IL-2 Inducible Kinase ITK is Critical for HIV-1 Infection of Jurkat T-cells

Successful replication of Human immunodeficiency virus (HIV)-1 depends on the expression of various cellular host factors, such as the interleukin-2 inducible T-cell kinase (ITK), a member of the protein family of TEC-tyrosine kinases. ITK is selectively expressed in T-cells and coordinates signaling pathways downstream of the T-cell receptor and chemokine receptors, including PLC-1 activation, Ca2+-release, transcription factor mobilization, and actin rearrangements. The exact role of ITK during HIV-1 infection is still unknown. We analyzed the function of ITK during HIV-1 replication and showed that attachment, fusion of virions with the cell membrane and entry into Jurkat T-cells was inhibited when ITK was knocked down. In contrast, reverse transcription and provirus expression were not affected by ITK deficiency. Inhibited ITK expression did not affect the CXCR4 receptor on the cell surface, whereas CD4 and LFA-1 integrin levels were slightly enhanced in ITK knockdown cells and heparan sulfate (HS) expression was completely abolished in ITK depleted T-cells. However, neither HS expression nor other attachment factors could explain the impaired HIV-1 binding to ITK-deficient cells, which suggests that a more complex cellular process is influenced by ITK or that not yet discovered molecules contribute to restriction of HIV-1 binding and entry.

. Characterization of ITK knockdown cells. (A) Jurkat cells expressing shRNAs targeting ITK or non-target (n.t.) shRNA were assayed for ITK expression via immunoblots using an ITK-specific antibody. Immunoblots were co-probed using anti-GAPDH antibody to show equal sample concentrations. (B) Expression of HIV-1 receptors CD4 and CXCR4 was analyzed in knockdown and wild-type cells by FACS analysis. Filled histograms represent isotype control and open histograms staining of CD4 or CXCR4 receptor. (C) Amount of total CDC42 and RAC1 was assayed by immunoblots using CDC42 and RAC1 specific antibodies. Immunoblots were co-probed using anti-GAPDH antibody to show equal sample concentrations. (D) Amount of activated form of CDC42 and RAC1 was assayed by pull-downs (input proteins shown in Fig. 1A and C), followed by immunoblot detection using CDC42 and RAC1 specific antibodies. Full-length blots are presented in Supplementary Fig. 1 of HIV-1 spread, Jurkat cells with the shRNA 258 showed a delayed and reduced virus production. Together, these results support the notion that ITK regulates multiple steps, early and late, during the HIV-1 infection, as described before [7][8][9] .
Single-round HIV-1-based luciferase viruses then were used to identify which stage of the viral life cycle was affected by ITK. Infection assays with HIV-1 enveloped particles (from clone L102, a variant of the C-terminally truncated (at amino acid 712) HIV-1 Env strain BH10) showed a strong dependency on ITK expression. HIV-reporter gene activity was highly reduced in the ITK knockdown cells compared to Jurkat cells that expressed ITK. However, transduction of Jurkat cells with VSV-G enveloped HIV reporter viruses resulted in luciferase activity independent of the ITK expression (Fig. 2B). We also tested the ITK inhibitor BIX02524 28 on HIV-1 infection using the luciferase reporter viruses (Fig. 2C). Jurkat cells treated with the ITK inhibitor showed a concentration-dependent reduced susceptibility for HIV-1 transductions that were mediated by the HIV envelope protein, but infections were not repressed when VSV-G pseudotypes were tested. These findings indicate that the ITK deficiency caused a specific HIV-1 restriction that involves the viral fusion and entry mechanism and excludes early downstream steps such as reverse transcription, integration, and protein expression.

Fusion of HIV-1 with the host cell membrane is inhibited in ITK knockdown cells. Because our
experiments suggested an ITK-mediated defect in viral entry, virion-based fusion assays were performed as described previously 29,30 . Viral particles were generated with incorporated ß-lactamase-Vpr chimeric proteins. After successful fusion with the host cell membrane, ß-lactamase protein was delivered to the cytoplasm where it cleaved a fluorescence substrate (CCF2). Cleavage of the dye caused a change in the emission spectrum from green (520 nm) to blue (447 nm). The viral fusion rate was calculated by flow cytometry based on the green to blue ratio of the cell population. For wild-type Jurkat and non-target shRNA expressing cells, the fusion rates were 6.7% and 6.04% respectively, when HIV-1 particles with HIV envelope were applied. However, fusion rates of the same virus stock were only 0.38-0.41% with the ITK deficient cells (Fig. 3A). VSV-G pseudotyped viruses that were used as the positive control showed equally high rates of viral fusion for wild-type, non-target shRNA expressing cells and ITK-deficient cells (Fig. 3B).

