Rapid, efficient and activation-neutral gene editing of polyclonal primary human resting CD4+ T cells allows complex functional analyses

CD4+ T cells are central mediators of adaptive and innate immune responses and constitute a major reservoir for human immunodeficiency virus (HIV) in vivo. Detailed investigations of resting human CD4+ T cells have been precluded by the absence of efficient approaches for genetic manipulation limiting our understanding of HIV replication and restricting efforts to find a cure. Here we report a method for rapid, efficient, activation-neutral gene editing of resting, polyclonal human CD4+ T cells using optimized cell cultivation and nucleofection conditions of Cas9–guide RNA ribonucleoprotein complexes. Up to six genes, including HIV dependency and restriction factors, were knocked out individually or simultaneously and functionally characterized. Moreover, we demonstrate the knock in of double-stranded DNA donor templates into different endogenous loci, enabling the study of the physiological interplay of cellular and viral components at single-cell resolution. Together, this technique allows improved molecular and functional characterizations of HIV biology and general immune functions in resting CD4+ T cells. A nucleofection-based method to enable efficient gene editing of primary human resting CD4+ T cells.

This resulted in a workflow ( Fig. 1a and Methods) where resting CD4 + T cells are nucleofected once with RNPs and then cultivated under specific conditions for up to 6 weeks until assessment of gene editing and functional characterization. To establish the KO efficacy by this protocol, we first targeted CD46, which encodes a single-span transmembrane cell surface receptor and functions as an important regulator of the complement system 17 . CD46 is homogenously expressed on resting CD4 + T cells and other hematopoietic cells 18 . Nucleofection of resting CD4 + T cells with RNPs containing CD46-targeting gRNAs did not affect cell viability compared to untreated cells (wild-type (WT)) with over 80% viable cells for up to 5 weeks (Fig. 1b). Based on the determination of absolute cell numbers (Extended Data Fig. 1c) we established that nucleofection of approximately 2.5-fold more cells than required for downstream functional characterization represents a reliable workflow.
Nucleofected cells maintained their resting phenotype as assessed by multiple established markers in comparison to activated CD4 + T-cell reference cultures (Fig. 1c,d and Extended Data Fig. 1e-g). Three different CD46-gRNAs, targeting the CD46 locus, were tested and the gene editing efficiency was quantified by deep sequencing 1 week after nucleofection (Extended Data Fig. 2b). The combination of the two most efficient CD46-targeting gRNAs (gRNA1 and gRNA2) resulted in editing of CD46 in 85% of cells (Extended Data Fig. 2b,c) and an almost complete loss (97.3 to 98.9%) of CD46 cell surface expression 2 weeks after nucleofection as assessed by flow cytometry (Fig. 1e) and confocal microscopy (Extended Data Fig. 2d). Similar results were obtained for the cell surface receptor P-selectin glycoprotein ligand-1 (PSGL-1, CD162). Again, the combination of two PSGL-1/SELPLG-targeting gRNAs (gRNA2 + 3) resulted in the best editing efficacy (93.4%) (Extended Data Fig. 3a-f) and exposure of the PSGL-1 receptor on the surface of resting CD4 + T cells became almost undetectable 2 weeks after nucleofection (Extended Data Fig. 3g). Thereafter, combinations of two or three pre-tested gRNAs were successfully used for all genes targeted. Of note, the KO efficiency per se was independent of the cultivation and activation conditions following nucleofection (Extended Data Fig. 3h).
Next, we wanted to test the applicability of this protocol for a multi-gene KO approach. As such, we targeted the expression of five genes with roles in HIV-1 infection (CD4, CXCR4, PSGL-1, TRIM5α and CPSF6) as well as CD46, in part using gRNAs pre-tested for each target for the efficiency of editing (Extended Data Fig. 4a) and protein loss was assessed in part by confocal microscopy (Extended Data Fig. 2d). The use of 12 different gRNAs in a single RNP complex nucleofection resulted in a reduction of the expression of all six proteins to almost undetectable levels at the cell surface as determined by flow cytometry 2 weeks after nucleofection for CD46, CXCR4, PSGL-1 and CD4 (Fig. 2a) or by immunoblotting for TRIM5α and CPSF6, 25 d after nucleofection (Fig. 2b). Of note, the viability of the six-gene KO cells was around 20% lower compared to CD46 KO only or untreated WT cells (Fig.  2c). Importantly, while polyclonal resting six-gene KO CD4 + T cells remained negative for activation markers under standard IL-7 and IL-15-containing cultivation conditions (Fig. 2e), they could be readily activated two weeks after nucleofection (Fig. 2d, n/a n/a n/a n/a n/a n/a . P values were corrected for multiple comparison (Tukey). ***P ≤ 0.001. b, Immunoblot analysis of cell lysates from the experiment shown in a 25 d after nucleofection. All targets were validated on the same membrane by re-probing. One representative experiment is shown (n = 2). c, Viability of cells with CD46-single KO, four-gene KO (CD46, CD4, CXCR4 and PSGL-1) or six-gene KO (as in a) were analyzed. WT cells served as