Limits...
Stable gene targeting in human cells using single-strand oligonucleotides with modified bases.

Rios X, Briggs AW, Christodoulou D, Gorham JM, Seidman JG, Church GM - PLoS ONE (2012)

Bottom Line: Stably EGFP-corrected cells were generated at a frequency of ~0.05% with an optimized oligonucleotide design combining modified bases and reduced number of phosphorothioate bonds.We provide evidence from comparative RNA-seq analysis suggesting cellular immunity induced by the oligonucleotides might contribute to the low viability of oligo-corrected cells.Further optimization of this method should allow rapid and scalable genome engineering in human cells.

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, Harvard Medical School, Boston, Massachusetts, United States of America.

ABSTRACT
Recent advances allow multiplexed genome engineering in E. coli, employing easily designed oligonucleotides to edit multiple loci simultaneously. A similar technology in human cells would greatly expedite functional genomics, both by enhancing our ability to test how individual variants such as single nucleotide polymorphisms (SNPs) are related to specific phenotypes, and potentially allowing simultaneous mutation of multiple loci. However, oligo-mediated targeting of human cells is currently limited by low targeting efficiencies and low survival of modified cells. Using a HeLa-based EGFP-rescue reporter system we show that use of modified base analogs can increase targeting efficiency, in part by avoiding the mismatch repair machinery. We investigate the effects of oligonucleotide toxicity and find a strong correlation between the number of phosphorothioate bonds and toxicity. Stably EGFP-corrected cells were generated at a frequency of ~0.05% with an optimized oligonucleotide design combining modified bases and reduced number of phosphorothioate bonds. We provide evidence from comparative RNA-seq analysis suggesting cellular immunity induced by the oligonucleotides might contribute to the low viability of oligo-corrected cells. Further optimization of this method should allow rapid and scalable genome engineering in human cells.

Show MeSH

Related in: MedlinePlus

Oligo-mediated targeting reporter system.(a) Reporter consists of a HeLa cell line with two stably integrated copies of EGFP with a mutated TTG start codon and a second potential start codon (AAG) downstream, both shown in red. Each mutated start codon can be targeted by sense or antisense oligos. Representative oligos used in this study are highlighted, in duplex form, with the mismatch shown by an asterisk. Further detail can be found on Table 1. (b) Sample flow cytometry dot plot of cells with no oligo (left) or transfected with F5-8 (right, Table 1). The frequency of oligo-induced correction of a start codon can be estimated as the %EGFP+/Propidium Iodide- cells (c) %EGFP+ cells generated after different oligo transfections, with and without 20 mM thymidine, assayed 36 h post transfections. Oligo sequences are shown 5′ to 3′, and the PTO bonds are highlighted in gray. The control oligos F5-1 and F5-6 did not produce any significant proportion of EGFP+ cells, neither did F5-2, which is complementary to the transcribed strand and encodes an ATG-restoring mutation. The oligos targeting the first and second potential start codons on the non-transcribed strand, F5-3 and F5-5 respectively, did produce EGFP+ cells. Lipo = lipofectamine only control. n = 4.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3351460&req=5

pone-0036697-g001: Oligo-mediated targeting reporter system.(a) Reporter consists of a HeLa cell line with two stably integrated copies of EGFP with a mutated TTG start codon and a second potential start codon (AAG) downstream, both shown in red. Each mutated start codon can be targeted by sense or antisense oligos. Representative oligos used in this study are highlighted, in duplex form, with the mismatch shown by an asterisk. Further detail can be found on Table 1. (b) Sample flow cytometry dot plot of cells with no oligo (left) or transfected with F5-8 (right, Table 1). The frequency of oligo-induced correction of a start codon can be estimated as the %EGFP+/Propidium Iodide- cells (c) %EGFP+ cells generated after different oligo transfections, with and without 20 mM thymidine, assayed 36 h post transfections. Oligo sequences are shown 5′ to 3′, and the PTO bonds are highlighted in gray. The control oligos F5-1 and F5-6 did not produce any significant proportion of EGFP+ cells, neither did F5-2, which is complementary to the transcribed strand and encodes an ATG-restoring mutation. The oligos targeting the first and second potential start codons on the non-transcribed strand, F5-3 and F5-5 respectively, did produce EGFP+ cells. Lipo = lipofectamine only control. n = 4.

