TZM-bl cells (JC53BL-13 derived from HeLa, Cat No: ARP-8129) were received through the NIH AIDS Reagent Program, (Division of AIDS, NIAID, NIH). The cells were cultured in DMEM medium with 10% FBS and 1X antibiotic and antimycotic solution (Cat No:15240062, Thermo Fisher scientific). Cells were cultured for at least two passages prior to gene-editing. The HIV infection in TZM-bl was measured by luciferase reporter expression which is under the control of HIV-Tat regulatory element (21).

Granulocyte colony stimulating factor (G-CSF) mobilized peripheral blood was collected from healthy donors after obtaining approval from Institutional Review Board (IRB). Peripheral blood mononuclear cells (PBMNCs) were separated using ficoll density gradient centrifugation. CD34⁺ cells were isolated from PBMNCs using EasySep CD34 positive selection kit (Cat No: 17896, STEMCELL Technologies) as per the manufacturer instructions. The purified CD34⁺ cells were pre-stimulated with Serum free essential medium (SFEM) containing hematopoietic stem cell specific cytokines such as SCF (240ng/ml), FLT3 (240ng/ml), TPO (80ng/ml) and IL-6 (40ng/ml) and small molecule cocktail of Resveratrol, Stem Reginin-1 and UM729 (RUS) as described in our previous work (22). All the donors were screened for delta32 genotype before conducting CCR5 gene editing ( Supplementary Figure 3A ). For CD34⁺CD90⁺ and CD34⁺CD90- cell sorting, the purified CD34⁺ cells were briefly cultured with above mentioned cytokines (6-12 hours) and stained with CD90 antibody (8µg/1x10⁶ cells) for 20 min at room temperature. After brief washes with PBS, CD90⁺ and CD90^- cells were sorted using BD FACS Aria III in purity mode. The purity of sorted cells was re-assessed by staining with CD34 and CD90 antibodies.

0.1 million HSCs were seeded on the upper chamber of the transwell plate with 100µl of basal medium and the lower chamber was loaded with 500µl of SFEM II cytokines (SCF (240ng/ml), FLT3 (240ng/ml), TPO (80ng/ml) and IL-6 (40ng/ml) medium containing 100 ng/mL of SDF-1α ligand. The plate was incubated at 37°C for 24 hours. On the next day, cells in the lower chamber were counted using trypan blue dye exclusion method. The counts were normalized with non-SDF-1α containing media. Percentage of migration was calculated with the number of cells in the bottom chamber divided by number of cells placed in the upper chamber.

Based on the experimental requirements, 300 to 500 HSCs were seeded in 3ml of semi-solid methylcellulose medium (Cat No: #04044, Stem cell technologies). After 14-16 days, hematopoietic progenitor colonies were enumerated under microscope. Based on the morphology, colonies were categorized as Burst forming unit-erythroid (BFU-E), Colony forming unit-erythroid (CFU-E), Colony forming unit-granulocyte-monocyte progenitor (CFU-GM) and Colony forming unit- granulocyte, erythrocyte, monocyte, megakaryocyte (CFU-GEMM).

Total RNA was isolated from sort purified HPCs and HSCs using RNeasy mini kit (Cat No:74104, Qiagen, Hilden, Germany) as per the manufacturer’s instructions. 1µg of total RNA was reverse transcribed into cDNA using PrimeScript 1^ststrand cDNA synthesis kit (Cat No: 6110A, Takara) and analyzed by Real-Time PCR (RT-PCR) with TB Green Premix Ex Taq ™II (Cat No: #RR820A, Takara) as per the manufacturer instructions. RT-PCR data were analyzed using the standard 2^−ΔΔCT method and presented as the fold expression normalized to the reference gene Ubiquitin C. Primers used for RT-PCR are listed in Supplementary Table 2 .

CD4 cells were isolated from peripheral blood of the healthy donor using CD4 positive selection kit (Cat No: #17852, stem cell technologies) as per the manufacturer’s instructions and the purity was analyzed using CD3 and CD4 FACS antibodies. Isolated cells were cultured using human XF T cell expansion medium (Cat No: 10981, stem cell technologies) along with CD3/CD28/CD2 T cell activator cocktail (Cat No: 10970, Stem cell technologies). CCR5 gene editing was performed on day 4 of expansion.

