Optimized Replicating Renilla Luciferase Reporter HIV-1 Utilizing Novel Internal Ribosome Entry Site Elements for Native Nef Expression and Function

Cells

All cells were propagated at 37°C in a humidified atmosphere of 5% CO₂. The HEK 293T/17 cell line (CRL-11268) was obtained from the American Type Culture Collection (ATCC, Manassas, VA). We previously generated the TZM-bl dual-reporter cell line, ³⁹ which is also available through the NIH AIDS Reagent Program, Division of AIDS NIAID, NIH (catalog no. 8129). TZM-bl and 293T cell lines were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS), penicillin (100 U/ml), streptomycin (100 μg/ml), l-glutamine (2 mM), and HEPES (25 mM). The A3R5.7 cell line was obtained from Drs. Jerome Kim and Robert McLinden at the U.S. Medical HIV Research Program (MHRP). The A3R5.7 cell line is derived from a human CD4⁺ CXCR4⁺ lymphoblastoid cell line (CEM/A3.01) ⁴⁰ and is engineered to express CCR5. ⁹ A3R5.7 was maintained in RPMI 1640 growth medium supplemented with 10% FBS, penicillin (100 U/ml), streptomycin (100 μg/ml), l-glutamine (2 mM), HEPES (25 mM), and geneticin G-418 sulfate (1 mg/ml) (Life Technologies, Carlsbad, CA). G-418-free growth medium was used for A3R5.7 experiments. The J2574-R5 cell line was kindly provided by Dr. Olaf Kutsch. J2574-R5 is a Jurkat-derived cell line, previously engineered by stable transduction with an HIV-1 LTR-GFP reporter cassette, ^41,42 followed by introduction of constitutive CCR5 expression.

Human primary PBMCs were provided by Tom Denny (Duke University) from the CT-VIMC (OPP1032325), part of the CAVD, funded by the Bill & Melinda Gates Foundation. PBMCs were thawed and placed for 48 h into “PBMC culture medium” [i.e., RPMI 1640 growth medium supplemented with 20% FBS, l-glutamine (2 mM), penicillin (100 U/ml), streptomycin (100 μg/ml), HEPES (25 mM), and interleukin-2 (IL-2; 30 U/ml) (Roche, Indianapolis, IN)] and supplemented with phytohemagglutinin (PHA)-P (5 μg/ml) (Sigma-Aldrich, St. Louis, MO). Following PHA-P stimulation, the medium was removed, and cells were placed into fresh “PBMC culture medium” and used immediately.

Proviral plasmid construction: IRES constructs

Previously, we described the construction of a LucR-expressing, replication-competent HIV-1 proviral DNA (designated pNL-LucR.T2A), as well as several derivative LucR.T2A proviruses in which the ectodomain of gp160/Env is encoded by heterologous HIV-1 strain env sequences, including pNL-LucR.T2A-BaL.ecto. ⁸ In the present study, we modified this approach by replacing the bicistronic LucR.T2A-nef fragment in pNL-LucR.T2A-BaL.ecto with a panel of bicistronic LucR.IRES-nef cassettes of different lengths. Detailed cloning schema and methods are available upon request and are summarized hereafter and in Supplementary Fig. S1 (Supplementary Data are available online at www.liebertpub.com/aid). Briefly, 11 different NheI-Kozak-LucR-spacer-IRES cassettes were inserted into the pNL-LucR.T2A-BaL.ecto backbone in place of pNL4-3 nt 8786, between the env stop codon (nt 8783–8785) and the nef start codon (8787–8789) (positions are based on the pNL4-3 backbone; GenBank ID: M19921.1). “LucR” corresponds to the entire Renilla reniformis luciferase reporter gene, including the stop codon, from phRL-CMV (Promega, Madison, WI) (nt 1068–2003; GenBank ID: AF362549). “Spacer” refers to one of two different 26-nt spacers (Spacer A: atcgatgccgccaccatggacaattg; Spacer B: atcgatgccgccaccatggagggtta).