Attachment of viral particles to ITK knockdown cells is impaired in a gp120 independent manner.
Because ITK expression influenced the HIV-1 fusion process, we analyzed the interaction of the gp120 envelope protein of HIV-1 with the host cells. Biotinylated gp120 protein was incubated with ITK expressing and ITK knockdown Jurkat cells. After 1 h of incubation, unbound protein was removed and cell-associated gp120 was visualized by staining with APC-Cy-7-coupled avidin and detected by flow cytometry. Compared to the negative control (cells incubated with biotinylated protein lysate without gp120), approximately 50% (34-66%) of the cells were positive for APC staining. However, no difference in gp120 binding was detected between ITK expressing and non-expressing cells (Fig. 4A).
We then analyzed the attachment of viral particles to target cells. Fluorescent viral particles were produced using viral vectors expressing a GFP-fused Gag polyprotein 31 . These GFP protein containing viruses were incubated with wild-type and ITK deficient cells. After removing unbound virus, cells were fixed with paraformaldehyde and fluorescence was analyzed by flow cytometry. Green HIV-1 particles with HIV-derived envelope protein were detected on the cell surface of ITK expressing wild-type and non-target shRNA cells, as demonstrated by a 22% and 23% increase in GFP fluorescence, respectively. In contrast, viral association with ITK knockdown cells was reduced, and the virus-associated fluorescence of ITK knockdown cells only reached 3% (Fig. 4B). Replacing the HIV-1 glycoprotein with the VSV-G protein resulted in green virions that strongly bound to all Jurkat cells irrespective of their ITK expression (Fig. 4C). Finally, we tested green virus particles that were produced without any viral glycoprotein. These GFP particles weakly bound to ITK knockdown cells, producing a fluorescence of 0.8% and 2.6%, whereas the ITK expressing cells associated strongly with these virions, showing a fluorescence of 19.8% and 18.5% (Fig. 4D).

Cell surface expression of heparan sulfate (HS) is reduced in ITK deficient cells.
Our data indicated that ITK knockdown cells have a very early block to HIV-1 infection due to processes affecting viral binding and membrane fusion. Because the expression of HIV receptors CD4 and CXCR4 were comparable in wild-type and ITK knockdown cells, other attachment factors may be affected by ITK. A potential alternative attachment protein is Siglec-1 (CD169) 32 . However, using flow cytometry analysis we found that neither wild-type nor ITK knockdown Jurkat cells expressed Siglec-1 on the cell surface, although it was detected on lipopolysaccharide treated monocyte-derived dendritic cells (data not shown). The presence of lymphocyte function-associated antigen 1 (LFA-1) can increase HIV-1 infection by improving viral attachment 33,34 , and ITK was shown to be involved in regulation of integrins such as LFA-1 20,35 . Therefore, cell surface expression of LFA-1 was analyzed using an anti-CD18 antibody. We also quantified the main interaction partner of LFA-1, ICAM-1, which was reported to enhance the infectivity of HIV-1 by incorporation into viral particles 36,37 . Although the amount of ICAM-1 was only slightly higher, LFA-1 expression was significantly enhanced in ITK knockdown cells compared to wild-type cells (Fig. 5). To test whether LFA-1 is important for HIV-1 infections, Jurkat cells lacking LFA-1 expression were generated using CRISPR-Cas9 38 ( Supplementary Fig. 3A). Single-round reporter virus infections in wild-type and LFA-1 knock-out cells showed that the LFA-1 expression did not change the susceptibility to HIV-1 ( Supplementary Fig. 3B). Similar, LFA-1 expression did not modulate the binding of HIV-1 particles to Jurkat cells ( Supplementary Fig. 3C). However, by analyzing the spreading replication in both Jurkat cell lines during 18 days using the TZM-bl reporter cells as read-out, LFA-1 wild-type cells showed a higher virus replication compared to the LFA-1 knock-out cells ( Supplementary Fig. 3D). LFA-1 deficiency did not impact the cell proliferation as measured with Carboxyfluorescein succinimidyl ester dye (CFSE) fluorescence for 96 hours by flow cytometry (Supplementary Fig. 3E). Thus, we conclude, that absence of ITK may have caused up-regulation of surface LFA-1, which is, however, unlikely to restrict entry of HIV-1.