The two sgRNAs targeting CCR5 gene (E2C5- UGACAUCAAUUAUUAUACAU (GRCh38.p10, Chromosome 3- 46372915 – 46372973), E3C5-CAGCAUAGUGAGCCCAGAAG (GRCh38.p10, Chromosome 3- 46373133 – 46373191) were designed based on the common hits identified using Synthego (https://design.synthego.com/#/), CHOPCHOP (https://chopchop.cbu.uib.no/) and benchling (https://www.benchling.com/crispr/) tools. Cas9-Ribonucleoprotein (RNP) was electroporated using Lonza 4D nucleofector with program DZ100 for HSCs, CN114 for TZM-bl cells and DN100 for CD4 cells. After 72 hours, cells were genotyped by sanger sequencing. The chromatographs obtained from the sequencing were analyzed by Inference of CRISPR Edits (ICE) tool from Synthego (https://ice.synthego.com/#/) The primers used for the region-specific amplification is listed in Supplementary Table 2 . The deletion induced by dual sgRNA system was analysed using gap PCR and quantified with ImageJ software (https://imagej.nih.gov/ij/download.html).

Macrophage differentiation of HSCs was carried out using the published protocol with minor modifications (23). Briefly, control and edited HSCs were plated in non-tissue culture treated polystyrene plates with macrophage differentiation medium (SFEM-II, SCF (100ng/ml), Flt3-L (50ng/ml), IL-6 (10ng/ml), IL-3 (10ng/ml), GM-CSF (10ng/ml) and M-CSF (10ng/ml). Non-adherent cells were collected every 72 hours and reseeded in the macrophage differentiation medium. Adherent cells were cultured using RPMI medium containing 10% FBS along with GM-CSF (10ng/ml) and M-CSF (10ng/ml). After 14-16 days, adherent cells were observed under microscope for morphology, harvested with Accutase (Material Number: 561527, BD Biosciences), stained for CD4 PE, CD14 FITC, CCR5 APC, CXCR4 APC, CD14 BV421, CD80 FITC, CD206APC, CD64 PE, CD163-PE CF594 and CD71 FITC antibodies and analyzed using BD FACS Aria III flow cytometer. Phagocytic potential of generated macrophages was validated using pHrodo Red E. coli BioParticles conjugate (Cat No, P35361, Thermo scientific) as per the manufacturer’s instructions. The proportion of phagocytosis was calculated by enumerating the phagocytosis positive and negative cells.

All the HIV related experiments were carried out under BSL-2 facility using BSL-3 practices after the approval of Institute Biosafety committee. HIV-1 p49.5 R5-tropic molecular clone (Cat No: ARP-11389) was obtained through NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH. The clone was contributed by Dr. Bruce Chesebro. For HIV production, plasmid was transfected into HEK293T cells using standard calcium phosphate method. 72 hours of post transfection, media containing viral particles was collected and filtered with 0.45µm filter and stored at -80°C as multiple aliquots. HIV production was monitored by measuring p24 antigen using ELISA kit, obtained from R&D biosystems (Cat No: DHP240B). Infectivity of the HIV-1 stock was determined using TZM-bl cells. HIV challenge assay with gene edited TZM-bl cells and macrophages was done using 25ng and 150ng of p24 respectively for 6 hours in growth medium containing polybrene (8µg/ml). HIV infectivity assay in HSPCs and its subsets were performed as described (24). In brief, 0.25 million HSPCs/HPCs/HSCs were seeded in 250µl of SFEM-II medium (with 8µg/ml of polybrene) in 48-well Retronectin coated plates and R5-tropic HIV-1 (200ng of p24) was added. The plate was subjected to spinfection for 30 mins at 900g and incubated at 37°C for overnight with 5% CO₂. On the next day, cells were washed with PBS by centrifugation, cultured for 4 to 6 days and stained with HIV Gag antibody. (KC 57- FITC, Beckman coulter).

Total genomic DNA was isolated from equal number of infected and non-infected HSPCs/HPCs/HSCs and subjected to PCR amplification using HIV specific primers listed in Supplementary Table 2 . The PCR was conducted using HotstarTaq master mix as per the standard PCR protocol using 50ng of DNA template.

