Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 May 3;25(5):1168-1186.
doi: 10.1016/j.ymthe.2017.03.012. Epub 2017 Mar 30.

In Vivo Excision of HIV-1 Provirus by saCas9 and Multiplex Single-Guide RNAs in Animal Models

Affiliations
Free PMC article

In Vivo Excision of HIV-1 Provirus by saCas9 and Multiplex Single-Guide RNAs in Animal Models

Chaoran Yin et al. Mol Ther. .
Free PMC article

Abstract

CRISPR-associated protein 9 (Cas9)-mediated genome editing provides a promising cure for HIV-1/AIDS; however, gene delivery efficiency in vivo remains an obstacle to overcome. Here, we demonstrate the feasibility and efficiency of excising the HIV-1 provirus in three different animal models using an all-in-one adeno-associated virus (AAV) vector to deliver multiplex single-guide RNAs (sgRNAs) plus Staphylococcus aureus Cas9 (saCas9). The quadruplex sgRNAs/saCas9 vector outperformed the duplex vector in excising the integrated HIV-1 genome in cultured neural stem/progenitor cells from HIV-1 Tg26 transgenic mice. Intravenously injected quadruplex sgRNAs/saCas9 AAV-DJ/8 excised HIV-1 proviral DNA and significantly reduced viral RNA expression in several organs/tissues of Tg26 mice. In EcoHIV acutely infected mice, intravenously injected quadruplex sgRNAs/saCas9 AAV-DJ/8 reduced systemic EcoHIV infection, as determined by live bioluminescence imaging. Additionally, this quadruplex vector induced efficient proviral excision, as determined by PCR genotyping in the liver, lungs, brain, and spleen. Finally, in humanized bone marrow/liver/thymus (BLT) mice with chronic HIV-1 infection, successful proviral excision was detected by PCR genotyping in the spleen, lungs, heart, colon, and brain after a single intravenous injection of quadruplex sgRNAs/saCas9 AAV-DJ/8. In conclusion, in vivo excision of HIV-1 proviral DNA by sgRNAs/saCas9 in solid tissues/organs can be achieved via AAV delivery, a significant step toward human clinical trials.

Keywords: AAV; Cas9; HIV; Tg26 mice; animal models; gRNA; gene therapy; genome editing; humanized BLT mice; viral latency.