“IRES” elements were amplified from several sources. The Israeli acute paralysis virus (IAPV) intergenic region (IGR) IRES (IAPVi) was amplified from a plasmid that was generously provided by Dr. Sunnie R. Thompson (UAB). ^43,44 The region amplified corresponds to viral genome nt 6399–6617 (GenBank ID: NC_009025) but with the following compensatory mutations: CA→TC (nt 6566–6567) and ATG→TGA (nt 6612–6614) (termed IAPV_uga/uc in Hertz et al. ⁴³ ). The cricket paralysis virus (CrPV) IGR IRES (CrPVi) was amplified from a plasmid generously provided by Dr. Sunnie R. Thompson (UAB). ^43,45 The region amplified corresponds to viral genome nt 6027–6216 (GenBank ID: AF218039). The HCV 5′-UTR (untranslated region) IRES (HCVi) was also amplified from a plasmid generously provided by Dr. Sunnie R. Thompson (UAB). ^45,46 The region amplified corresponds to viral genome nt 18–341 (GenBank ID: AB691953). The EMCV IRES (GenBank ID: NC_001479) was amplified from the plasmids pIRES-EGFP (Clontech, Mountain View, CA) and pNL-ENG1-IRES-BaL.ecto (also termed pNLENG1i-BaL.ecto or simply “NLENG1i”). ^7,27,28

These two EMCV IRES elements differ in that the pNLENG1i-BaL.ecto contains the “wild-type” (A)₆ (or “6A”) bifurcation loop and pIRES-EGFP contains the (A)₇ (or “7A”) bifurcation loop, which leads to known differences in translation efficacy. ⁴⁷ From each vector, two regions were amplified corresponding to viral genome nt 258–833 and nt 399–833, for “full-length” and “truncated” (TR) forms, ^47,48 respectively. In addition, all four forms were also amplified to include a spacer or “attenuator” at the 3′ end of the IRES, just before the ATG codon of nef, and are referred to as (att)enuated. The 29-nt attenuator (tctagactcacaaccccagaaacagacat) corresponds to an XbaI restriction site (tctaga) fused to nt 2692–2716 of the mouse β-globin major 5′-UTR (GenBank ID: J00413). ⁴⁹ Overall, eight different EMCV IRES elements were created, referred to as 6Ai, 6ATRi, 7Ai, 7ATRi, att6Ai, att6ATRi, att7Ai, and att7ATRi.

Proviral plasmid construction: Nef mutants

A series of Nef mutants (see Fig. 1A) were created in the pNL-BaL.ecto, ⁸ pNL-LucR.T2A-BaL.ecto, ⁸ and pNL-LucR.6ATRi-BaL.ecto (described above) proviral genomes using standard cloning techniques. To briefly summarize these methods, the Nef^G2A mutant is based on that reported in Welker et al. in which the Gly-to-Ala change at amino acid position 2 (G2A) resulted in an inability to N-myristoylate (Δmyr) the N-terminus of Nef. ³⁸ The Nef^KR mutant [mutation of six basic amino acid residues important for Nef association with the membrane (Lys and Arg to Val or Ala, see Fig. 1A)] is based on the equally named mutant reported in Bentham et al. ⁵⁰ The Nef^STOP mutant was created by combining the G2A mutation with the mutation of the fourth and fifth codons to stop codons. The presence of the T2A motif followed by an XbaI site adds three extra amino acids (Pro-Ser-Arg, or “PSR”) to the N-terminus of Nef (T2A-Nef). Therefore, two additional mutants were created, T2A^PSD-Nef and T2A^PSS-Nef, to mutate the basic Arg (R) residue to Asp (D) or Ser (S), respectively. Furthermore, several of the mutations were combined to produce additional functional Nef mutants (see Fig. 1A). Detailed cloning schema and methods are available upon request.

OPEN IN VIEWER

FIG. 1.