Mentions: To improve human genome engineering with oligonucleotides, we worked with a well-characterized reporter system [19] consisting of a HeLa cell line with two stably integrated copies of a modified EGFP gene (mEGFP). This version of EGFP has a non-functional start codon (TTG) which can be rescued by targeting oligos (Figure 1a, Table 1) encoding for a functional ATG, thus the oligo-mediated targeting efficiency can be determined by flow cytometry as the percentage EGFP+ cells (Fig. 1b). We confirmed previous findings [27] where targeting mEGFP with a 25 bp long oligo complementary to the non-transcribed strand and carrying a centrally located mismatch and six PTO bonds at each end (F5-3, Fig. 1c) delivered with Lipofectamine 2000 yielded a substantial proportion of EGFP+ cells after 48 hours (~0.5%), and this efficiency was further increased to ~2% by slowing down DNA replication with thymidine treatment. Other cationic lipid transfection reagents resulted in lower efficiencies, possibly due to low nuclear accumulation of the oligo (Figures S1,S2). Alternative delivery methods such as electroporation and nucleofection failed to produce any significant proportion of EGFP+ cells (not shown). An oligo complementary to the transcribed strand, F5-2, did not produce any EGFP+ cells significantly different from background. Transfection of an oligo encoding an alternative ATG-restoring mutation 9 bp away from the first site resulted in ~0.4% EGFP+ cells (F5-5, Fig. 1a,c), whereas an oligo carrying a non-coding mismatch (F5-6, Fig. 1c) did not produce EGFP+ cells, demonstrating that the expression of EGFP depends on the targeting oligo sequence restoring the ORF. To further verify that EFGP+ cells had undergone the desired genomic modification, EGFP+ and – cells were sorted out by fluorescent-activated cell sorting (FACS) and genotyped by allele-specific qPCR (AS-qPCR). The EGFP+ population was estimated by AS-qPCR to carry 11–13% converted DNA, which matches the 12.5% expected if these cells have undergone a single oligo incorporation at one of the two genomic mEGFP during DNA replication (1/8 strands at the end of S phase) but not yet proceeded through cell division. In the EGFP- population, the proportion of corrected alleles was estimated to be ~1%, which may either represent corrected cells that had not yet produced functional EGFP protein or PCR artifacts caused by residual targeting oligonucleotides in the cells [28].


Stable gene targeting in human cells using single-strand oligonucleotides with modified bases.

Rios X, Briggs AW, Christodoulou D, Gorham JM, Seidman JG, Church GM - PLoS ONE (2012)

Oligo-mediated targeting reporter system.(a) Reporter consists of a HeLa cell line with two stably integrated copies of EGFP with a mutated TTG start codon and a second potential start codon (AAG) downstream, both shown in red. Each mutated start codon can be targeted by sense or antisense oligos. Representative oligos used in this study are highlighted, in duplex form, with the mismatch shown by an asterisk. Further detail can be found on Table 1. (b) Sample flow cytometry dot plot of cells with no oligo (left) or transfected with F5-8 (right, Table 1). The frequency of oligo-induced correction of a start codon can be estimated as the %EGFP+/Propidium Iodide- cells (c) %EGFP+ cells generated after different oligo transfections, with and without 20 mM thymidine, assayed 36 h post transfections. Oligo sequences are shown 5′ to 3′, and the PTO bonds are highlighted in gray. The control oligos F5-1 and F5-6 did not produce any significant proportion of EGFP+ cells, neither did F5-2, which is complementary to the transcribed strand and encodes an ATG-restoring mutation. The oligos targeting the first and second potential start codons on the non-transcribed strand, F5-3 and F5-5 respectively, did produce EGFP+ cells. Lipo = lipofectamine only control. n = 4.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3351460&req=5

