation, acetylation and bound proteins, as indicated by comparison to data from ChIP-seq experiments (41–44) (Figure 4B), again showing favored integration near marks of active transcription. These patterns matched closely for the Mu-mediated and restriction-enzyme-based methods. The genome-wide patterns for SCID-X1 gene therapy closely paralleled those seen for γ-retroviral vectors, as reported previously (3,6,7). Thus, conclusions on genome-wide distributions of integration sites from the Mu-mediated method parallel those from extensive previous studies.Figure 4.

Bottom Line: Human genetic diseases have been successfully corrected by integration of functional copies of the defective genes into human cells, but in some cases integration of therapeutic vectors has activated proto-oncogenes and contributed to leukemia.Here, we show that a new method based on phage Mu transposition in vitro allows convenient and consistent recovery of integration site sequences in a form that can be analyzed directly using DNA barcoding and pyrosequencing.The method also allows simple estimation of the relative abundance of gene-modified cells from human gene therapy subjects, which has previously been lacking but is crucial for detecting expansion of cell clones that may be a prelude to adverse events.

View Article: PubMed Central - PubMed

Affiliation: Department of Microbiology, University of Pennsylvania School of Medicine, 3610 Hamilton Walk, Philadelphia, PA 19104-6076, USA.

ABSTRACT
Human genetic diseases have been successfully corrected by integration of functional copies of the defective genes into human cells, but in some cases integration of therapeutic vectors has activated proto-oncogenes and contributed to leukemia. For this reason, extensive efforts have focused on analyzing integration site populations from patient samples, but the most commonly used methods for recovering newly integrated DNA suffer from severe recovery biases. Here, we show that a new method based on phage Mu transposition in vitro allows convenient and consistent recovery of integration site sequences in a form that can be analyzed directly using DNA barcoding and pyrosequencing. The method also allows simple estimation of the relative abundance of gene-modified cells from human gene therapy subjects, which has previously been lacking but is crucial for detecting expansion of cell clones that may be a prelude to adverse events.

Show MeSH
Related in: MedlinePlus