TALENs-mediated homozygous CCR5Δ32 mutations endow CD4+ U87 cells with resistance against HIV‑1 infection

Introduction

C-C chemokine receptor type 5 (CCR5) is a co-receptor for human immunodeficiency virus (HIV)-1 present on the surface of target cells, including cluster of differentiation (CD)4+ T lymphocytes (1). A small fraction of Europeans (1%) are homozygous for a 32-bp deletion within the coding region of both CCR5 alleles (CCR5Δ32/Δ32 genotype), which produces a polymorphic form of CCR5 that is not present on the cell surface, and thus confers strong protection against HIV-1 infection (2). CCR5 antagonists can block HIV-1 entry into target cells; at present, one small molecule CCR5 antagonist has been approved for clinical use (3).

In 2007, an HIV-1-infected patient with acute myeloid leukemia received transplantation of bone marrow stem cells from a donor with the CCR5Δ32/Δ32 genotype, and the viral load in this patient has since been undetectable (4,5). Therefore, replacement of host CD4+ T lymphocytes with engineered CCR5Δ32/Δ32 genotype cells is believed to represent a method by which HIV-1 infection may be cured. Various gene-targeting techniques could be used to produce genetically engineered cells, including zinc finger nucleases (ZFNs) (6,7), transcription activator-like effector nucleases (TALENs) (8–10) and the RNA-guided CRISPR/Cas9 nuclease system (11,12), which can be used to induce random mutations (deletion and/or insertions) or insert a specific gene at specific loci.

Various techniques have been used to disrupt CCR5 in hematopoietic stem and progenitor cells (HSPCs), CD4+ T lymphocytes and induced pluripotent stem cells (iPSCs) (7,13–17). Disruption of CCR5 by ZFNs can efficiently inhibit HIV-1 infection of CD4+ T cells (7). In addition, ZFN modification of CCR5 in primary human CD4+ T cells protects cells from infection with CCR5- and CXCR4-trophic HIV-1 strains (6). TALENs recognize only one nucleotide, instead of the three required for ZFNs (9), and can target sites in the CCR5 loci with less cytotoxicity than ZFNs (8). This technique has been reported to protect CCR5-expressing T cells from R5-tropic HIV (10). In addition, Wang et al (11) recently silenced CCR5 via Cas9 and CCR5-specific single-guide RNA in CEM cells, whereas Hou et al (12) extended this to CXCR4 in primary CD4+ T cells.

Although bi-allelic disruption of the CCR5 gene can prevent infection of target cells, including CD4+ T lymphocytes, concerns have been raised suggesting that cells with non-functional CCR5 may lose some important immune functions (18); however, individuals with the CCR5Δ32/Δ32 genotype do not experience any discernable deleterious clinical effects (19,20). Recently, Ye et al (21) homozygously reproduced the naturally existing CCR5Δ32 mutation in iPSCs by combining the TALENs or CRISPR/Cas9 technique with the PiggyBac technique, as a ‘TTAA’ tetranucleotide sequence happens to be located close to the to-be-deleted 32 bp region. The established CCR5Δ32/Δ32 iPSC clones maintained pluripotency and resistance to HIV-1 infection, further indicating that the CCR5Δ32/Δ32 genotype is safe for cells.

Site-specific, size-controlled and homozygous DNA deletion remains a major challenge in mammalian genome engineering. The present study established an efficient method to homozygously reproduce the natural CCRΔ32 mutation in CD4+ U87 cells using a TALENs-mediated homologous recombination technique. Engineered CD4+ U87 cells with the CCR5Δ32/Δ32 genotype exhibited significant resistance to HIV-1 infection.

Materials and methods

Cell culture

CD4+ U87 cells were acquired from American Type Culture Collection (Manassas, VA, USA). CD4+ U87 cells were originally derived from glioma cells expressing CCR5 and CXCR4, and were stably transfected with a CD4 receptor gene to mimic CD4+ T lymphocytes, and a puromycin gene resistance for selection. CD4+ U87 cells were cultured in Dulbecco's modified Eagle's medium (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 1% penicillin and streptomycin, 1%, amphotericin B, 1% sodium pyruvate, 1% L-glutamine and 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.) in a CO2 incubator at 37°C.