Figures

None
Figure 1
Figure 1
saCas9/sgRNA Efficiently Excises the HIV-1 Genome and Suppresses HIV-1 Luciferase Reporter Expression (A) Location of sgRNA target sites and PCR primers (arrows). (B and C) Efficiency comparison between saCas9/sgRNA and the spCas9/sgRNA system using the EcoHIV-eLuc reporter assay (B) and direct PCR genotyping (C). The excision efficiency (%) indicates the density of the fragmental deletion band(s) over the total bands. Arrows point to the representative fragmental 296-bp deletion between LTR-1 and LTR-3 target sites at the third nucleotide from PAM and to an additional 180-bp insertion, which were validated by TA cloning and Sanger sequencing (C). (D and E) Three LTR sgRNAs paired with viral structural sgRNAs induced robust inhibition of EcoHIV-eLuc activity by the ONE-Glo Luciferase Assay (D) and fragmental deletion/insertion by direct PCR genotyping with the indicated primers (E). The representative fragmental deletion in the dotted box in (E) was corroborated with Sanger sequencing (see Figure S1). HEK293T cells in six wells of the 96-well plate were co-transfected with the pNL4-3-EcoHIV-eLuc reporter (10 ng/well), the pCMV-renilla-luciferase reporter (2 ng/well), and the indicated Cas9/sgRNA-expressing vectors as follows: for the spCas9 system (B), pLV-EF1a-spCas9-T2A-RFP (80 ng/well) and the indicated sgRNA expression vectors (60 ng/well each for paired gRNAs); for the saCas9 system (B and D), pX601-saCas9/LTR sgRNA expression vectors (100 ng/well each for indicted pairs). After 48 hr, eLuc activity in the cell lysates was measured with the ONE-Glo luciferase assay system and renilla-luciferase activity was measured with the Renilla-Luciferase Assay System. Data represent the mean ± SD of four independent transfections with percentage changes in eLuc after renilla-luciferase normalization as compared with the corresponding empty pX601 vector. Two additional wells of the transfected cells were used for direct PCR genotyping (C and E). Similar experiments were repeated two to three times.
Figure 2
Figure 2
Efficacy of Multiplex sgRNAs in an All-in-One AAV Vector (A and B) Duplex sgRNAs/saCas9 exhibited stronger inhibition of EcoHIV-eLuc activity (A) and excision of the HIV-1 genome (B). The excision efficiency (%) indicates the density of the fragmental deletion band over the total bands (B). (C–J) Quadruplex sgRNAs/saCas9 exhibited stronger inhibition of EcoHIV-eLuc activity (C) and higher excision efficiency at both 5′-LTR-1/LTR-3 with GagD (D and E) or GagD/PolB with 3′-LTR (F–J). The primer T361/T458 (D) detected fragmental deletions between 5′-LTR-1/LTR-3 and GagD (arrows) and an additional insertion was observed in both the duplex and quadruplex groups (arrowhead). T710/T458 detected the deletion between 5′-LTR-3 and GagD (E). Deletions between GagD or PolB and 3′-LTR-1/LTR-3 were detected by the primer T758/T363 (F) and T689/T363 (G), respectively. The primer pair T689/T711 detected the deletion between PolB and 3′-LTR-1 (H). The genomic DNA was normalized with saCas9 and β-actin (J). HEK293T cells in a 96-well plate were co-transfected with the EcoHIV-eLuc reporter plus two individual monoplex sgRNA-expressing vectors (100 ng each), one duplex sgRNA-expressing vector (100 ng) with an empty pX601 vector (100 ng) to achieve an equal amount of DNA (200 ng, A and B), or one duplex/quadruplex sgRNA-expressing vector and the sgRNA control (100 ng each, C–J). After 48 hr, the ONE-Glo Luciferase Assay (A and C) was performed as described in Figure 1, and direct PCR genotyping (B and D–J) was performed using the indicated primer pairs.
Figure 3
Figure 3
AAV-DJ-Mediated Delivery of the Multiplex sgRNAs/saCas9-Expressing Vector Effectively Excises the HIV-1 Integrated Genome in NSCs from Tg26 Transgenic Mice (A–C) Similar infection efficiency of monoplex (A), duplex (B), and quadruplex (C) in NSCs at 20 days post-infection with a fMOI of 10. The infected cells were determined by immunofluorescence cytochemistry with anti-HA tag antibody. (D–F) Direct PCR analysis validated the efficient delivery of transgene saCas9 (D) and LTR-1/GagD (E) and excision of 5′-LTR/GagD and GagD/3′-LTR (F) at 2 days post-infection. Relative fold changes or excision efficiency (%) were calculated via densitometry. (G) Quadruplex sgRNAs/saCas9 showed stronger cleavage efficacy than the duplex sgRNAs/saCas9 at 20 days post-infection with an fMOI of 10. The primers T361/T458 detected the deletions between 5′-LTR-1/LTR-3 and GagD, and the primers T758/T645 detected the deletions between GagD and 3′-LTR-1/LTR-3 (F and G). GAPDH was used for normalization of genomic DNA. Control-882 AAV-DJ virus was used as a nonfunctional control. GAPDH, glyceraldehyde 3-phosphate dehydrogenase; NSC, neural stem cell.
Figure 4
Figure 4
