Sequence-specific correction of genomic hypoxanthine-guanine phosphoribosyl transferase mutations in lymphoblasts by small fragment homologous replacement

Abstract

Oligo/polynucleotide-based gene targeting strategies provide new options for achieving sequence-specific modification of genomic DNA and have implications for the development of new therapies and transgenic animal models. One such gene modification strategy, small fragment homologous replacement (SFHR), was evaluated qualitatively and quantitatively in human lymphoblasts that contain a single base substitution in the hypoxanthine-guanine phosphoribosyl transferase (HPRT1) gene. Because HPRT1 mutant cells are readily discernable from those expressing the wild type (wt) gene through growth in selective media, it was possible to identify and isolate cells that have been corrected by SFHR. Transfection of HPRT1 mutant cells with polynucleotide small DNA fragments (SDFs) comprising wild type HPRT1 (wtHPRT1) sequences resulted in clones of cells that grew in hypoxanthine-aminopterin-thymidine (HAT) medium. Initial studies quantifying the efficiency of correction in 3 separate experiments indicate frequencies ranging from 0.1% to 2%. Sequence analysis of DNA and RNA showed correction of the HPRT1 mutation. Random integration was not indicated after transfection of the mutant cells with an SDF comprised of green fluorescent protein (GFP) sequences that are not found in human genomic DNA. Random integration was also not detected following Southern blot hybridization analysis of an individual corrected cell clone.

FIG. 1.

(A) Schematic of the target region encompassing HPRT1 Exon 3 and the PCR strategy for generating the 579-bp targeting SDF. The strategy for the assessment of genomic DNA targeting as well as the RFLP analysis of the 1025-bp diagnostic product is also depicted. The sequence represents Exon 3 and the LT1-1B1 cell line 152G>C (R51P) mutation. The mutation is indicated as (“g” = 152 G > C). (B) Schematic representation of the RT-PCR and RFLP analysis of HPRT1 mRNA. The XhoI digestion for the 256-bp amplicon will result in a 190- and 66-bp fragment.

FIG. 2.

PCR/RFLP analysis of genomic DNA from transfected cells. The 1025-bp amplification product was either digested with XhoI (lanes 3, 5, and 7) or not digested (lanes 2, 4, and 6). LT1-1B1 cell DNA was isolated before transfection (lanes 2 and 3) and after transfection (lanes 4 and 5). Lanes 6 and 7 were derived from control TK6 cell DNA. Lane 1 is the 100-bp molecular weight marker.

FIG. 3.

(A) RT-PCR/RFLP analysis of mRNA-derived HPRT1 cDNA from LT1-1B1 cells transfected with ssSDFs (lanes 2 and 3), LT1-1B1 cells transfected with dsSDFs (lanes 4 and 5), mutant LT1-1B1 controls (lanes 6 and 7), and HPRT wt TK6 cells (lanes 8 and 9). The 256-bp product from LT1-1B1 cells transfected with ssSDFs (lanes 2 and 3) and from LT1-1B1 cells transfected with dsSDFs (lanes 4 and 5) were digested with XhoI (lanes 2 and 4). Assessment of mRNA from untransfected mutant LT1-1B1 controls (lanes 6 and 7) and HPRT wt TK6 cells (lanes 8 and 9) showed no XhoI digestion of the LT1-1B1 cell cDNA (lane 6) and the expected XhoI cleavage of the TK6 cDNA (lane 8). The RT-PCR products in lanes 3, 5, 7, and 9 are not digested by XhoI and serve as controls. (B) Sequence analysis of the mRNA-derived cDNA.

FIG. 4.

Southern blot hybridization RFLP analysis of genomic HPRT1 digested with TaqI. The blot was hybridized with a 309-bp ³²P-labeled probe. DNAs were as follows: lane 1, untransfected LT1-1B1; lane 2, ssSDF transfected LT1-1B1; lane 3, TK6 cells (normal lymphoblast control); lane 4, K562 (human erythroleukemia cell line, normal control); lane 5, HFL (human fetal liver normal control); and lane 6, lambda HindIII DNA marker. DNA from untransfected LT1-1B1 cells gave a 4173-bp band, while all the DNA derived from HPRT1wt cells and the SFHR corrected cells showed the product bands 2089- and 2084-bp resulting from TaqI digestion at the site of the corrected mutation.

FIG. 5.

Nested PCR analysis of genomic DNA with and without gel purification following transfection with the GFP-SDF. (A) Lanes 1, 3, and 5 show the product of the GFP3/GFP4 amplification using gel purified genomic DNA as a template 1, 2, and 4 weeks post-transfection, respectively. Lack of amplification is in support of the absence of random integration. The quality of genomic DNA was ensured in an independent amplification using primers coding for a known genomic sequence. Lanes 2, 4, and 6 ascertain the gel purified genomic DNA quality with HPRT1 primers (5′-CACAGTTCACTCCAGCCTCA-3′, sense, 5′-CCAGCAGGTCAGCAAAGAAT-3′, antisense) at weeks 1, 2, and 4 post-transfection, respectively. (B) The GFP3/GFP4 PCR analysis of genomic DNA that was not gel purified indicates that the GFP-SDF disappears from the cells as a function of time. Lanes 1–4 indicates the presence of GFP sequences on weeks 1–4 post-transfection, respectively. At 4 weeks post-transfection GFP-SDF could barely be detected.