Construction of CCR5Δ32 donor DNA plasmids

CCR5Δ32 DNA plasmids were constructed by overlap extension PCR. To mimic the naturally occurring CCR5Δ32, mutation, the 32 bp DNA fragment (3,299–3,330 bp) was deleted from the wild-type CCR5 (Gene ID:1234, https://www.ncbi.nlm.nih.gov/gene/1234). Two sets of primers, F1 (5′-CACAAGATTTTATTTGGTGAGA-3′) and R1 (5′-CTATCTTTAATGTATGGAAAATGAGAGCTG-3′), and F2 (5′-TTTCCATACATTAAAGATAGTCATCTTGGG-3′) and R2 (5′-ATACATAAGGAACTTTCGGAGT-3′), were designed for both sides of the 32 bp DNA fragment, as indicated in Fig. 1. The two homologous arms, 836 and 786 bp in lengths, were separately amplified by PCR with the primers F1/R1 and F2/R2, respectively, and were then used as DNA templates for the next round of PCR with the primers F1 and R2. The products (1,602 bp in length) were confirmed to contain the correct CCR5Δ32 sequence by gene sequencing (data not shown), and were finally ligated into EcoRI/BamHI digested T-easy vectors (Promega Corporation, Madison, WI, USA).

Figure 1. Schematic diagram of CCR5Δ32 donor DNA construction. Homologous arms A and B were first amplified separately with F1 and R1, and F2 and R2, respectively. Their products, 836 and 786 bp in length, were then mixed and annealed with each other, followed by the next round of polymerase chain reaction with F1 and R2. The final product (A+B, 1,602 bp in length) was finally ligated into the EcoRI/BamHI digested T-easy vectors. CCR5, C-C chemokine receptor type 5.

Designation and selection of TALENs

CCR5-TALENs were designed with the right arm mostly overlapping the to-be-deleted 32 bp region and the left arm 14–18 bp upstream (Fig. 2A). It is critical to ensure arms are mostly within the specific regions to prevent the donor DNA and the mutated CCR5Δ32 from further targeting by CCR5-TALENs. In addition, the CCR5Δ1-TALENs were designed with both arms overlapping the to-be-deleted 31 bp region (Fig. 2B). The TALENs plasmids were constructed by one-step ligation using the Fast TALE™ TALEN Assembly kit (Sidansai Biotechnology Co., Ltd., Shanghai, China). Altogether, six pairs of TALENs plasmids were designed and constructed to target both CCR5 and CCR5Δ1. After a preliminary transfection, two pairs of TALENs plasmids, L4309/R4324 for CCR5, and L4336/R4352 for CCR5Δ1, with the highest targeting efficiencies were selected for use in the subsequent experiments. The plasmids were designated CCR5-TALENs and CCR5Δ1-TALENs accordingly.

Figure 2. CCR5-TALENs and CCR5Δ1-TALENs. (A) CCR5-TALENs, three left (L4307, L4309 and L4310) and two right arms (R4324 and R4325) were designed. The binding domains of the three left arms were located mostly within the to-be-deleted 32 bp region. The spacer lengths ranged between 14 and 18 bp. (B) CCR5Δ1-TALENs, three left (L4334, L4335 and L4336) and two right arms (R4351 and R4352) were designed. All binding domains of the five arms more or less overlapped the to-be-deleted 31 bp region. The spacer lengths ranged between 15 and 18 bp. CCR5, C-C chemokine receptor type 5.

Transfection

CD4+ U87 cells (~1×106) were mixed with 8 µg paired CCR5-TALENs plasmids (each 4 µg) and 2 µg CCR5Δ32 donor DNA in a cuvette (Bio-Rad Laboratories, Inc., Hercules, CA, USA) with 100 µl Opti-minimal essential medium (Gibco; Thermo Fisher Scientific, Inc.). Electroporation was conducted at 150 V using a transfection system (NEPA21; Nepa Gene Co., Ltd., Chiba, Japan). Transfected CD4+ U87 cells were transferred into one well of a 6-well plate and cultured at 37°C for 3 days, after which the cells underwent a second and third round of transfection using the same conditions. After three rounds of transfection, DNA was extracted using the Blood Genomic DNA Extraction Mini kit (Tiangen Biotech Co., Ltd., Beijing, China) for T7 endonuclease 1 (T7E1) enzyme analysis, and ≥50 transfected cells were cultured individually. A single clone with the CCR5Δ32 mutation was screened by monoculture and sequencing. In order to generate homozygous CCR5Δ32 mutations, transfection of the CD4+ U87 cells with CCR5Δ32/Δ1 was performed under the same conditions using the CCR5Δ1-TALENs plasmids for a further two rounds in place of the CCR5-TALENs plasmids, and no donor DNA was added.