control. Means are shown (n = 2). d,e, Resting cells were nucleofected, cultivated for 2 weeks and then either activated or not for one additional week before analyzing expression of the indicated surface receptors (d) and T-cell activation markers CD25, CD69, CD38 and HLA-DR (e) by flow cytometry. Means are shown (n = 2). f, Off-target analysis in resting CD4 + T cells following six-gene KO (as in a). Specific primers were designed to target the top two off-target coding loci predicted for each gRNA used (Extended Data individual turnover pathways, subcellular localization and resulting half-lives, the timespan required for a pronounced depletion is expected to be a specific property of each target protein and needs to be determined individually 19 . We compared the editing of three protein-encoding genes chosen as examples for different protein turnover rates and established their functions in viral infections or cell migration: CD46, described above, CXCR4, a seven-transmembrane chemokine cell surface receptor and HIV co-receptor 20 and the HIV-1 restriction factor SAMHD1, a soluble deoxynucleoside triphosphate triphosphohydrolase localized in the nucleus and cytoplasm of resting CD4 + T cells 3 . Following nucleofection of resting CD4 + T cells with gene-specific RNP complexes, CXCR4 surface expression became undetectable within 1 week of cultivation (Fig. 3a), whereas CD46 was only partially depleted at this time point and reached a near complete loss of surface exposure at week 2 ( Fig. 1e) or 3 (Fig. 3b). The kinetic analysis of SAMHD1 levels by immunoblotting of gene-edited cells revealed an even slower decay of the lentiviral restriction factor with residual levels still detectable after 4 weeks and complete depletion of cellular SAMHD1 pools seen only 6 weeks after nucleofection ( Fig. 3c and Extended Data Fig. 4b).
Functional characterization of single-gene knockouts for cell migration and virus infection. This time-dependent loss of specific protein expression in resting CD4 + T cells had marked functional consequences. Depletion of CXCR4, which acts as receptor for the α-chemokine SDF-1α, abrogated the ability of CXCR4 KO cells to migrate across a Transwell membrane in response to its natural ligand ( Further, consistent with the role of CD46 as the major receptor for measles morbillivirus (MeV) 18 , infection of resting CD4 + T cells carrying a CD46 KO with a MeV-GFP reporter virus was reduced up to 4.2-fold compared to untreated WT cells or NTC-nucleofected cells (Fig. 3g). Finally, to assess the impact of the progressive loss of the restriction factor SAMHD1 on HIV-1 infection in resting CD4 + T cells from healthy donors, we challenged SAMHD1 KO cells and NTC cells with a replication-competent X4 HIV-1 GFP reporter virus, which carries the Vpx-interaction motif in the Gag p6 protein (HIV-1* GFP), at different time points after nucleofection. Two weeks after nucleofection, reduced levels of SAMHD1 ( Fig. 3c and Extended Data Fig. 4b) in SAMHD1 KO cells enhanced HIV-1* GFP infection significantly, by 2.5-fold, compared to NTC reference cells (Fig. 3h, top). This difference steadily increased for infections at subsequent time points reaching a maximum of ∼ninefold when cells were infected 6 weeks after nucleofection, at which time SAMHD1 was undetectable in KO cells (Fig. 3c,h (top) and Extended Data Fig. 4b). In contrast, HIV-1* GFP virions engineered to package the Vpx protein (HIV-1* GFP + Vpx) that recruits SAMHD1 to a cullin 4A-RING E3 ubiquitin ligase (CRL4-DCAF1), which targets the enzyme for proteasomal degradation 5,6 , did not show a significant difference of infection between SAMHD1 KO and NTC cells (Fig. 3h (bottom) and Extended Data Fig. 7 for absolute numbers of HIV-1 GFP + cells). This is consistent with results from earlier work describing the effects of the lentivirus-encoded antagonist Vpx on SAMHD1 (refs. 3,4,7 ). Together, these results establish that RNP nucleofection and cultivation of resting CD4 + T cells according to our protocol enables the functional characterization of genetically depleted host cell factors for the infection of HIV-1 and other viruses (like MeV).
CPSF6, but not PSGL-1 and MX2, is a critical factor for HIV-1 infection in resting CD4 + T cells. We next investigated the relevance of three host factors that have been shown to affect HIV-1 infection in other cell systems [23][24][25][26] , but whose role during infection of resting CD4 + T cells has not been defined. We generated KOs for either PSGL-1, MX2 or CPSF6. As described above, surface-exposed PSGL-1 was completely depleted 2 weeks after nucleofection. PSGL-1 has been shown to reduce virion infectivity and post-entry steps when expressed on HIV-1 producer or target cells (activated cells or cell lines), respectively 23,24,27 . For the first time, we were able to assess the potential impact of PSGL-1 on entry and post-entry steps in resting CD4 + T target cells. A CXCR4 KO was used as positive control, diminishing HIV-1 fusion and productive infection to background levels ( Fig. 3e,f). In contrast, PSGL-1 KO cells retained their virion fusion capacity and permissivity to HIV-1 infection comparable to NTC control cells (Fig. 3e,f). Thus, PSGL-1 does not act as an entry or post-entry restriction factor for HIV-1 in resting CD4 + T target cells.