All animal experiments were conducted after obtaining approval from institute animal ethical committee, Christian Medical College, Vellore, Tamil Nadu, India. The Nonirradiated NOD, B6. SCID Il2ry^-/- Kit W^41/41 (NBSGW) mice (Jackson laboratory) were bred in inhouse animal facility. Depending on the experimental requirements, 0.5-1x10⁵ gene-edited HSCs were infused into busulfan conditioned 7-8 weeks old female NBSGW mice via tail vein injection. After 16 weeks of infusion, mouse bone marrow, peripheral blood and splenic cells were harvested and analysed for human cell engraftment using human CD45 and mouse CD45.1 antibodies. The percentage of human cell engraftment was calculated using the formula: = (% hCD45)/(% hCD45 + % mCD45.1) x 100. The gene editing in the engrafted cells was analysed by extracting the DNA using Quick DNA Extract (Cat No: QE0905T, Lucigen) and subjected for sequencing with human CCR5 specific primers and ICE analysis. Multilineage repopulation was analyzed with lineage specific antibodies listed in Supplementary Table 1 .

All the statistical analysis was performed using PRISM GraphPad 8 package (GraphPad Software Inc., San Diego, CA, USA). Data analysis was done using two tailed unpaired t-test and multiple T-test using Holm-sidak method as indicated in figure legends. Error bars denotes ± SEM. Number of independent experimental replicates (n), number of donors used are indicated in the figure legend. P value < 0.05 is considered as statistically significant.

Naturally occurring deletion (Δ32) in the coding region of CCR5 is shown to ablate CCR5 expression. Generating CRISPR-Cas9 mediated CCR5 Δ32bp genotype or deletion of coding region in HSPCs needs incision by two sgRNAs of similar efficiency and this procedure increases the chance of off- target editing and chromosomal rearrangements (13, 25, 26). To test the impact of small InDels on CCR5 expression, we used single sgRNA (E2) that targets coding region and compared it with dual sgRNA (E2 and E3) approach that deletes 246 bp in TZM-bl cells ( Supplementary Figure 1A ). High frequency of InDels (>80%) with single sgRNA and deletion (>85%) with dual sgRNA were detected by ICE analysis of Sanger sequencing reads and by gap PCR, respectively ( Figure 1A ). Both the single sgRNA and dual sgRNA approaches resulted in reducing CCR5 expression to <12% and <3%, respectively, compared to un-edited control. ( Figures 1B–D ).

Next, we challenged the single and dual sgRNA edited TZM-bl cells with R5-tropic HIV. While the control cells showed cell death ( Figure 1F ), and high frequency of HIV infection as indicated by luciferase reporter expression ( Figure 1G ) and by intracellular HIV gag p24 staining ( Figures 1E, H ), all the edited conditions showed complete HIV resistance. These observations suggest that single sgRNA mediated InDels are sufficient to provide HIV resistant phenotype similar to deletion of CCR5 coding region.

HIV infection in HSPCs has been reported and this can potentially limit the use of autologous HSPCs for CCR5 gene editing as the transplanted cells can later serve as reservoir for HIV infection (27, 28). HIV receptor-CD4 and co-receptor-CCR5 are crucial for R5-tropic HIV infection and therefore, we examined the expression of CD4/CCR5 on the G-CSF mobilized HSPCs from the healthy donors. In agreement with previous findings (29, 30), HSPCs expressed HIV receptors. While 25% of HSPCs expressed only CD4 (CD34⁺CD4⁺CCR5^- cells) around 6% of HSPCs had both the receptors (CD34⁺CD4⁺CCR5⁺ cells) ( Figure 2A and Supplementary Figure 1C ). To identify the subpopulation of HSPCs that expresses both CD4 and CCR5 (CD4/CCR5) receptors, we sorted HSPCs as CD34⁺CD90^- HPCs and CD34⁺CD90⁺ HSCs (31) ( Supplementary Figure 1B ). The CD34⁺CD90⁺ HSCs contained 7-fold reduced CD4⁺CCR5⁺ receptors than CD34⁺CD90^- HPCs (8.4% vs 1.2%) ( Supplementary Figure 1C and Figure 2B ).

To validate this finding, we infected the CD34⁺ HSPCs, CD34⁺CD90^- HPCs and CD34⁺CD90⁺ HSCs with R5-tropic HIV. Consistent with the pattern of HIV receptor expression, the CD34⁺CD90⁺ HSCs displayed lesser HIV gag p24 staining than the HPCs and HSPCs (3-fold and 2-fold, respectively) ( Figures 2C, D ). Next, on examination of HIV-1 proviral DNA integration, the HSPCs and CD34⁺CD90^- HPCs had proviral DNA but not the CD34⁺CD90⁺ HSCs ( Figure 2E ). To further confirm that CD34⁺CD90⁺ HSC fraction effectively resist R5- tropic HIV infection, HIV infected HSPCs, HPCs and HSCs were cultured in monocyte stimulation medium for 7 days. While, both HSPCs and CD34⁺CD90^- HPCs had 2-fold increase in viral outgrowth post-stimulation, such increase was not detected in CD34⁺CD90⁺ HSCs ( Figure 2F ).