We also analyzed the presence of HS, a linear sulfated glycosaminoglycan that has been implicated being important for attachment of HIV-1 (for review 39 ). HS usually is part of a proteoglycan (HSPG), consisting of a cell surface protein with one or more HS chains covalently attached 40 . Only a small number of HSPGs (<20) have been identified, whereas hundreds of proteins have the capacity to interact with HS (for review 41 ). We measured the HS level on the cell surface of our Jurkat cells with and without ITK expression by flow cytometry using an antibody reacting with the 10E epitope. That epitope is present on many types of HS including N-acetylated and N-sulfated glucosamine residues. Much to our surprise, ITK deficient cells lacked most of the HS that was detected on ITK expressing cells. While wild-type cells showed moderate expression, HS was not detectable on knockdown cells (Fig. 6A). To investigate the importance of HS for HIV-1 attachment, HS was enzymatically removed by a mixture of heparinase I and III. Jurkat wild-type cells, which positively stained for HS expression, had a significantly reduced amount of HS molecules on their cell surface after the enzymatic treatment (Fig. 6B). However, the susceptibility to HIV-1 attachment of HS-stripped cells did not differ from that of not digested cells (Fig. 6C). To understand if the lack of HS in the ITK deficient cells affected also the presence of HSPGs, we tested CD47 expression. CD47 is one of the highly expressed proteins in Jurkat cells that is modified by HS 42 . We detected CD47 on all four types of Jurkat cells, with moderately reduced surface levels on ITK knockdown cells (Fig. 6D). Together with previous findings that CD47 depletion did not reduce HIV-1 infections 43 , we conclude that the light reduction of cell surface CD47 expression is unlikely causing the replication block we detected in ITK deficient cells.

Discussion
The T-cell specific TEC kinase ITK is involved in regulation of various processes, including actin reorganization, Ca 2+ mobilization, and control of transcription factors. Therefore, ITK plays an important role during cell differentiation and activation (for review see 44 ). Because proper T-cell activation is required for HIV-1 infection, a function for ITK as a host factor for HIV-1 replication was expected, and recent studies confirmed a block of HIV-1 at various steps of the viral life cycle after ITK inhibition [7][8][9] . Here we show that down-regulation of ITK expression via RNA interference (RNAi) led to severely reduced HIV-1 expression and spreading in human Jurkat T-cells as well as reduced membrane fusion of HIV-1 caused by an impaired virus particle attachment. Consistent with Readinger et al. 7 , our experiments confirm that ITK affects viral entry, but not reverse transcription.
The lower capacity for virus binding might be explained by defects in actin reorganization and receptor clustering, as ITK was shown to be important for control of the actin cytoskeleton and local enrichment of adapter molecules as well as gp120-induced cytoskeleton reorganization 7,22 . In our experiments the typical F-actin depolymerization in Jurkat cells incubated with HIV-1 was detectable in wild-type cells but not in ITK knockdown cells (data not shown). This observation is in agreement with the reduced expression of active RHO GTPases in ITK knockdown cells (Fig. 1C), but the absence of attached viral particles likely contributed to the missing actin depolymerization. In contrast, HIV-1 particles that were pseudotyped with VSV-G were able to undergo membrane fusion and infection in ITK-deficient cells. These observations indicate that the step of viral attachment and fusion is the impaired function in ITK knockdown cells. Although VSV entry also requires actin reorganization mediated by RHO GTPase signaling, the route of entry and the involved molecules differ. HIV-1 enters cells in a pH independent way at the plasma membrane; whereas VSV-G mediates pH dependent fusion via endocytosis in clathrin coated vesicles 45 . Thus, for HIV-1 entry the RHO GTPases RAC1, CDC42 and RHOA play a dominant role 6,46-48 , whereas RHOB and RHOC seem to be more important for VSV entry 49 . Our data indicate further that ITK-deficiency in Jurkat cells diminishes a late stage in HIV replication affecting the viral infectivity. In our study, we did not explore this late stage block in detail because we focused on HIV attachment and entry. However, other studies found that after HIV-1 infection ITK deficiency induces a late stage block of replication