In KO cells, the host restriction factor MX2 and the host dependency factor CPSF6 became almost undetectable 4 weeks after nucleofection (Fig. 3i). Consistent with their reported role as cellular interactors of the HIV-1 capsid during nuclear entry of the pre-integration complex and proviral integration 25,26 , depletion of MX2 or CPSF6 proteins did not significantly impact on HIV-1 entry (Fig. 3j). To assess the effects of these KOs on productive HIV-1 infection, resting CD4 + T cells were challenged with HIV-1 GFP at two different multiplicities of infection (MOIs) and reporter gene expression was determined by flow cytometry 3 d later. Infection levels of MX2 KO and NTC cells were found to be largely comparable (Fig. 3k). While MX2 depletion increased HIV-1 infection rates in single-round analyses in other cell types between 3-and 11-fold 25,28 , our results suggest that MX2 is not a relevant HIV-1 restriction factor in resting CD4 + T cells. In contrast, loss of CPSF6 markedly reduced the efficacy of HIV-1 infection, by up to 9.8-fold, compared to the NTC reference in these primary cells (Fig. 3k). Previously HIV-1* GFP + Vpx CRISPR-Cas9-mediated transgene knock in into multiple loci in resting CD4 + T cells. Beyond the disruption of individual gene expression to identify their overall function, gene editing allows the introduction of specific mutations or protein tags for detailed mechanistic studies, for example by knock in (KI) via homology-directed repair (HDR) 30 . In activated T cells, KI levels of up to 69% have been reported 30 . To assess the suitability for locus-specific KIs in resting CD4 + T cells, we designed a strategy to modify the SAMHD1 locus by insertion of an enhanced GFP-encoding reporter gene (eGFP) including a stop codon immediately downstream of the nuclear localization signal (NLS) of SAMHD1. In principle, this should result in GFP expression from the endogenous SAMHD1 promoter, while transcription of SAMHD1 itself is disrupted (Fig. 4a). As a donor template for HDR we used a double-stranded DNA (dsDNA)    CD4-GFP GFP-RAB11A GFP-BATF with around 550 bp homology arms from the region targeted by SAMHD1-gRNA2. While nucleofection of resting CD4 + T cells with dsDNA reduced cell viability in the absence of GFP reporter expression (Fig. 4b, top), nucleofection of dsDNA together with the RNP complex targeting the exon 1 of SAMHD1 resulted in GFP expression in >20% of these cells (Fig. 4b, bottom), indicating that the dsDNA donor template was successfully integrated and the GFP reporter was expressed. In parallel, the eGFP KI into SAMHD1 was validated by a locus-specific PCR fragment amplification (Extended Data Fig. 8a). The use of single-stranded DNA (ssDNA) instead of dsDNA as donor template increased cell viability, in line with a recent report 30 , but markedly decreased KI efficiency (Extended Data Fig. 8b).
To examine whether this approach is widely applicable we performed KIs of the eGFP reporter cassette into the locus of three different genes (BATF, RAB11A and CD4; Fig. 4c) to generate fusion proteins, using reagents previously reported for activated T cells 30 . This protocol resulted in the expression of GFP fusion proteins from the endogenous promoters in resting CD4 + T cells 2 weeks after nucleofection as detected by flow cytometry (Fig. 4d) and by confocal microscopy in a typical subcellular localization ( Fig. 4e; BATF, nuclear; RAB11A, intracellular compartment reminiscent of endosomes; CD4, plasma membrane). The KI efficiency and detectable expression of the GFP fusion proteins in resting cells varied from 0.9% (BATF) to 10.5% (RAB11A) (Fig. 4d), most likely reflecting differences in locus-specific HDR efficiency related to the sequence of the homology arms, gRNA efficiency or the endogenous promoter activity. Of note, the expression of GFP-BATF was induced upon T-cell activation (8.7-fold, Fig. 4d), in line with a previous report 31 , providing an example for an endogenous fluorescent reporter of activation in primary human T cells.