Decreased expression and functional impairment of Nef in J2574-R5 cells infected with LucR.T2A reporter viruses compared to nonreporter viruses. (A) Schematic representation and N-terminal amino acid sequence detail of Nef in nonreporter (NL4-3 and NL-BaL.ecto) and LucR.T2A reporter (NL-LucR.T2A and NL-LucR.T2A-BaL.ecto) constructs and the various mutations of nef that were incorporated into either background. Mutations that abrogate either myristoylation (Nef^G2A), basic residue-mediated membrane association (Nef^KR), or expression (Nef^STOP) of Nef are indicated. Additionally, for LucR.T2A reporter viruses, the presence of the T2A motif followed by an XbaI site incorporates three extra amino acids (Pro-Ser-Arg or “PSR”) to the N-terminus of Nef (T2A-Nef). Therefore two additional mutants were created, T2A^PSD-Nef and T2A^PSS-Nef, to mutate the basic Arg (R) residue to an Asp (D) or Ser (S), respectively. (B) Western blot evaluation of Nef levels in 293T cells 48 h after transfection with LucR.T2A reporter and nonreporter proviral DNA. Replica blots were prepared in parallel and probed with anti-p24 (Gag) to demonstrate equivalent transfection efficiencies. (C, D) Evaluation of expression (C) and function (D) of Nef in a Jurkat-derived T cell line, J2574-R5, after infection with 293T transfection-derived LucR.T2A reporter and nonreporter viruses prepared in (B). J2574-R5 cells express GFP upon HIV-1 infection, which was exploited to normalize the number of infected cells analyzed by Western blot, and for gating on HIV-1-infected (GFP⁺) cells for flow cytometric analysis. (C) Western blot evaluation of Nef levels in J2574-R5 cells 48 hpi with anti-Nef monoclonal antibody (mAb) clone EH1. Replica blots were prepared in parallel and probed with anti-p24 (Gag) to demonstrate loading of equivalent numbers of infected cells. (D) Flow cytometric evaluation of Nef function, as measured by CD4 (left panels) and MHC-I (right panels) down-modulation on infected (GFP⁺) cells at 24 and 48 hpi, respectively. Levels of CD4 and MHC-I down-regulation were compared to nonreporter viruses (NL4-3 or NL-BaL.ecto) encoding wild-type Nef (filled histograms). Insets illustrate relative numbers of cells (y-axis) that are GFP⁺ (x-axis) and analyzed in each panel. Color images available online at www.liebertpub.com/aid

Generation of virus stocks and determination of virus infectivity

293T transfection-derived virus stocks were produced by transfection of proviral DNA using FuGENE 6 according to the manufacturer's protocol (Promega) as previously described. ^4,8 Medium was changed after 24 h, and viral supernatants were harvested 72 h posttransfection, clarified at 300×g for 12 min, and frozen at −80°C. Virus stocks were analyzed for HIV-1 p24 antigen concentration by ELISA (PerkinElmer, Groningen, The Netherlands) as indicated and were titered on the TZM-bl reporter cell line by enumeration of beta-galactosidase (β-gal)-stained colonies [TZM-bl infectious units (IU)] as previously described. ³⁹ The titered virus stocks were further analyzed for infectivity and LucR gene expression in the TZM-bl and A3R5.7 cell lines. Briefly, flat- or round-bottom 96-well plates were seeded at 1×10⁴ TZM-bl cells per well 1 day prior to infection or at 9×10⁴ A3R5.7 cells per well on the day of infection, respectively.

On the day of infection, virus was serially diluted starting with a normalized TZM-bl IU input multiplicity of infection (MOI) of 1. The linear range of the LucR readout was examined by infecting TZM-bl cells in triplicate at an MOI equivalent to 0.01, 0.05, 0.1, 0.5, and 1.0 TZM-bl IU (based on β-gal enumeration) in the presence of DEAE-dextran (10 μg/ml). At 48 h postinfection (hpi), 50 μl of 5×Passive Lysis Buffer (PLB; Promega) was added (for a final concentration of 1×) directly to each well containing cells. Samples were frozen at −80°C until use. Samples were brought to room temperature and 20 μl of each sample was analyzed for provirally encoded LucR [Renilla Luciferase Assay System (Promega; catalog no. E2810)] and cell-encoded firefly luciferase (LucF) [as a measure of virus infectivity, irrespective of LucR reporter gene expression; Luciferase Assay System (Promega; catalog no. E4530)] using a Victor X Light Luminometer with single-channel injector (PerkinElmer). Similarly, the linear range of the LucR expression after infection of (nonreporter) A3R5.7 cells with normalized and serially diluted Env-IMC-LucR.T2A and Env-IMC-LucR.IRES viruses [in the presence of DEAE-dextran (5 μg/ml)] was determined at 48 hpi, using the Renilla Luciferase Assay System, including its 5× Renilla Luciferase Assay Lysis Buffer (Promega).

Replication kinetics in PBMCs and determination of LucR stability

PBMCs (2×10⁷) from four different donors were thawed and individually stimulated with PHA-P (5 μg/ml) and IL-2 (30 U/ml) for 48 h, as described above. Cells were then combined (8×10⁷ PBMCs in total) and depleted of CD8⁺ cells using MACS CD8 MicroBeads according to the manufacturer's protocol (Miltenyi Biotec Inc., Auburn, CA). Following depletion, 2×10⁶ cells were infected with 1×10⁶ TZM-bl IU (MOI of 0.05) of each virus to be analyzed, with and without the protease inhibitor indinavir sulfate (1 μM; obtained through the NIH AIDS Reagent Program, Division of AIDS NIAID, NIH). The cells were incubated with virus in screw-cap Eppendorf tubes in a total volume of 1 ml PBMC culture medium and rotated for 4 h at 37°C. The cells were then washed twice with prewarmed “PBMC culture medium” and transferred to a 48-well plate for the duration of the experiment.