including virion assembly and release 7,9 . In addition, it was shown that the HIV-1 protein Nef binds and activates ITK, suggesting that Nef may play a role in the recruitment and activation of ITK, and thus contributes to viral egress 8 . The expression levels of the HIV-1 receptors CD4 and CXCR4 in wild-type and ITK knockdown cells were similar, with slightly higher CD4 levels in the ITK deficient cells. We conclude that the attachment of HIV is not much affected by gp120 binding to CD4/CXCR4, which appears to be needed only to induce membrane fusion. By analyzing alternative attachment factors on Jurkat cells, we found that ITK knockdown cells expressed higher levels of the integrin LFA-1, slightly reduced levels of CD47, and no HS compared to wild-type cells. The membrane glycoprotein LFA-1, an integrin which is composed of the integrin alpha L chain ITGAL (CD11) and the beta 2 chain ITGB2 (CD18) is expressed on T-cells, were it functions in recruitment of the cells to the site of infection and interacts with antigen-presenting cells 50 . For binding to one of the interaction partners ICAM-1 (CD54), ICAM-2 (CD102) or ICAM-3 (CD50), LFA-1 needs to be activated first. During production of HIV-1 cellular ICAM-1 is selectively packaged into new particles by interaction with the viral matrix protein 36 . In addition, elevated attachment and entry of "cell-free" HIV-1 to CD4 T-cells is mediated by incorporated ICAM-1 37,51-54 . Since LFA-1 and ICAM-1 are also integral components of the HIV-1 virological synapse 33,55 , a supramolecular structure which mediates efficient viral transmission between infected and uninfected T-cells 33 , it is speculated that interactions of this adhesion molecules are not only important for infection with free virus particles but also for viral cell to cell transmission. We saw a moderate effect of the LFA-1 deficiency on spreading replication of HIV-1, consistent with the previous observation that T-cells deficient in LFA-1 were less able to support cell-cell transfer of HIV-1 33 . However, our experiments suggest that LFA-1 expression in Jurkat cells is not important for HIV-1 binding and entry. CD47 (also known as integrin-associated protein) is a ubiquitously expressed glycoprotein of the immunoglobulin superfamily that plays a critical role in self-recognition. CD47 was not investigated further in this study, because we found the down-regulation rather moderate on ITK-deficient cells and a previous study demonstrated that CD47-deficiency does not affect HIV-1 replication 43 .
The anti-HS antibody used in our study recognized the 10E4 epitope that is present in many types of HS. Thus, it is likely that ITK knockdown cells do not express HS on the cell surface. Alternatively, ITK knockdown cells could express HS variants that do not react with the anti-HS antibody. HS moieties on cell surfaces are known to provide attachment sites for many viruses, including foamy retroviruses 56,57 . In line with the importance of HS for foamy virus infection, infection of ITK knockdown cells by prototype foamy viruses revealed a 10-fold restriction compared to infection of ITK-expressing Jurkat cells (data not shown). Studies have also shown that HS is important for attachment of HIV-1 to target cells 39,[58][59][60][61][62][63] . Most of these studies concluded that HS interacts with the viral envelope protein; however, other reports show that attachment of envelope protein-free HIV particles to cells is caused at least in part by HS 64,65 . In these studies, treatment of HeLa cells with heparinase I reduced binding of envelope protein free HIV particles by 25%, and the highest heparinase concentrations reduced binding by 50% 64 . We found that HIV-1 particles without a viral glycoprotein bound to Jurkat wild-type cells but not to ITK knockdown cells and that the enzymatic removal of HS from the cell surface of Jurkat cells by a mixture of heparinase I and III did not reduce this binding. These data indicate that Jurkat T-cells contain an HS-independent attachment factor that does interact with the viral membrane. This factor, however, has yet to be identified. The striking correlation between very low HS expression in ITK knockdown cells and the absence of attached HIV-1 particles may point to a more complex mechanism that cannot be completely described by cell surface staining using the anti-HS antibody. Finally, binding of the HIV particles in Jurkat T-cells might be regulated by a HS-modified protein, and the observed lack of HS may be the result of down-regulation of such a protein.
