Table 2.
Number of predicted copies of integrated vectors in the analyzed samples
Block | DNA, ng | Diploid genomes corresponding to inputa | Tumor cells (%)b | VCNc | Random integrationd |
Clonal integratione | ||
---|---|---|---|---|---|---|---|---|
1.00% | 0.10% | 0.01% | ||||||
A3 | 45 | 6,300 | 50% | 1 | 31.5 | 3.15 | 0.315 | 3,150 |
B1 | 57 | 7,980 | 25% | 1 | 19.95 | 1.995 | 0.1995 | 1,995 |
C1 | 5 | 700 | 75% | 1 | 5.25 | 0.525 | 0.0525 | 525 |
AAV, adeno-associated virus; H&E, hematoxylin and eosin; LOD, limit of detection; VCN, vector copy number. Example for block A3: 45 ng of DNA corresponded to 6,300 cells (because each human cell contains ∼7.2 pg of DNA). Assuming that 50% of this sample contains tumor cells (as suggested by ISH) and has a VCN of 1, and assuming that integration happens randomly in 1% of the tumor cells, 31 different copies of integrated vector could be expected in the sample. This is likely below the LOD, especially if the integration rate was 0.1% or 0.01%. However, if all cells were derived from the same integrated clone, then 3,150 integrated vector copies (i.e., with the same integration site) would be expected. This is likely above the LOD.
Average weight of DNA contained in one human diploid genome is 7.2 pg.
Percentage tumor cells in sample estimated on H&E-stained serial sections.
Estimated VCN vector copy number in tumor cells based on in situ hybridization.
Predicted AAV vector sequence number if clonal integration present in tumor cells.