T7E1 enzyme analyses

The genomic region, encompassing the TALENs targeting site, was amplified by polymerase chain reaction (PCR), and PCR products were denatured and annealed. Some heteroduplex DNA was formed as a result of random mutations or homologous recombination. The annealed DNA was then digested with 5 units of T7E1 enzyme (Beijing Viewsolid Biotech Co., Ltd., Beijing, China) at 37°C for 15 min. The heteroduplex DNA fragments were cut at the mismatch point and 1% agarose gel electrophoresis was performed to separate the DNA fragments. The TALENs targeting rate was calculated according to the DNA band intensities measured by grayscale technique (Photoshop CS6; Adobe Systems, Inc., San Jose, CA, USA).

Nested PCR

In order to avoid contamination of undegenerated CCR5Δ32 donor DNA, two pairs of primers were designed to amplify the targeting sites. The first round of PCR was performed with primers F3 (5′-TTCATCATCCTCCTGACAATCG-3′) and R3a (5′-CTCAAGAATCAGCAATTCTC-3′), and product length was 1,048 bp. Since R3a was located 18 bp downstream from the primer R2 used to amplify donor DNA, there was no chance for R3a to anneal to donor DNA when contaminated. The products from the first round of PCR were purified by gel extraction, and the second round of PCR was performed with primers F3 and R3b (5′-TGGTCCAACCTGTTAGAGCTAC-3′) to amplify the targeting sites; product length, 479 bp. The PCR product was ligated into T-easy vectors and then transformed into DH5a competent cells, and successfully transformed clones were gene sequenced using Sanger sequencing.

Determination of the p24 antigen

Cells were challenged with BaL-HIV-1 obtained from State Key Laboratory for Infectious Disease Prevention and Control (Beijing, China), a CCR5-trophic virus isolate, at a multiplicity of infection of 0.06, for 4 h at 37°C with 8 µg/ml Polybrene. The challenged cells were rinsed three times to remove the free virus and were cultured as above for 12 days. The culture supernatants were collected every 48 h and replaced with fresh medium. p24 content in the culture supernatants was assessed in triplicate by ELISA (632200; Clontech Laboratories, Inc., Mountainview, CA, USA) according to the manufacturer's protocol (22,23). This experiment was performed in a P-3 laboratory situated in the Chinese Center for Disease Control and Prevention (Beijing, China) strictly according to the guidelines.

Results

TALENs-mediated homozygous CCR5Δ32 mutation

CCR5-TALENs and CCR5Δ1-TALENs plasmids were used to induce homozygous CCR5Δ32 mutation (Fig. 3). CD4+ U87 cells were initially transfected with CCR5-TALENs plasmids and the CCR5Δ32 donor DNA fragments carried in T-easy vectors. After three rounds of transfection without any antibiotic selection, the CCR5 gene in ≤50% of CD4+ U87 cells was targeted by T7E1 enzyme analysis (Fig. 3A). Two of the 29 (6.9%) single-cell cultured clones were revealed to carry bi-allelic mutations, and one of these contained a 1 bp deletion (Δ1) on one allele and a 32 bp deletion (Δ32) on the other (representing 1.7% of the transfected alleles). The latter was confirmed to carry the natural CCR5Δ32 mutation by gene sequencing (Fig. 3C). The bi-allelic mutated CD4+ U87 cells with the CCR5Δ32/Δ1 genotype underwent a further two rounds of transfection with CCR5Δ1-TALENs, without any donor DNA. It was assumed that the mutated CCR5Δ32 alleles themselves could be used as donor DNA for the predicted homologous recombination.