It appears that this unknown factor mediates HIV-1 attachment in a gp120 Env-independent way. We hypothesize that HIV with viral glycoproteins that weakly bind to receptors, e.g., HIV-1 Env BH10, need this additional attachment for infection of Jurkat cells and infection of ITK-deficient Jurkat cells may need viral glycoproteins that have a stronger interaction with receptors, e.g., VSV-G. ITK is not expressed in HEK293T or HeLa cells; however, if these cells are genetically modified to express CD4 and CXCR4, they are fully permissive for HIV-1 infection (data not shown). In addition, we found ITK expression in other HIV-1 permissive T-cell lines to be absent or much lower than in Jurkat cells (data not shown). These observations suggest that ITK is not strictly required for HIV-1 infection and that the ITK-dependence of HIV-1 may reflect an entry-associated pathway that is present in Jurkat cells and maybe different in other cell lines. Altogether, our results support the premise that ITK is an important protein that modulates the permissivity of Jurkat T-cells for HIV-1 infection that involves an unknown mechanism for HIV-1 attachment. Small hairpin RNA. Small hairpin RNA (shRNA)-mediated ITK knockdown was carried out using a set of lentivirus particles expressing different shRNAs against ITK (MISSION shRNA cat. no. SHGLY-NM_005546; Sigma), or non-target shRNAs (cat. no. SHC002; Sigma). Lentivirus particles were produced by transfecting HEK293T cells with 700 ng shRNA expression plasmid, 280 ng of pMDLg/pRRE 66 packaging construct, 110 ng pRSV-Rev 66 and 60 ng of VSV-G expression plasmid. Jurkat cells were spin-transduced via centrifugation for 1 h at 31 °C at 2000 rpm by lentivirus particles expressing shRNA. Two days after transduction cells were cultured in the presence of 2 μg/ml puromycin (Applichem, Darmstadt, Germany) for a two-week selection period.

Methods
Calcium measurements. Calcium flux analyses of human T-cells were performed according to 67 . Briefly, 3 × 10 5 cells/mL in phenol red-free RPMI 1640 (Life Technologies) containing 10% FCS were loaded with 5 µg/mL Indo-1 AM (MoBiTec, Göttingen, Germany) at 37 °C for 45 min, followed by an additional incubation for 45 min in medium without Indo-1. Cells were kept on ice before equilibration at 37 °C for 5 min, directly before measurement. Changes in intracellular calcium were monitored using a flow cytometer LSRI (BD Biosciences, Heidelberg, Germany). Cells were illuminated using the 325 nm laser-line of a helium-cadmium laser. Fluorescence emissions at 405/30 nm (calcium-bound Indo) and 510/20 nm (free Indo) were detected simultaneously, analyzing the ratio of bound to free Indo over time. After monitoring the baseline activity for 1 min, the cells were stimulated by 10 µg/mL CD3 mAb (clone UCHT1, BD Biosciences) and cells were measured for another 6 min. To confirm proper Indo-1 loading, the cells were then treated with 10 µg/mL ionomycin (Sigma-Aldrich). Kinetics were analyzed using FlowJo v7.6.3 software (Tree Star, Ashland, USA). q-PCR. Expression levels of ITK transcripts in Jurkat cells stably expressing sh ITK RNA were quantified by q-PCR using the Applied Biosystems 7500 Real-Time PCR system (Applied Biosystems, California, United States). Total RNA was isolated from cells via the RNeasy Mini kit (Qiagen, Hilden, Germany) and 1 µg of total RNA was used to produce cDNA by Quantitect Reverse Transcription Kit (Qiagen). Q-PCR-reactions with cDNA were performed in triplicates using the SYBR green master mix (Applied Biosystems) according to the manufacturer's instructions. ITK was amplified with the primers ITK_Exon1_F1 (5′-TGAACAACTTTATCCTCCTGGAAGA) and ITK_Exon1_R2 (5′-GGTTAACACAAAGAAGCGGACTTTA). After initial incubations at 50 °C for 2 min and 95 °C for 10 min, 40 cycles of amplification were carried out for 15 s at 95 °C, followed by 1 min at 60 °C. ITK levels were normalized to the human reference gene hypoxanthine-guanine phosphoribosyltransferase1 (primers HPRT1_fw 5′-GCTTTCCTTGGTCAGGCAGT and HPRT1_rv 5′GCTTGCGACCTTGACCATCT).