Knock in of eGFP upstream of SAMHD1 allows studies into the physiological interplay of the cellular restriction factor, HIV-1 and an accessory lentiviral protein. To apply the KI approach to a functional analysis in the context of HIV-1 infection in resting CD4 + T cells, we introduced the eGFP gene upstream of the SAMHD1 locus to generate a GFP-SAMHD1 fusion protein (Fig. 5a), which is a Vpx-degradable (Extended Data Fig. 9) and enzymatically active analog of non-tagged SAMHD1. Of note, the new dsDNA template carries the SAMHD1-gRNA2 binding site (Fig. 5a). To avoid cleavage of the donor template by the RNP, the PAM sequence in the dsDNA was mutated and a corresponding plasmid tested for gRNA in vitro digestion. Unexpectedly, mutating the PAM sequence alone decreases the cutting efficiency, but this was not sufficient to fully abrogate in vitro cleavage of the DNA (Fig. 5b, left), in contrast to a recent report 32 . This problem was overcome by mutating the complete gRNA-binding sequence of SAMHD1-gRNA2, while preserving the amino acid sequence (Fig. 5b, right). Five days after nucleofection this KI strategy resulted in 2.3% viable, GFP-positive cells (Extended Data Fig. 10), which were subsequently enriched by cell sorting. The sorted cell population carried the KI cassette in the SAMHD1 locus as verified by a locus-specific PCR amplification specific for the eGFP integration (Fig. 5c). Similarly, immunoblots of the sorted, GFP-positive cell population showed reactivity of the GFP-SAMHD1 fusion protein (∼100 kDa) to both anti-GFP and anti-SAMHD1 antibodies (Fig. 5d). Sorted, GFP-negative cells and plasmid-transfected 293T cells served as references.
For functional analysis, sorted GFP-SAMHD1-expressing resting CD4 + T cells or unmanipulated, sorted WT reference cells were challenged with X4 HIV-1* BFP virions with Vpx (+Vpx) or without and analyzed on day 3 after virus challenge by flow cytometry. As expected 3,4 , both cultures were largely refractory to infection with X4 HIV-1* BFP (Fig. 5e, top quadrants, left, 2.58% and 0.88%) and expression of GFP-SAMHD1 remained intact compared to WT cells (Fig. 5e, left). In contrast, infection with HIV-1* BFP virions containing Vpx was efficient, yielding 20.28% of BFP + CD4 + T cells, the majority of which also showed a strong depletion of GFP-SAMHD1 (Fig. 5e, top quadrants, middle top). Consistent with the model that virion-incorporated Vpx rapidly targets endogenous SAMHD1 for degradation, depletion of GFP-SAMHD1 was readily observed, yet addition of the reverse transcriptase inhibitor efavirenz (EFV) still prevented progression of the replication cycle and productive infection (Fig. 5e, right). Thus, this KI strategy enables the introduction of a functional reporter system into an endogenous locus to study virus-host interactions at single-cell resolution in quiescent T cells.

Discussion
Together, our results establish a set of protocols that overcomes the resistance of resting human CD4 + T cells to genetic manipulation and combined with improved cultivation conditions enables mechanistic studies of this primary cell type. These protocols allow a high KO efficiency with the option of simultaneous multi-gene KOs, enhanced cell viability while preserving a resting state of T cells and the KI of fluorescently modified genes into various endogenous loci to conduct functional characterizations of interest. The RNP nucleofection protocol, which combines two to three pre-tested gRNAs per target gene, in combination with optimized cell culture conditions, resulted in remarkably high editing rates (typically >98%). This approach yielded cell populations, in which protein expression of the specific targets was reduced to undetectable levels. Further characterizations revealed that this reduction in protein expression also translated into a loss of function, demonstrating that this KO protocol facilitates insight into physiological virus-host interactions. Its efficacy and speed alleviate the need for any selection process and allows the analysis of a polyclonal, potentially heterogeneous primary cell population.
Notably, our protocol did not result in CD4 + T-cell activation as assessed by activation marker expression and various markers of cell proliferation and gene-edited cells maintained their general resistance to HIV-1 infection as well as their sensitivity to the lentiviral Vpx protein for infection enhancement. We found that cultivation of resting CD4 + T cells in the presence of low concentrations of human IL-7 and IL-15 preserved cell viability for up to 6 weeks without inducing cell proliferation or expression of activation markers. This provided the basis for efficient gene editing and functional downstream analyses following loss of the respective target proteins of interest. A recent review 33 , discussed the importance of IL-7 and IL-15 for the viability of primary mouse T cells ex vivo. Notably, this cytokine combination induces both T-cell activation and proliferation, referred to as T-cell receptor-independent homeostatic proliferation, in mice, suggesting a species-specific response to these ILs.
In contrast to the commonly used post-activation protocols, cells edited by this protocol maintain the key physiological properties of resting CD4 + T cells. Our approach allowed the simultaneous depletion of up to six individual proteins or the KI of a functional reporter gene. The workflow described herein thus overcomes major limitations of a previously reported protocol for gene editing of resting CD4 + T cells 12 in that it (1) maximizes KO efficiency at enhanced cell viability, (2) preserves cells' resting activation state over weeks of culture and (3) enables simultaneous multi-gene KOs as well as KIs of modified genes. This approach opens avenues not only for the functional characterization of individual gene products in bulk populations of truly resting CD4 + T cells, but also for more complex mechanistic studies by targeting entire cellular pathways, analysis of functional redundancy and compensatory mechanisms as well as introduction of specific mutations, tags and functional reporters.