The presence of high levels of antiviral restriction factors have shown to restrict HIV-1 infection in resting CD4 T cells (32–34). To examine whether such phenomenon exists in the CD34⁺CD90⁺ HSCs, we investigated the expression of anti-viral restriction factors. Real time PCR quantification indicated the high expression of IFITM2, and SAMHD1 in the CD34⁺CD90⁺ HSC fraction when compared with the CD34⁺CD90^- HPCs. In particular, SAMHD1 which is shown to restrict HIV infection (33) was expressed 2.5-fold higher in CD34⁺CD90⁺ HSCs ( Figure 2G ). All these findings suggest that reduced expression of HIV receptors and presence of higher levels of antiviral restriction factors are mediating the resistance of CD34⁺CD90⁺ HSCs to HIV infection. These findings also suggest that CD34⁺CD90⁺ HSCs are ideal target cells for CCR5 gene manipulation for HIV gene therapy.

Recent study showed that the gene editing of BCL11A binding site in CD34⁺CD90⁺ HSCs reduced the requirement of target cell population by 10-fold for gene-manipulation and resulted in durable engraftment in non-human primates (35). We reported that sort enriched CD34⁺CD90⁺ HSCs can be preserved with a small molecule cocktail of Resveratrol, UM-729 and SR-1 (RUS) (22).

To test whether RUS treatment could retain the stemness of sort-enriched CD34⁺CD90⁺ HSCs for CCR5 gene editing, the purified HSPCs were sorted for CD34⁺CD90⁺ HSCs. The CD34⁺CD90⁺ HSCs sorting procedure enriched the most primitive HSCs, marked as CD34⁺CD133⁺CD90⁺CD45RA^-CD38^-CD49f⁺ cells (22), by >80% ( Supplementary Figure 2A ). The sorted cells were pre-stimulated with cytokines for 48 hours with or without RUS before gene editing with Cas9 RNP targeting CCR5. The CCR5 gene edited cells cultured with RUS displayed increased retention of CD34⁺CD90⁺ cells ( Supplementary Figure 2B and Figure 3A ) and high frequency of CD34⁺CD133⁺CD90⁺CD45RA^-CD38^-CD49f⁺ HSCs ( Figure 3B ). Consistent with surface expression analysis, the colony formation assay (CFU) showed high frequency of GEMM colonies ( Supplementary Figure 2C ). The RUS treated CD34⁺CD90⁺ cells showed 2-fold increase in CXCR4 expression, a factor crucial for stem cell homing in the bone marrow ( Supplementary Figure 2D ) and thereby resulted in 2-fold greater response towards SDF-1α cytokine ligand ( Figure 3C ). Additionally, sodium nitroprusside (SNP) treatment mediated activation of nitric oxide signaling is shown to increase the CXCR4 expression, transwell migration and homing of HSPCs isolated from umbilical cord blood (36). Remarkably, RUS treatment suffices to improve the SDF1-α mediated transwell migration of the CD34⁺CD90⁺ HSCs at comparable levels as that of SNP treated CD34⁺CD90⁺ HSCs ( Figure 3C ). RUS treatment also showed modest increase in the editing frequency (79% vs 60% in vehicle) ( Figure 3D ).

To confirm that RUS treated CD34⁺CD90⁺ HSCs have superior engraftment potential in vivo, we transplanted the CD34⁺CD90⁺ HSCs that were cultured for 48hrs with cytokines and with or without RUS into NBSGW mice. The RUS treated cells displayed a higher human cell chimerism in bone marrow ( Figure 3E ), spleen ( Figure 3F ) and peripheral blood ( Figure 3G ) than the vehicle treatment. Strikingly, infusion of 50,000 RUS treated CD34⁺CD90⁺ cells was sufficient to repopulate approximately 56% of mouse bone marrow ( Figure 3H ). The RUS treatment did not alter the multi lineage repopulation capacity of the CD34⁺CD90⁺ cells and we detected lineages such as HSPCs (CD34), T cells (CD3), myeloid cells (CD13), B cells (CD19) and monocytes (CD14) ( Figure 3I ). These results demonstrate that RUS supplementation during culture of HSCs preserves stemness and provides an improvised culture condition for CCR5 gene editing.