Figure 3. TALENs-mediated homozygous recombination. (A) T7E1 enzyme analysis of transfected CD4+ U87 cells after each transfection with CCR5-TALENs. The targeting efficiencies after first, second and third rounds of transfection were 14.80, 38.20 and 50.04%, respectively. M, 100 bp marker; wild-type, wild-type cells; lanes 1–3, transfected cells after the first, second and third rounds of transfections. (B) PCR of CCR5Δ32/Δ1 genotype cells. M, 100 bp marker; lane 1, prior to transfection; lane 2, after CCR5-TALENs transfection; lane 3, after CCR5Δ1-TALENs transfection; the CCR5Δ32 band (168 bp) was 2.87-times brighter than the CCR5Δ1 band; lane 4, wild-type cells (negative control); lane 5, CCR5Δ32 donor DNA (positive control). (C) Sequencing of wild-type CD4+ U87 cells and those with CCR5Δ32/Δ1 mutations. (D) PCR of 34 single-cell cultured clones of CD4+ U87 cells with CCR5Δ32/Δ1 mutations post-transfection with CCR5Δ1-TALENs. The CCR5Δ32 band was observed in all clones; however, as a single band, it was only seen in clones 16, 18 and 25 post-transfection with CCR5Δ1-TALENs. M, 100 bp marker; lanes 1–34, single-cell cultured cells post-transfection with CCR5Δ1-TALENs; wild-type, wild-type cells (negative control); donor, CCR5Δ32 donor DNA (positive control). (E) Gene sequencing of clones 16, 18 and 25. CCR5, C-C chemokine receptor type 5; CD4, cluster of differentiation 4; PCR, polymerase chain reaction; T7E1, T7 endonuclease 1; TALENs, transcription activator-like effector nucleases.

As expected, PCR revealed that the genomic DNA of cells with the CCR5Δ32/Δ1 genotype contained similar levels of CCR5Δ32 and CCR5Δ1 nucleic acids. However, after two rounds of transfection, the CCR5Δ32 alleles were gradually enriched and became the major alleles with increasing homologous recombination. The intensity of the CCR5Δ32 band was eventually close to 3-fold (2.87-fold, as determined by gray scale measurement) that of the CCR5Δ1 band (Fig. 3B).

At this point, a total of 50 single cells were randomly selected for single cell culture, and DNA was separately extracted from 34 successfully cultured clones. Subsequent PCR analysis using primers F4 (5′-CTCCCAGGAATCATCTTTACC-3′) and R4 (5′-TCATTTCGACACCGAAGCAG-3′), with a short product length of 200 bp, indicated that clones 16, 18 and 25 were homozygous for the CCR5Δ32 mutation, whereas in clones 5, 13, 14, 17 and 32 CCR5Δ1 appeared to be randomly mutated (Fig. 3D). Subsequent gene sequencing confirmed that clones 16, 18 and 25 (representing 8.8% of the transfected alleles, and 37.5% of the targeted alleles) were homozygous for the CCR5Δ32 mutation (Fig. 3E).

Off-targeting analysis of CCR5-TALENs and CCR5Δ1-TALENs

A Basic Local Alignment Search Tool (BLAST) search of the National Center for Biotechnology Information (NCBI) database (https://blast.ncbi.nlm.nih.gov/Blast.cgi) revealed CCR2 to have the most similar genomic DNA sequence to CCR5; therefore, the off-target effects of the two TALENs on this gene were investigated. An analysis investigating the homology between the potential targeting region of CCR2 (4,312–4,711 bp) and that of CCR5 (4,023–4,422 bp), which was the targeting region of the CCR5-TALENs and the CCR5Δ1-TALENs, revealed a score of 73%, analyses of all other potential targeting regions revealed scores <50%. The present study amplified the potential targeting region of CCR2 in CD4+ U87 cells with CCR5Δ32/Δ32 mutations by PCR, followed by T7E1 analysis and gene sequencing (data not shown). The present study confirmed that no off-target CCR2 mutations were generated.

HIV-1 challenge test

Wild-type CD4+ U87 cells and the CCR5Δ32/Δ32 genotype clones were challenged with BaL-HIV-1 for 4 h to assess resistance to infection. The level of p24 in the culture supernatants was assessed every 48 h for the following 12 days by ELISA. In wild-type cultures, p24 was detected after 48 h and peaked after 8 days. Conversely, p24 was not detected in the supernatants of CCR5Δ32/Δ32 genotype cells at any time point (Fig. 4).

Figure 4. HIV-1 challenge. The mean concentration of p24 in the supernatant at 2, 4, 6, 8, 10 and 12 days after challenge was 58.47±2.35, 162.23±4.78, 458.78±27.34, 613.35±26.78, 580.35±24.73 and 483.34±30.85 in wild-type CD+ U87 cells, whereas almost no p24 was detected in CCR5Δ32/Δ32 genotype cultures. CCR5, C-C chemokine receptor type 5; CD4, cluster of differentiation 4; HIV-1, human immunodeficiency virus-1.