These protocols enable robust gene editing combined with functional characterization of resting CD4 + T cells. This provides methodology to decipher the role of specific restriction factors, host dependency factors and nucleic acid sensors for determining the balance of resistance and susceptibility of these cells to HIV-1 infection, the establishment of viral latency and why the infection in resting cells is not productive [34][35][36] .
Some limitations toward high-throughput applications remain. First, the specific time point at which the protein of interest is sufficiently depleted for functional analysis needs to be determined for each target gene. Second, and owing to the nonproliferative nature of resting CD4 + T cells, the number of cells per donor that is available for analysis remains somewhat limited. In summary, applying these KO and KI protocols will yield insights into the processes governing infection, latency, reactivation and immune recognition of HIV-1, but also other T-cell tropic viruses such as measles virus,

Fig. 5 | GFP-SAMHD1 endogenously expressed in resting CD4 + T cells is functional in the context of HIV-1 infection and degraded by particle-packaged
Vpx. a, KI-targeting strategy to introduce a N-terminal GFP fusion into the endogenous SAMHD1 locus. The dsDNA template from a was introduced into a plasmid and digested using the SAMHD1-gRNA2-containing RNP. b, Digestion of the dsDNA template with either PAM sequence mutated only (left) or the sequence complementary to SAMHD1-gRNA2 mutated completely, including the PAM sequence (right). No digestion or BstBI digestion (digestion ctrl) served as references. One experiment is shown (n = 2). c, GFP-positive KI resting CD4 + T cells were sorted by flow cytometry and lysed and a PCR specific for the eGFP integration into the SAMHD1 locus was performed. Untreated WT cells and cells nucleofected with dsDNA template only served as references. A PCR specific for the CD46 locus was used as loading control (bottom). One experiment is shown (n = 2). d, Sorted cells from c, either positive or negative for GFP, were immunoblotted for both SAMHD1 and GFP. WT cells served as reference; vinculin was the loading control. The 293T cells transfected with expression plasmids encoding either GFP, GFP-SAMHD1 or SAMHD1-GFP served as references. One representative experiment is shown (n = 2). e, Cells were challenged with equivalent infectious units of HIV-1* BFP virions with (+Vpx) or without the lentiviral SAMHD1 antagonist Vpx and analyzed by flow cytometry on day 3. Reverse transcriptase inhibitor EFV served as a specificity control. One representative experiment is shown (n = 2).
human T-lymphotropic virus, human herpesviruses 6 and 7, human cytomegalovirus and human herpes simplex virus type 2 (ref. 37 ), as well as virus-unrelated studies of activation, proliferation and differentiation of this cardinal human cell type.

Online content
Any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/ s41592-021-01328-8.

Methods
Primary human CD4 + T cells. CD4 + T cells were isolated from heparinized blood retained in leukocyte reduction system chambers from healthy blood donors with approval by the Ethics Committee of the LMU München (project no. 17-202 UE), in principle as reported 38 . In brief, blood cells were diluted with PBS (Gibco) and CD4 + T cells were isolated via EasySep Rosette Human CD4 + T cell enrichment kits (STEMCELL Technologies) according to the manufacturer's protocols. Resting CD4 + T cells were kept in RPMI 1640 GlutaMAX (Gibco) supplemented with 10% (v/v) fetal bovine serum (FBS; Sigma) and penicillin/streptomycin (100 IU ml −1 ; Thermo Fisher Scientific) with the addition of IL-7 and IL-15 (2 ng ml −1 each; Peprotech) and cultivated in 96-well flat-bottom plates at a cell density of 1 × 10 6 cells ml −1 . The culture medium was replaced every 3 d. For activation, Dynabeads Human T-Activator CD3/CD28 (Gibco) were added to CD4 + T cells at a ratio of 1:10 (bead to T cell) and kept in medium containing IL-2 (50 IU ml −1 ; Biomol). Fresh beads were added to the culture every other week. Here, only 2 µl of RNP complexes for each gRNAs were used. For co-editing of up to six genes, only 0.5 µl of each RNP complex were used. Cells were nucleofected using program EH-100 on the 4D-Nucleofector system 12 . Then, 100 µl of pre-warmed RPMI (without supplements) was added to each well and cells were transferred to 48-well plates and allowed to recover for 10 min at 37 °C. Subsequently, complete culture medium supplemented with IL-7 (Peprotech; 200-07) and IL-15 (Peprotech; 200-15) (2 ng ml −1 each) was added. A list of gRNA sequences used in this study is provided in Extended Data Table 1. The PCR was then cleaned up using the NucleoSpin Gel and PCR clean-up columns (Macherey-Nagel), followed by Sanger sequencing that was performed by Eurofins. Sequencing results were analyzed with the TIDE webtool (http://tide.nki.nl/) 39 . For Miseq, 1 µl of lysate was used to perform PCR-I and subsequently PCR-II followed by Illumina MiSeq analysis as described previously 40 .