The CD4 T-cells and CD34⁺HSPCs are the currently explored grafts for CCR5 gene editing. To utilize the CD34⁺CD90⁺ HSCs for CCR5 gene editing, The RUS treated CD34⁺CD90⁺ HSCs were edited for CCR5 with different concentrations of Cas9-RNP complex and >90% InDels were observed with 100 pmol of RNP ( Figure 4A and Supplementary Figure 3B ). The CCR5 edited cells produced similar number and pattern of multilineage colonies as the control cells in the CFU assay ( Figure 4B ). The edited CD34⁺CD90⁺ HSCs were single cell sorted in methocult medium and the analysis of clonal colonies showed that >85% cells were bi allelic edited cells ( Figure 4C ).

In-vitro macrophage differentiation of CCR5 edited CD34⁺CD90⁺ HSCs generated comparable macrophage cell yield ( Supplementary Figure 3C ), expression of respective lineage markers CD14, CD71, CXCR4, and CD4 ( Figure 4D ) and subsets of macrophages (M1 and M2) to that of control ( Figure 4E ). Importantly, 90% of the macrophages lacked the CCR5 expression ( Figure 4D ). The lack of CCR5 expression has not affected the phagocytosis, an important function of macrophages ( Supplementary Figures 3D, E ).

Next, we infected the macrophages with the R5-tropic virus. We observed an active R5-tropic HIV infection in control macrophages whereas >80% of CCR5 modified cells were resistant to infection as determined by HIV gag p24 staining ( Figure 4F and Supplementary Figure 3F ). Of note, InDel patterns generated during CCR5 editing of CD34⁺CD90⁺ HSCs were conserved in the TZM-bl cells, CD4 T-cells, HSPCs, and the macrophages ( Figure 4G ). Such InDel pattern in CD4 T cells resulted in CCR5 null CD4 cells suggesting the uniform functional outcome ( Figure 4H ). All these suggest that CCR5 editing in CD34⁺CD90⁺ HSCs generates CCR5 null lineage cells which are HIV resistant and functionally intact.

As RUS treatment preserves the engraftment potential of CD34⁺CD90⁺ HSCs, we hypothesized that a low dose of RUS treated CCR5 edited CD34⁺CD90⁺ HSCs could repopulate the mouse bone marrow. We tested this by sorting the CD34⁺CD90⁺ HSCs, cultured it for 48hours with RUS cocktail and gene edited 1x10⁵ CD34⁺CD90⁺ HSCs with Cas9-RNP targeting CCR5 loci or Cas9- tracrRNA as a control. The crRNA less Cas9- tracrRNA control will not induce DNA double strand breaks and thus helps to better understand any gene-editing associated engraftment defect.

At 16^th week post transplantation, into NBSGW mice, we observed that the CCR5 edited cells engrafted as efficiently as the control cells with the mean human cell chimerism in bone marrow ( Supplementary Figure 4A and Figure 5A ), spleen ( Figure 5B ) and peripheral blood ( Figure 5C ) of about > 70%, 60% and 20% respectively. This analysis indicates that CCR5 gene editing does not affect the engraftment of CD34⁺CD90⁺ HSCs and a dose of 1x10⁵ CD34⁺CD90⁺ HSCs is sufficient to repopulate 70% of the bone marrow. Multilineage analysis in bone marrow showed the formation of myeloid and lymphoid lineages with no lineage bias in the CCR5 edited group ( Figure 5D ).

Tissue trafficking of CCR5^- cells is crucial to eradicate the established HIV-1 reservoirs. To test whether CCR5 edited cells can lodge in to different mouse tissues, we analysed the CCR5 editing frequency in the cells recovered from different tissues. InDel analysis showed reconstitution of CCR5 edited cells in bone marrow, spleen, peripheral blood and brain. Importantly, the frequency of gene-editing was maintained from that of the infused product ( Figure 5E ). InDel pattern analysis confirmed that the HSPCs with the prominent +1 and -1 InDels were retained on long-term repopulation, while the cells with 14bp deletion diminished, suggesting that HSPCs with 14bp deletion is not competent for long-term repopulation ( Figure 5F ). Similar loss of 13bp deletion has been reported with gene editing of γ-globin promoter (35).

To test the functional proficiency, macrophages were generated from the engrafted human cells and that showed > 90% of cells lacking CCR5 expression in the CCR5 edited group with no defects in macrophage generation or maturation when compared with the control ( Figure 5G and Supplementary Figure 4B ). When these macrophages were challenged with infectious R5-tropic HIV, > 80% of cells showed resistance in the CCR5 edited group, consistent with the genotype and CCR5 expression ( Figure 5H ).