Discussion

The present study successfully reproduced the naturally existing CCR5Δ32/Δ32 genotype in CD4+ U87 cells by combining TALENs with a traditional homologous recombination technique. This study reported a novel technique capable of producing site-specific, size-controlled and homozygous DNA deletion within the mammalian genome. Introduction of the CCR5Δ32/Δ32 mutation, rather than entirely knocking out CCR5, may ease concerns regarding the potentially unfavorable clinical effects of CCR5 knockout (18). TALENs-mediated homologous recombination achieved a significantly higher frequency of recombination, without application of selection, compared with traditional homologous recombination techniques without off-target DNA integration into the genome (24).

In order to ensure that the donor CCR5Δ32 genes are saved from further targeting by TALENs, the binding domain of one arm of the TALENs must be located mostly within the to-be-deleted 32 bp region of CCR5 (Fig. 2A and B). Once CCR5 is randomly mutated with insertions or deletions, including CCR5Δ1, new TALENs may be needed to target the randomly mutated CCR5, even though the insertions or deletions did not occur within the binding domains of the original TALENs arms. After repeated transfection with CCR5-TALENs, gene sequencing indicated that the CCR5Δ32/Δ1 genotype remained unchanged. These results suggested that a new pair of TALENs may be required even though mutation did not occur within the binding domains of CCR5-TALENs. Six pairs of CCR5Δ1-TALENs were therefore designed; one of which was revealed to work well. These results indicated that these TALENs are not only specific for the binding domains, but may also be specific for the 3-dimensional structures surrounding the binding domains. This property increases safety of these TALENs, and suggests that off-target effects will be rare. In the present study, the originally designed CCR5-TALENs were no longer functional when a 1 bp deletion had occurred 3 bp downstream of the right arm of the CCR5-TALENs (Fig. 2).

To the best of our knowledge, homozygous deletion of a predetermined length of genomic DNA using ZFNs, TALENs or CRISPR/Cas9 has not been previously reported. The mechanism involved, therefore, requires discussion. Theoretically, homology-directed repair occurs in a manner most consistent with the synthesis-dependent strand annealing (SDSA) model of double strand break (DSB) repair (25,26). However, since constructed CCR5Δ32 donor DNA templates were used, when DSB occurred, one out of two or both 5′ to 3′ broken ends of each DSB, bearing the to-be-deleted sequence, will not find the homologous sequence on the donor DNA templates to anneal with; therefore, SDSA would not be able to start. However, the present study suggested that this non-homologous end could be trimmed off by some unidentified DNA polymerase. Therefore, the DSB repair process with the predetermined 32 bp deletion used in this study may involve the five steps as illustrated in Fig. 5: i) DSB, double strand DNA was broken apart in the targeting region by FokI connected with TALENs; ii) invasion, both 5′ to 3′ ends invaded into the opened double strands of donor DNA and annealed with the corresponding homologous sequences, at least one bearing a non-homologous ‘floating’ tail; iii) trimming, the non-homologous tail is excised by an unidentified DNA polymerase; iv) SDSA, SDSA starts with donor DNA sequences as templates; and v) re-annealing, the two newly synthesized DNA single strands leave the donor DNA templates, re-anneal, and may continue SDSA with each other as templates, if necessary, until the DSB is completely repaired.

Figure 5. Mechanism underlying TALENs-mediated homologous recombination. DSBs were induced by FokI connected with TALENs. The two 5′ to 3′ ends invaded into the opened double strands of donor DNA templates and annealed with the corresponding regions with at least one of them bearing a non-homologous ‘floating’ tail. These non-homologous tails were excised by some unidentified DNA polymerase, followed by SDSA on the donor DNA templates. The two newly synthesized DNA single strands leave the donor DNA templates, re-anneal to each other and may continue to carry on SDSA with each other as templates if necessary until the DSB is completely repaired. CCR5, C-C chemokine receptor type 5; DSB, double-strand break; SDSA, synthesis-dependent strand annealing; TALENs, transcription activator-like effector nucleases.