For the analysis of off-target sites, cells were collected 1 week after nucleofection and lysed. The top two off-target sites in open reading frames were selected using the Synthego CRISPR design tool (https://design.synthego.com). A PCR specific for each off-target site was performed and analyzed by Illumina Miseq. Primers used are listed in Extended Data Table 1.
Immunoblotting. CD4 + T cells were collected and washed twice with PBS. Cell pellets were resuspended in RIPA buffer supplemented with proteinase inhibitors (Roche) and phosphatase inhibitors (Thermo Fisher Scientific) and kept on ice for 30 min followed by freezing at −80 °C. Cell lysates were thawed on the day of the experiment, spun clear at 10,000g for 10 min at 4 °C and lysates were transferred to new tubes. The protein concentration was quantified by BCA assay (Thermo Fisher Scientific), following the manufacturer's protocol. Lysate samples were separated by tris-glycine denaturing SDS-PAGE (Thermo Fisher Scientific). Proteins were blotted onto 0.2-mm nitrocellulose membranes (GE Healthcare), blocked in 5% milk (Roth) in TBS-T for 1 h and incubated with the indicated primary antibody in 1% BSA/TBS-T or 5% milk, depending on the antibody used and subsequently with the corresponding secondary antibodies for 1 h (1:10,000 dilution in 5% milk). ECL (Bio-Rad) was used as substrate and the chemiluminescent signals were detected on a Fusion Fx (Vilber). The following human-specific antibodies were used: anti-SAMHD1 (proprietary chicken monoclonal antibody of the Keppler laboratory), anti-CPSF6 (rabbit, polyclonal, Cell Signaling, cat. no. 75168S) 1:1,000 dilution, anti-vinculin (mouse, hVIN-1, Sigma Aldrich, cat. no. V9264) 1:2,000 dilution, anti-MX2 (rabbit, polyclonal, Novus Biologicals, cat. no. NBP1-81018) 1:250 dilution, anti-TRIM5α (rabbit, clone D6Z8L, Cell Signaling, cat. no. 14326S) 1:1,000 dilution, anti-SAMHD1 (mouse, clone OTI3F5, Origene, cat. no. TA502024) 1:250 dilution and anti-GFP (rabbit polyclonal, Chromoteck, PABG1-20) 1:1,000 dilution. The following HRP-conjugated secondary antibodies were used in a dilution of 1:10,000: goat anti-mouse IgG (rat adsorbed, polyclonal, Bio-Rad, cat. no. STAR77), goat anti-chicken IgY (H&L, polyclonal, Abcam, cat. no. ab6877) and goat-IgG anti-Rabbit IgG (H + L, polyclonal, Dianova, cat. no. AFK-600). In six-gene KO cells (Fig. 2b) TRIM5α and CPSF6 expression was analyzed by consecutive re-probing of nitrocellulose membranes, followed by staining for vinculin (loading control). For immunoblots of GFP-SAMHD1 KI CD4 + T cells, due to the limited number of cells available, cells were activated first with Dynabeads Human T-Activator CD3/CD28 and IL-2 for 1 week to allow expansion, subsequently collected and sorted with the FACSAria Fusion cell sorter (BD) and lysed as described above. Gating strategies are described in Extended Data Fig. 1d and Supplementary Data. Full-length blots are provided as Source Data.
WES system western blot. Cells were lysed and the protein concentration was quantified as described above. Then, 0.6 µg of total protein was evaluated by separation and immunodetection employing HIV-1 plasmids. For HIV infection, the HIV-1 GFP proviral clone NLENG1-IRES 42 was used, referred to as HIV-1 GFP in the current study. The Vpx-binding motif DPAVDLL from SIVmac Gag was introduced into the p6 of NLENG1-IRES (referred to as HIV-1* GFP in this study; GenBank accession no. OK558601), obtained from the plasmid HIV-1* NL4-3 (ref. 3 ) using the restriction sites SphI and AgeI. For packaging of Vpx into virions, the Vpx expression construct pcDNA3.1 Vpx SIVmac239-Myc was used and the pcDNA3.1 empty vector served as negative control 3 . GFP was replaced by mtagBFP to obtain HIV-1* BFP (GenBank accession no. OK558602). An insert obtained from pCDH-mtagBFP vector 43 was introduced into the HIV-1* GFP using the restriction sites SphI and AgeI. For the virion fusion assay, the R5 HIV-1 proviral clone HIVivo 44 , kindly provided by M. Nussenzweig (Laboratory of Molecular Immunology, The Rockefeller University, New York, NY, USA), was used in combination with pCMV-BlaM-Vpr during virus production (see below). X4 HIVivo (GenBank accession no. OK589863) was generated introducing the X4 envelope gene from NLENG1-IRES into the R5 HIVivo backbone using the restriction sites EcoRI and HpaI. Snapgene was used to design the cloning strategy and the primers needed.