Quadruplex sgRNAs/saCas9 AAV-DJ/8 Induces HIV-1 Proviral DNA Excision and a Robust Reduction in HIV-1 RNA Transcription in Most Organs/Tissues of Tg26 Transgenic Mice (A and B) Tg26 mice were injected via the tail vein with purified AAV-DJ/8 virus (1.535 × 1012 GC/mouse). Two weeks after injection, mice were euthanized and their tissues were collected for genomic DNA extraction and PCR genotyping. The positive control represented excision of the EcoHIV-eLuc reporter in HEK293T cells transfected with AAV-saCas9/sgRNA (LTR-1 plus GagD) vector. The negative control represented the mice injected with mock AAV virus. (C) An additional injection of AAV-DJ/8 was given at 2 weeks after the first injection and tissue samples were harvested 2 weeks later for PCR genotyping with the indicated primers. The representative deletion fragments, indicated by the dotted boxes in (B), were verified with TA cloning and Sanger sequencing (see Figure S6). (D) Diagram showing the location of RT-qPCR primer pairs for HIV transcripts. (E–G) RT-qPCR analysis showing a robust decrease (p < 0.001) in the levels of HIV-1 RNA transcripts in sgRNAs/saCas9-treated mice as compared with the control mice injected with mock AAV virus. Data represent the mean ± SD of triplicate experiments after normalization with housekeeping gene Ppia.
Figure 5
Figure 5
Bioluminescence Imaging Analysis Shows that Quadruplex sgRNAs/saCas9 AAV-DJ/8 Induces Excision of EcoHIV-eLuc In Vivo (A) AAV-DJ/8 was administered via a retro-orbital injection into the blood sinus of the right eye of each mouse (n = 3) right after EcoHIV-eLuc inoculation via the same injection route. Representative bioluminescence imaging of one mouse on days 6, 9, 12, and 19 post-EcoHIV inoculation is shown. All of the images were measured for radiance (photon/second/Sr2) and pseudocolored with the same rainbow scale for fair comparison. (B) Total flux of bioluminescence (total photons/second, P/S) measured from the entire body of each group (n = 3/group until day 19 [n = 2]) on each indicated day post-EcoHIV-eLuc inoculation. (C) Bioluminescence output from EcoHIV-eLuc-infected cells in each mouse was measured from the ROI defined on the right eye for comparison. Data represent the mean ± SD, with statistical significance indicated by an asterisk in each comparison (p < 0.05, Student’s one-sided t test). (D) The bioluminescence output from the neck lymph node of each mouse (n = 3 until day 19 [n = 2]) was used for comparison of the efficacy of saCas9/gRNA treatment longitudinally on HIV excision (p values are two-sided using a linear mixed-effects model).
Figure 6
Figure 6
PCR Genotyping and qPCR Analysis Validated the Gene Transduction and Excision Efficiency of EcoHIV-eLuc by Quadruplex sgRNAs/saCas9 AAV-DJ/8 in Various Organs/Tissues of Mice (A–C) Various extents of gene transduction in indicated tissues/organs for saCas9 (A), GagD (B), and LTR-1 (C) at 19 days after infection with quadruplex sgRNAs/saCas9 AAV-DJ/8 (upper panel) or empty control AAV-Cre-DJ/8 (lower panel). (D–F) Conventional PCR genotyping showing various extents of EcoHIV DNA excision between GagD and the 3′-LTR in each indicated tissue/organ (upper panel) compared with the negative control (lower panel). (G–I) qPCR analysis of three indicated fragmental excision types in different tissues/organs selected from B1M1 (G), B1M2 (H) and B1M3 (I). Data represent the mean ± SD of triplicate reactions and are expressed as relative levels over the internal uncut (non-targeting) region of EcoHIV DNA in the same tissues/organs. BxMx, box and mouse number.
Figure 7
Figure 7
Excision of HIV Proviral DNA by Quadruplex sgRNAs/saCas9 AAV-DJ/8 in Humanized BLT Mice Inoculated with HIVNL-BaL-eLuc at 2, 3, and 4 Weeks after AAV Treatment Conventional PCR (35 cycles) using primer pair T361/T458 (for amplifying 5′-LTR/Gag) was used to detect the presence of HIV proviral DNA, which showed a weak wild-type band (1.4 kb) in the lung tissue of the B6M3 negative control without saCas9/sgRNA treatment (C). Nested PCR (35 cycles) using primer pair T361/T946 (for 5′-LTR/Gag) with the template from the first-round PCR (22 cycles with the primer pair T361/T458) identified the wild-type DNA fragment (1.13 kb) of HIV-1 proviral DNA in most, if not all, organs/tissues of each BLT mouse (A–F) and the PCR product of fragmental deletions resulted from HIV-1 excision in the heart, colon, and vagina of HIV-infected BLT mice treated with saCas9/sgRNA (A). Nested PCR with the primer pair T758/T363 (for Gag/3′-LTR) using the template from the first-round PCR (25 cycles, T758/T425) identified the fragmental deletion bands in organs/tissues of saCas9/sgRNA-treated BLT mice (A, B, D, and E) but not the untreated BLT mice (C and F). The fragmental deletion bands resulted from saCas9 editing were framed with dotted boxes followed by TA cloning and Sanger sequencing for validation (see Figures S10 and S11). The positive control represents the excision of the EcoHIV-eLuc reporter in HEK293T cells transfected with AAV-saCas9/sgRNA (LTR-1 plus GagD) vector. BxMx, box and mouse number.

Comment in

Similar articles

See all similar articles

Cited by 65 articles

See all "Cited by" articles

Publication types

MeSH terms

Substances

Feedback