The homologous recombination frequency after the first three rounds of transfection was 1.7% when the TALENs targeting rate was ≤50.4%. However, after the last two rounds of transfections it reached 8.8%, when the TALENs targeting rate was only 23.5% (Fig. 3). However, dividing these two homologous recombination frequencies by the TALENs targeting rate, indicates that the homologous recombination frequency after the first three rounds of transfection was 3.4%, and after the last two rounds of transfection was 37.5%. To better determine the frequencies of TALENs targeting and homologous recombination, three types of frequency were discussed: i) Frequency of targeting (T7E1 analysis), which reflects how many cells with mutated CCR5 were generated; ii) frequency of homologous recombination, which reflects the number of clones containing the CCR5Δ32 allele (formula: F2=the number of CCR5Δ32 alleles/2× the number of sequenced clones); iii) frequency of homologous recombination in targeted CCR5, which reflects the accuracy of DNA autonomous repair (formula: F3=the number of CCR5Δ32 alleles/the number of targeted alleles including CCR5Δ32 alleles). This homologous recombination difference may be attributed to homologous recombination, as longer homologous arms will have higher homologous recombination frequencies. The total length of the homologous arms for the first three rounds of transfection was only 1,602 bp, whereas for the last two rounds of transfections the homologous arms were the whole chromatids. This significant difference may explain the 10-fold difference in homologous recombination frequency between the last two rounds and first three rounds of transfection. The homologous recombination frequencies could be further improved if the homologous arm length was increased. Notably, since 3 of the 8 targeted mutations edited by CCR5Δ1-TALENs that occurred during DSB repair processes in the last two rounds of transfection were caused by homologous recombination, the DSB induced by TALENs may have a tendency to be repaired, restoring the ‘original’ sequences.

A major disadvantage of the present study is choosing the CD4+−U87 cell line to generate homozygous CCR5Δ32 mutations instead of directly using T lymphocytes. T lymphocytes could be used directly once the technique is ready for clinical application. The present study chose the CD4+−U87 cell line to preliminarily establish the technique simply because these cells grow fast and are easy to manipulate. Furthermore, they mimic T lymphocytes very well in terms of expressing CD4, CCR5 and CXCR4, which is required for the HIV challenge test following gene editing. Since this modified TALENs technique has been well established in editing the U87 cell line, our further studies aim to optimize the technique, and make it easier to alter the CCR5 gene in T lymphocyte cells.

To the best of our knowledge, off-target occurrence should be avoided when using genome editing for therapeutic applications. To minimize off-target modification, the following strategies were employed in the present study: i) Potential off-target sites prediction, after obtaining homozygous CCR5Δ32 mutations in three clones, potential off-target sites were analyzed in the targeting regions of CCR5-TALENs and CCR5Δ1-TALENs. Homology (73%) was detected between CCR5 (4,023–4,422 bp) and CCR2 (4,312–4,711 bp). The CCR2 gene was revealed to possess the highest homology to the targeting region of CCR5, whereas other genes exhibited lower homology (≤50%) when aligned with CCR5 (4,023–4,422 bp), as determined using a BLAST search in NCBI. ii) Detecting off-target modification by T7E1 analysis and gene sequencing; since the CCR2 gene exhibited the highest homology to the targeting region of CCR5, a primer was designed to specifically amplify CCR2 (4,312–4,711 bp) with corresponding template DNA extracted from wild-type cells, CCR5Δ32/Δ1 cells, and clones 16, 18 and 25. Amplified 400 bp CCR2 fragments were analyzed by T7E1 analysis, and the results demonstrated that none were cleaved by T7E1, indicating that no potential off-target effects occurred in the CCR2 homologous region (data not shown). In addition, 400 bp PCR products were ligated into T-easy vector and sequenced (50 successfully transformed clones for each PCR product were sent for gene sequencing) and the sequencing results confirmed that no potential off-target effects occurred in the CCR2 homologous region (data not shown). Since no off-target effects were detected in the CCR2 gene, it is very unlikely that other off-target effects will occur in genes with low homology to CCR5. Furthermore, TALENs, instead of CRISPR/Cas9, were used in the present study to edit the target gene due to the following reasons: i) TALENs exhibit an efficient editing efficacy, although it is lower compared with CRISPR/Cas9 (21); ii) notably, TALENs exhibit much lower off-target modification than CRISPR/Cas9 (27).

In conclusion, to the best of our knowledge, the present study reproduced the CCR5Δ32/Δ32 genotype without selection for the first time in CD4+ U87 cells. This mutation conferred resistance against HIV-1 infection. Our future studies aim to adapt this technique in HSPCs or CD4+ T lymphocytes, producing clinically useful cells for therapeutic use in HIV-positive patients.

Acknowledgements

The present study was supported by grants from the Ministry of Science and Technology of China (grant no. 2013ZX10001-004-002-005), the Science and Technology Department of Hubei Province (grant no. 2012FFA037), the Hubei Provincial Department of Education (grant nos. B2013111 and B2015484) and the Science and Technology Department of Shiyan City Government (grant no. 069S, 2013).

Glossary

Abbreviations

Abbreviations:

References