HIV-1 production. Sucrose-cushion-purified HIV-1 stocks were produced as previously described 45 . HIV-1* GFP virus stocks, carrying virion-packaged Vpx, were produced by co-transfection of the proviral HIV-1* GFP DNA and the indicated Vpx expression constructs 7 . In brief, 293 T cells were seeded at a density of 8 × 10 6 cells in a 15-cm dish. After 24 h, cells were co-transfected with a mixture of 37.5 µg HIV-1 plasmid and 112.5 µl of L-PEI (3 µl of L-PEI for every µg of DNA, stock concentration of 1 µg µl −1 ; Polysciences) in DMEM without any additives for 30 min. After this time, the DNA/PEI solution was added to the cells. After 72 h, the supernatant was collected and virus was purified via sucrose-cushion centrifugation. For virions to incorporate Vpx, the transfection was performed as described above, but using 37.5 µg of HIV-1* GFP or HIV-1* BFP together with 18.75 µg of pcDNA-Vpx (SIVmac239-Myc) or the corresponding pcDNA3.1 empty control vector. For virus production for the virion fusion assay, the transfection was performed as described above, combining 37.5 µg of X4 HIVivo and 12.5 µg of pCMV-BlaM-Vpr.
HIV-1 fusion assay. T cells were incubated with virions containing BlaM-Vpr at 37 °C for 4 h. Subsequently, cells were washed twice in CO 2 -independent medium (Thermo Fisher Scientific) and then loaded with CCF2/AM dye (Thermo Fisher Scientific), as described previously 21,22 . Briefly, 2 µl of CCF2/AM (1 mM) was mixed with 8 µl of solution B and 10 µl of probenecid (250 mM stock; MP Biomedicals) in 1 ml of CO 2 -independent medium supplemented with 10% FBS (v/v). Cells were incubated in 100 µl of loading solution for 16 h at room temperature. Cells were then washed twice with PBS and fixed with 4% (v/v) paraformaldehyde (PFA) for 1.5 h. Subsequently, cells were washed and resuspended in FACS buffer. The shift in emission fluorescence of CCF2 after cleavage was monitored by flow cytometry.
HIV-1 infection. The titer of individual virus stocks was determined on SUP-T1 cells using virus-encoded GFP or BFP signals measured by flow cytometry as readout for productive infection. Primary resting CD4 + T cells were infected with virus stocks at different MOIs as indicated for each experiment. Where indicated, infections were performed either by co-incubation of virus and cells without additional centrifugation or by spinoculation for 2.5 h at 650g and 37 °C. After 3 d, cells were washed twice with PBS and fixed with 4% (v/v) PFA for 1.5 h. Cells were then washed and resuspended in FACS buffer. The percentage of GFP-or BFP-positive cells was monitored by flow cytometry. Drug or antibody controls were added to cells 30 min before HIV-1 challenge. The following drugs were used: EFV (stock, 10 mM; Sigma Aldrich), AMD3100 (stock, 16 µg ml −1 ; Sigma Aldrich), anti-CD4 clone SK3 (stock, 25 µg ml −1 ; (cat. no. 344602) BioLegend) and T20 (stock, 90 mg ml −1 ; Enfuvirtid; Roche). For infection of GFP-SAMHD1 KI CD4 + T cells, GFP-positive resting CD4 + T cells, 1 week after nucleofection, were sorted using a FACSAria Fusion cell sorter (BD), allowed to rest for 16 h and then challenged with virus.
Chemokine-migration assay. Resting CD4 + T cells were used 1 week after nucleofection for this assay (CXCR4 KO or NTC). After removing the membrane from the Transwell system, 500 µl of RPMI medium supplemented with 0.2 % of FCS and the chemokine SDF-1α (1,000 ng ml −1 ; Peprotech) were added at the bottom of the 6.5-mm Transwell with 3.0-µm pore (24-well plate; Corning). The polycarbonate membrane was added into the corresponding wells and for each condition, 200 µl containing 2.5 × 10 5 cells were transferred to the top of the membrane. The 24-well plate was incubated at 37 °C for 3 h. Subsequently, the membrane was removed and the total number of cells in the bottom chamber of the Transwell was quantified by flow cytometry using BD Trucount Absolute Counting Tubes.

Measles virus infection.
CD46 KO and WT reference resting CD4 + T cells were challenged with a measles GFP reporter virus (MeV-vac-eGFP), Schwarz-ATU-GFP 46 , at an MOI of 1. After 24 h, cells were stained with an APC-conjugated antibody to detect CD46 surface expression (clone TRA-2-10; BioLegend) and analyzed by flow cytometry. The MeV-vac-eGFP stock was generated as described 46 .