Aliquots of supernatants from each sample were taken at the indicated time points and stored at −80°C until all of the samples were collectively analyzed by quantitative HIV-1 p24 ELISA (PerkinElmer). Aliquots of cells were also removed, lysed with 5× Renilla Luciferase Assay Lysis Buffer (Promega), and stored for LucR analysis, as above, at the time supernatants were collected (equivalent cell numbers were collected at all time points). When applicable, indinavir was readded at day 3, and medium removed for sampling was replenished throughout the experimental time course.

Replicate aliquots of the same supernatant samples were also used to determine the stability of LucR expression over several replication cycles, essentially as previously described. ⁸ Briefly, the supernatants collected on days 5, 9, and 16, as well as the cryopreserved 293T transfection-derived virus stocks, were used to infect TZM-bl cells in triplicate with three 5-fold dilutions. The cells were infected in the presence of DEAE-dextran (40 μg/ml) for 4 h, then the medium was replaced with prewarmed cell culture medium. After a 48-h incubation period, the cells were lysed with 5× PLB (Promega) and analyzed for both LucF and LucR activity, as above. Overall infectivity was measured by cell-encoded LucF activity while virus-encoded LucR activity was determined to assess the relative viral reporter gene expression. Values from those input dilutions yielding relative light unit (RLU) values within the linear range were then used to calculate the ratios of LucR to LucF.

Infection of cells for the determination of Nef, CD4, and MHC-I expression

J2574-R5 cells (1×10⁶) were resuspended in vials with 500 μl of medium and 500 μl of each virus. Cells and virus were brought to a final volume of 1 ml in the presence of DEAE-dextran (5 μg/ml) and rotated at 37°C for 4 h. Cells were then transferred to a 24-well plate at which time the final volume was brought up to 2 ml with fresh medium, keeping the concentration of DEAE-dextran at 5 μg/ml, and then incubated overnight at 37°C. At 18–24 hpi, a portion of the infected cells was analyzed for CD4 expression by flow cytometry (see below), and at 48 hpi the remaining infected cells were analyzed for MHC-I expression and prepared for Western blot analysis (see below).

For A3R5.7 and PBMCs, 5×10⁶ cells were resuspended in vials with 500 μl of their respective medium and 500 μl of each virus (or uninfected controls), rotating them at 37°C for 4 h to promote infection [DEAE-dextran (5 μg/ml) was included for A3R5.7 cells]. The contents of the vials were then transferred into T-12.5-cm² flasks, at which point the final volume was brought up to 2 ml with fresh medium, and then incubated overnight at 37°C. The next morning 2 ml of fresh medium was added to each flask. At 72 hpi, 1 ml of medium was removed and replaced with 2 ml of fresh medium. The day after (∼96 hpi) the cells were counted, and aliquots from each flask were taken to determine, by flow cytometry, the number of infected cells by intracellular p24 staining, as well as MHC-I down-modulation (see below). Samples were also prepared for Western blot (see below).

Flow cytometric analysis for the number of infected cells and CD4 and MHC-I expression levels

To examine Nef functionality, infected cells were identified by flow cytometry by either GFP expression (for J2574-R5) or intracellular p24 staining with FITC-conjugated mouse anti-HIV-1 p24 (clone KC57; Beckman Coulter, Brea, CA) monoclonal antibody (mAb) (for A3R5.7 and PBMCs). Within these gates (i.e., positivity for GFP or FITC), levels of cell surface MHC-I and CD4 were determined using the following mouse mAbs for flow cytometry: APC-conjugated antihuman CD4 (clone L200) and APC-conjugated antihuman MHC-I (clone G46-2.6, recognizing a monomorphic epitope on HLA-A, HLA-B, and HLA-C) purchased from BD Biosciences, Pharmingen (San Diego, CA). Cells were assayed using a BD FACSCalibur flow cytometer and the resulting data were analyzed with FlowJo software (TreeStar Inc., Ashland, OR).