pone-0036697-g001: Oligo-mediated targeting reporter system.(a) Reporter consists of a HeLa cell line with two stably integrated copies of EGFP with a mutated TTG start codon and a second potential start codon (AAG) downstream, both shown in red. Each mutated start codon can be targeted by sense or antisense oligos. Representative oligos used in this study are highlighted, in duplex form, with the mismatch shown by an asterisk. Further detail can be found on Table 1. (b) Sample flow cytometry dot plot of cells with no oligo (left) or transfected with F5-8 (right, Table 1). The frequency of oligo-induced correction of a start codon can be estimated as the %EGFP+/Propidium Iodide- cells (c) %EGFP+ cells generated after different oligo transfections, with and without 20 mM thymidine, assayed 36 h post transfections. Oligo sequences are shown 5′ to 3′, and the PTO bonds are highlighted in gray. The control oligos F5-1 and F5-6 did not produce any significant proportion of EGFP+ cells, neither did F5-2, which is complementary to the transcribed strand and encodes an ATG-restoring mutation. The oligos targeting the first and second potential start codons on the non-transcribed strand, F5-3 and F5-5 respectively, did produce EGFP+ cells. Lipo = lipofectamine only control. n = 4.
Mentions: To improve human genome engineering with oligonucleotides, we worked with a well-characterized reporter system [19] consisting of a HeLa cell line with two stably integrated copies of a modified EGFP gene (mEGFP). This version of EGFP has a non-functional start codon (TTG) which can be rescued by targeting oligos (Figure 1a, Table 1) encoding for a functional ATG, thus the oligo-mediated targeting efficiency can be determined by flow cytometry as the percentage EGFP+ cells (Fig. 1b). We confirmed previous findings [27] where targeting mEGFP with a 25 bp long oligo complementary to the non-transcribed strand and carrying a centrally located mismatch and six PTO bonds at each end (F5-3, Fig. 1c) delivered with Lipofectamine 2000 yielded a substantial proportion of EGFP+ cells after 48 hours (~0.5%), and this efficiency was further increased to ~2% by slowing down DNA replication with thymidine treatment. Other cationic lipid transfection reagents resulted in lower efficiencies, possibly due to low nuclear accumulation of the oligo (Figures S1,S2). Alternative delivery methods such as electroporation and nucleofection failed to produce any significant proportion of EGFP+ cells (not shown). An oligo complementary to the transcribed strand, F5-2, did not produce any EGFP+ cells significantly different from background. Transfection of an oligo encoding an alternative ATG-restoring mutation 9 bp away from the first site resulted in ~0.4% EGFP+ cells (F5-5, Fig. 1a,c), whereas an oligo carrying a non-coding mismatch (F5-6, Fig. 1c) did not produce EGFP+ cells, demonstrating that the expression of EGFP depends on the targeting oligo sequence restoring the ORF. To further verify that EFGP+ cells had undergone the desired genomic modification, EGFP+ and – cells were sorted out by fluorescent-activated cell sorting (FACS) and genotyped by allele-specific qPCR (AS-qPCR). The EGFP+ population was estimated by AS-qPCR to carry 11–13% converted DNA, which matches the 12.5% expected if these cells have undergone a single oligo incorporation at one of the two genomic mEGFP during DNA replication (1/8 strands at the end of S phase) but not yet proceeded through cell division. In the EGFP- population, the proportion of corrected alleles was estimated to be ~1%, which may either represent corrected cells that had not yet produced functional EGFP protein or PCR artifacts caused by residual targeting oligonucleotides in the cells [28].

Bottom Line: Stably EGFP-corrected cells were generated at a frequency of ~0.05% with an optimized oligonucleotide design combining modified bases and reduced number of phosphorothioate bonds.We provide evidence from comparative RNA-seq analysis suggesting cellular immunity induced by the oligonucleotides might contribute to the low viability of oligo-corrected cells.Further optimization of this method should allow rapid and scalable genome engineering in human cells.

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, Harvard Medical School, Boston, Massachusetts, United States of America.

ABSTRACT
Recent advances allow multiplexed genome engineering in E. coli, employing easily designed oligonucleotides to edit multiple loci simultaneously. A similar technology in human cells would greatly expedite functional genomics, both by enhancing our ability to test how individual variants such as single nucleotide polymorphisms (SNPs) are related to specific phenotypes, and potentially allowing simultaneous mutation of multiple loci. However, oligo-mediated targeting of human cells is currently limited by low targeting efficiencies and low survival of modified cells. Using a HeLa-based EGFP-rescue reporter system we show that use of modified base analogs can increase targeting efficiency, in part by avoiding the mismatch repair machinery. We investigate the effects of oligonucleotide toxicity and find a strong correlation between the number of phosphorothioate bonds and toxicity. Stably EGFP-corrected cells were generated at a frequency of ~0.05% with an optimized oligonucleotide design combining modified bases and reduced number of phosphorothioate bonds. We provide evidence from comparative RNA-seq analysis suggesting cellular immunity induced by the oligonucleotides might contribute to the low viability of oligo-corrected cells. Further optimization of this method should allow rapid and scalable genome engineering in human cells.

Show MeSH
Related in: MedlinePlus