GST pull-down. PAK1-RHO binding domain (RBD) was expressed in E. coli as fused GST protein in the pGEX-2TK plasmid (GE Healthcare, Munich, Germany). The total bacterial lysate was prepared using standard protocols and the quality was checked in SDS-Gels 68,69 . Bacterial lysates containing the PAK1-RBD were used as a bait and subsequently to pull-down activated GTP-bound CDC42 or RAC1 from total Jurkat cell lysates (pray) including the wild-type, scramble and ITK knockdown cells 70  done using the software FlowJo 7.6.3 (Tree Star, Ashland, USA)). Cells were grown under normal culture conditions and fluorescence of CFSE was measured in a time interval of 24 hours, beginning directly after staining and ending after 96 hours. As negative control, cells were incubated 1 h before CFSE staining with 100 µg/ml Cisplatin (Accord Healthcare, Freilassing, Germany), a cytostatic drug which inhibits cell growth. CRISPR/Cas9. For LFA-1 knock-out, Cas9 and sgRNA (TGCCCGACTGGCACTGATAGAGG) targeting ITGAL (pLenti-ITGAL) were delivered into Jurkat cells by lentiCRISPRv2-based lentiviral transduction 38 . Lentiviral particles were produced by transfecting HEK293T cells with 800 ng pLenti-ITGAL, 800 ng packaging construct psPAX2 (Trono laboratory, obtained from Addgene, Cambridge, USA) and 300 ng VSV-G expression plasmid. Jurkat cells were spin-transduced via centrifugation for 90 min at 1200 rpm at room temperature. Three days after transduction medium was replaced with medium containing 2 μg/ml puromycin (Applichem, Darmstadt, Germany). ß-lactamase-based virion fusion assays. ß-lactamase containing viral particles where generated by transfecting 293 T cells with 3rd generation lentiviral plasmids; pMDLg/pRRE and pRSV-Rev for packaging, pSIN PPT CMV Luc ires GFP as viral genome, pMD.G or pL102 for pseudotyping and pMM310 for ß-lactamase-Vpr chimeric protein expression 29 . Two days post transfection viral supernatant was removed and concentrated 10 fold by centrifugation through a 20% sucrose gradient. For fusion assays 1 × 10 6 Jurkat cells were seeded in a 24 well plate, in a total volume of 500 µl. Then 150 µl of HIV-enveloped or 80 µl VSV-G pseudotyped virus was added. After incubation for 3.5 h at 37 °C cells were washed twice with CO 2 independent, serum-free media. To allow uptake of the fluorescence substrate, cells were resuspended in 500 µl loading solution, which was prepared directly before use and consists of: CO 2 independent, serum-free media, 10 mM HEPES, 1% probenecid, 0.015% solution A (CCF2-AM) and 0.08% solution B (100 mg/ml Pluronic-F127R, 0.1% acetic acid). Solution A and B were obtained from the GeneBLAzer Detection Kit from Invitrogen. After incubation for 15 h at 25 °C, cells were washed with PBS and fluorescence was measured with BD FACS canto II (BD Biosciences) analysis were done using the software FlowJo 7.6.3 (Tree Star, Ashland, USA).
Biotinylation. HIV-1 envelope protein was obtained from the NIH AIDS Reagent Program (HIV-1 gp120IIIB (CHO), No. 11784) or purified with Ni-NTA Agarose beads from cell lysate after transfection with pEFgp120-linker-myc-6HIS plasmid. Proteins were labeled with activated NHS-biotins (EZ-link Sulfo-NHS-Biotion, 21326, Thermo Fisher Scientific, Waltham, United States), which covalently binds to primary amines. 5 µg protein and 100-fold molecular excess of biotin were incubated 30 min rotating at room temperature, following manufacturer's recommendations. Unbound biotin was removed by washing using desalting columns (UFC503024, Merck Millipore, Darmstadt, Germany). Lysate of un-transfected 293 T cells containing no gp120 protein was processed in parallel as negative control. gp120 binding assay. To analyze attachment of biotinylated gp120 protein to Jurkat cells, 5 × 10 5 cells were washed and incubated with 5 µg protein at RT for 1 h in a total volume of 100 µl. After washing cells were re-suspended in 150 µl buffer containing 2 µg streptavidin APC-C7 (554063, BD Bioscience) and incubated for 30 min at RT to stain attached gp120. After repeated washing 1% paraformaldehyd was used for fixation overnight and fluorescence was analyzed by flow cytometry. For binding and washing HEPES ++ buffer was used (binding buffer: 50 mM HEPES, 5 mM MgCl 2 , 1 mM CaCl 2 , 5% BSA, 0.1% NaN 3 , washing buffer: 50 mM HEPES, 5 mM MgCl 2 , 1 mM CaCl 2 , 5 mM NaCl).

Attachment assay.
To study binding of viral particles to Jurkat cells, virus with Gag-EGFP fusion protein was generated by transfection of 293 T cells with 800 ng pCHIV.eGFP(Env−) and 800 ng pCHIV(Env−) 31 and co-expression of expression plasmids for HIV Env L102 or VSV-G. Fluorescent virus was incubated with target cells for 2 h at 37 °C. Cells were washed two times with PBS and re-suspended in PBS 2% PFA for flow cytometry analysis.