Production of knock in DNA templates. The plasmids containing the donor DNA template for each KI approach were synthesized by Twist Bioscience (pTwist KI-Template GFP; GenBank accession no. OK558599) and pTwist KI-Template GFP-SAMHD1 (GenBank accession no. OK558600). The DNA template was amplified by PCR from these plasmids using specific primers. The PCR reaction contained 5 µl 5× High-fidelity PCR buffer (Thermo Fisher Scientific), 5 µl 5× GC PCR buffer (Thermo Fisher Scientific), 1 µl dNTPs (10 mM stock; Thermo Fisher Scientific), 1.5 µl dimethylsulfoxide, 2.5 µl forward primer (10 µM stock), 2.5 µl reverse primer (10 µM stock), 1 µl Phusion (NEB), 1 µl (10 ng For ssDNA production, a PCR product containing a phosphate group on one of the two strands is required (see below). For this reason, additionally, either the forward primer or the reverse primer were replaced with a primer of the same sequence but with an additional phosphorylation. The ssDNA was produced with the Guide-it Long ssDNA Production System (Takara Bio) according to the manufacturer's protocol. After the PCR, a PCR clean-up was performed with the NucleoSpin Gel and PCR clean-up (Macherey-Nagel) according to the manufacturer's protocol. Finally, the DNA concentration was determined by NanoDrop (Thermo Fisher Scientific). Sequences of KI templates used are listed in Extended Data Table  1. To knock in eGFP into the SAMHD1 locus (Fig. 4a), a dsDNA template with homology arms of around 550 bp each, including an eGFP reporter gene followed by a stop codon and a polyadenylation (polyA) signal was used to disrupt endogenous SAMHD1 expression.
In vitro digestion of knock in DNA templates. To test whether the KI dsDNA templates are cleaved by the gRNAs used to generate the double-strand break in the target cell genome, an in vitro restriction was performed. The Cas9-RNP complex (1 µM) as well as a single-cutter restriction enzyme (BstBI) for the specific plasmid were used and samples were incubated at 37 °C for 2 h. Subsequently, 1 µl of proteinase K (20 mg ml −1 ) was added and samples were incubated at 56 °C for 10 min, separated and visualized on an agarose gel (1 %).
Knock in of resting CD4 + T cells. For KIs, the same nucleofection conditions as for the KO generation were used (P3 buffer and program EH-100; Lonza). In addition to the RNP, the donor DNA template was added to the P3 cell suspension at 1 µg, unless stated otherwise. Additional information on the overall strategies to generate KIs into different loci in resting CD4 + T cells, including DNA templates, gRNAs and primers is included in Extended Data Table 1.

Confocal microscopy.
Resting CD4 + T cells were used 2 weeks after nucleofection (KO for CD46, CD4, PSGL-1 or CXCR4). Cells were stained with antibodies against either CD46 (APC clone, TRA-2-10; BioLegend), CXCR4 (APC clone, 12G5; BD), PSGL-1 (Alexa Fluor 647, KPL-1; BD) or CD4 (APC clone, RPA-T4; BD). Cells were collected, washed once with PBS and resuspended in 50 µl staining solution (FACS buffer and specific antibodies) and kept for 20 min at 4 °C. To amplify the signal for CD46 a secondary antibody was used (goat anti-mouse IgG (H + L), Alexa Fluor 647; (cat. no. A-21236), Invitrogen). After this time, cells were washed, fixed with 4% PFA/PBS for 10 min at room temperature and washed again. Cell were then mounted with ProLong Diamond Antifade Mountant (Thermo Fisher Scientific) and analyzed with a spinning disk confocal microscope (Nikon). For the KI experiments, activated KI cells were used 2 weeks after nucleofection. Cells were washed, as described above and fixed with BD Cytofix for 10 min at room temperature. After washing, cells were permeabilized with Perm Buffer III (BD) for 10 min on ice. Cells were then washed twice with Perm/Wash buffer (BD) and resuspended in Perm/Wash buffer containing the primary antibody (GFP clone, PABG1; Chromoteck) for 30 min on ice. Cells were then washed twice with Perm/Wash buffer and resuspended in Perm/Wash buffer containing the secondary antibody (anti-rabbit IgG (H + L) and Alexa Fluor 647 (Invitrogen, cat. no. A-21236)) for 30 min on ice. Cells were then washed twice and mounted with ProLong Diamond Antifade Mountant and analyzed by spinning-disk confocal microscopy. Imaris Viewer (Oxford Instruments) was used to analyze images.
Material availability. All materials are available upon request to keppler@mvp. lmu.de. This includes chicken anti-human SAMHD1 monoclonal antibody and proviral constructs pHIV-1* GFP, pHIV-1* BFP and pX4 HIVivo. These proviruses will also be made available through the National Institutes of Health AIDS Reagent Program. pTwist KI-Template GFP and pTwist KI-Template GFP-SAMHD1 are available from Addgene (plasmids 177988 and 177987, respectively).

Reporting Summary. Further information is available in the Nature Research
Reporting Summary linked to this article.

Data availability
The data in this paper are shown in the main figures and Extended Data figures. Additional information is available as Source Data Files for Figs. 1-4