Western blot analysis

Provirally transfected 293T cells were washed with phosphate-buffered saline (PBS) and lysed with Laemmli sample buffer (Sigma-Aldrich) 60–72 h posttransfection. When indicated, virions were concentrated from the same six-well plate cultures (three wells per sample, ∼7.5 ml) by layering cell-free supernatants over a 20% sucrose cushion before ultracentrifugation at 25,000 rpm (∼110,000×g) for 2 h (Beckman Optima XL-90 polyallomer tubes; Beckman Coulter). Supernatants were removed and virion pellets were lysed directly in 150 μl Laemmli sample buffer. Cell lysates and virion lysates were sonicated for 40 s and then heated for 10 min at 95°C and frozen at −20°C until use.

J2574-R5 cells were infected as above. At 48 hpi, cells were washed with PBS, counted, pelleted, and then lysed with a low volume of Laemmli sample buffer. The percentage and total number of infected cells (i.e., GFP⁺ cells) were determined by flow cytometric analysis of an aliquot from each sample (see above), and the volume of Laemmli sample buffer was adjusted accordingly to achieve 1×10⁴ infected cells per μl of lysate. Similarly, the percentage and total number of infected A3R5.7 or PBMCs were determined at 96 hpi via intracellular p24 staining by flow cytometric analysis (see above). Based on the available total number of cells and the percentage of infection determined by flow cytometry, aliquots of cells were lysed with a volume of Laemmli sample buffer in order to achieve a final concentration of 1×10⁴ infected cells per μl of lysate. J2574-R5, A3R5.6, and PBMC cell lysates were sonicated for 40 s and then heated for 10 min at 95°C and frozen at −20°C until use.

Before use, all samples were reheated at 95°C for 5 min and the proteins resolved by denaturing SDS-PAGE (10% resolving gel) and transferred to a nitrocellulose membrane. Membranes were blocked with 5% nonfat dried milk in Tris-buffered saline with Tween 20 (TBST) or SuperBlock T20 (PBS) Blocking Buffer (Thermo Scientific, Rockford, IL) and then incubated with the appropriate primary antibodies at 4°C overnight. Mouse mAb to LucR (clone 5B11.2; Chemicon International, Temecula, CA) was used at a 1:1000 dilution. Polyclonal rabbit HIV-1 Nef antiserum (obtained from the NIH AIDS Reagent Program, Division of AIDS NIAID, NIH, contributed by Ronald Swanstrom, catalog no. 2949, Lot# 10-070932) ⁵¹ was used at 1:1000 dilution (Fig. 2B). Mouse mAb to HIV-1 Nef (clone EH1; kindly provided by Dr. James Hoxie at the University of Pennsylvania), directed against the 13 C-terminal residues of SF-2 Nef and reactive against NL4-3 Nef, was used at a 1:1000 dilution (Figs. 1, 4, and 5). Mouse mAb to HIV-1 p24 (Gag) (also known as “ARF”) was prepared from the HIV-1 p24 hybridoma (clone 183-H12-5C; obtained from the NIH AIDS Reagent Program, Division of AIDS NIAID, NIH, contributed by Drs. Bruce Chesebro and Hardy Chen, catalog no. 1513) ⁵² and used at a 1:1000 dilution. Membranes were probed with horseradish peroxidase (HRP)-conjugated goat polyclonal secondary antibodies to mouse IgG or rabbit IgG (SouthernBiotech, Birmingham, AL) and developed with Immobilon Western Chemiluminescent HRP Substrate (Millipore, Billerica, MA) or Amersham ECL Western Blotting Detection Reagents (GE Healthcare UK Limited, Little Chalfont, Buckinghamshire, UK). PageRuler Prestained Protein Ladder (Thermo Scientific) was used to estimate the molecular weight of bands.

OPEN IN VIEWER

FIG. 2.

