Figure 1.
Gene targeting is inefficient in human somatic cells.
When targeting vector is transfected into human cells, random integration by non-homologous recombination occurs at least 2 to 3 orders of magnitude more frequently than homologous recombination-mediated targeted integration. The LIG4-dependent NHEJ pathway has been thought to be responsible for random integration, but recent evidence indicates a contribution from LIG4-independent mechanisms that rely on LIG1/3 (DNA ligase I or IIIα). The gene-targeting efficiency is calculated by dividing the number of targeted clones with that of drug-resistant clones analyzed (see Materials and Methods for details).
Figure 2.
HPRT targeting vectors with long, but not short, homology arms stimulate NHEJ-independent random integration.
(A) Schematic representation of HPRT targeting vectors pHPRT8.9-Puro(+) and pHPRT8.9-Puro(−). (B) Schematic representation of HPRT targeting vectors pHPRT2.2-Puro(+) and pHPRT2.2-Puro(−). (C) Structural features of the HPRT targeting vectors. (D) Integration frequency of HPRT targeting vectors and pPGK-Puro (a non-targeting vector) in human Nalm-6 wild-type and LIG4-null cells. The ratio of integration frequency in LIG4-null to wild-type cells is indicated in the right column. At least six independent experiments were performed for each vector. Note that pPGK-Puro harbors little or no homology to the human genome. Grey lines indicate the lengths of plasmid backbones, and § denotes a 14-bp sequence.
Figure 3.
The short-arm vector pHPRT2.2-Puro(−) functions as a genuine targeting vector.
(A) Gene-targeting efficiency of pHPRT8.9-Puro(−) and pHPRT2.2-Puro(−) in wild-type and LIG4-null cells. (B) Random and targeted integration frequencies of pHPRT8.9-Puro(−) and pHPRT2.2-Puro(−) in wild-type and LIG4-null cells. At least three independent experiments were performed for each vector.
Figure 4.
NHEJ-independent random integration is significantly decreased when a homology arm is deleted from the HPRT targeting vector.
(A) Integration frequency of pHPRT8.9-Puro vectors and their derivatives in Nalm-6 wild-type and LIG4-null cells. The ratio of integration frequency in LIG4-null to wild-type cells is indicated in the right column. At least five independent experiments were performed for each vector. The data for the arm-proficient vectors are the same as that in Figure 2D. (B) Integration frequency of pHPRT2.2-Puro vectors and their derivatives in Nalm-6 wild-type and LIG4-null cells. The ratio of integration frequency in LIG4-null to wild-type cells is indicated in the right column. At least three independent experiments were performed for each vector. Symbols are as in Figure 2D, and the data for the arm-proficient vectors are the same as that in Figure 2D.
Figure 5.
Integration frequency of targeting vector correlates with the lengths of homology arms and repetitive DNA sequences.
Integration frequencies of pHPRT8.9-Puro(+), pHPRT2.2-Puro(+), and twelve other (non-HPRT) gene-targeting vectors are shown as a function of the total length of homology arms (A), 5’-arm length (B), 3’-arm length (C), the total length of SINE/LINE sequences present in the arms (D), 5’-arm SINE/LINE length (E), and 3’-arm SINE/LINE length (F). The R2 values shown in the graphs were calculated from the fourteen points. See also Figures S4 and S5 in File S1 for details.
Figure 6.
Integration frequency of targeting vector correlates well with the lengths of homology arms and repetitive DNA sequences, particularly in the absence of LIG4.
(A, B) Integration frequencies of seven gene-targeting vectors in wild-type and LIG4-null cells are shown as a function of the total length of homology arms (A) and the total length of SINE/LINE sequences present in the arms (B). (C, D) Estimated frequencies of LIG4-dependent integration are shown as a function of the total length of homology arms (C) and the total length of SINE/LINE sequences present in the arms (D). The values were calculated by subtracting the integration frequency in LIG4-null cells from that in wild-type cells. See text for more details.