Derivation of a panel of novel LucR.IRES reporter viruses. (A) Schematic representation of the insertion of various LucR.IRES elements into the NL4-3 or NL-Env.ecto (e.g., NL-BaL.ecto) backbones (also see Supplementary Fig. S1). Eleven LucR.IRES elements were inserted between the NL4-3 env and nef genes. The nucleotide (nt) sequence is shown for the junction between the stop codon (taa) of env and the start codon (atg) of LucR, which contains an NheI restriction site (gctagc) and Kozak sequence (ccacc). Each LucR.IRES contains a 26 nt “IRES spacer” between the LucR gene and internal ribosome entry site (IRES) element. For the encephalomyocarditis virus (EMCV) IRES, two regions were amplified corresponding to “full-length” and “truncated” (TR) forms (where the 3′ nucleotide sequences are identical). Each of these EMCV IRES elements was also designed to contain either the “wild-type” (A)₆ (i.e., “6A”) or (A)₇ (i.e., “7A”) bifurcation loop, which corresponds to known differences in translation efficacy. In addition, all four forms were also engineered to include a 29 nt spacer or “(att)enuator” at the 3′ end of the IRES, just before the start codon of nef, resulting in eight EMCV IRES variants (6Ai, 7Ai, 6ATRi, 7ATRi, att6Ai, att7Ai, att6ATRi, and att7ATRi). LucR.T2A (described elsewhere ⁸ ) is also shown for comparison. The LucR gene (excluding stop codon) and 18 amino acid T2A sequence (underlined; “EGR…PGP”) were fused in-frame. The closed arrowhead (▼) indicates the cotranslational cleavage point between the penultimate (Gly, G) and last amino acid (Pro, P) of T2A. The first two amino acids (→; Met-Gly or “MG”) of Nef are also depicted, as well as the translated XbaI restriction site (Ser-Arg or “SR”) that was introduced by design (see Fig. 1A). Cleavage of the LucR.T2A-Nef bicistron results in the N-terminus of Nef containing three additional amino acids preceding the Met (Pro-Ser-Arg “PSR”). The length of each element (“X”) is shown to the right. The asterisk (*) indicates that the element is 1 nt longer in the 7A variant than in the 6A variant. (B) Western blot evaluation of Nef (upper panel; polyclonal rabbit HIV-1 Nef antiserum) and LucR (middle panel) levels in 293T cells 48 h after transfection with nonreporter (NL-BaL.ecto), LucR.T2A reporter (NL-LucR.T2A-BaL.ecto), and the 11 LucR.IRES reporter proviral DNA. Replica blots were prepared in parallel and probed with anti-p24 (Gag) (lower panel) to demonstrate equivalent transfection efficiencies. The LucR.T2A reporter and three LucR.IRES reporter constructs chosen for further analysis are denoted by a closed diamond (♦). (A, B) IAPV, Israeli acute paralysis virus; CrPV, cricket paralysis virus; HCV, hepatitis C virus; i, IRES. Color images available online at www.liebertpub.com/aid

A3R5/Env-IMC-LucR neutralization assay

To analyze neutralization of HIV-1 infection in A3R5.7 cells, we followed the standardized protocol essentially as described. ¹⁰ Briefly, appropriate volumes of 293T transfection-derived virus (NL-LucR.T2A-BaL.ecto or NL-LucR.6ATRi-BaL.ecto) that would yield approximately 1.5×10⁵ LucR RLU under the sampling conditions were added to each well of a 96-well plate with different bNAbs (see below) diluted in a series of eight 3-fold dilutions from starting concentrations of either 25 or 5 μg/ml. After incubation for 45–90 min at 37°C in 96-well plates, 9×10⁴ A3R5.7 cells (in 100 μl) were added to each well in a final volume of 250 μl. After incubation for 96 h at 37°C, 125 μl of medium was removed from each well and 30 μl of 5× Renilla Luciferase Assay Lysis Buffer (Promega) was added (for a final concentration of 1×) directly to each well containing cells. Samples were frozen at −80°C until use. At that point, samples were brought to room temperature and 20 μl of each sample was analyzed for LucR activity, as described above. Percent neutralization was calculated as previously described. ⁸

The 2F5, 4E10, IgG1b12, PG9, PG16, and 2G12 bNAbs were obtained from Polymun GmbH (Vienna, Austria). CH01 and VRC-CH31 were manufactured by Catalent (Madison, WI) with permission from the Center for HIV/AIDS Vaccine Immunology (CHAVI). 2F5 and 4E10 are directed against the HIV-1 gp41 membrane-proximal external region (MPER). ^53,54 IgG1b12 (b12) and VRC-CH31 are directed against the HIV-1 gp120 CD4-binding site ^55,56 ; CH01, PG9, and PG16 are directed against the HIV-1 gp120 V1-V2 conformational epitope ^{56 –59} ; and 2G12 is directed against a distinct HIV-1 gp120 epitope. ⁶⁰

Nef expression and function in LucR.T2A reporter viruses

Nef is a 27-kDa myristoylated protein that is not essential for HIV-1 replication in cell culture (reviewed in Foster and Garcia ⁶¹ ). In vivo, the maintenance of high viral loads and the progression to AIDS appear to require functional Nef. ^62,63 Among its many functions, Nef has been shown to down-regulate CD4 and MHC-I, modulate signal transduction pathways, and enhance viral infectivity (reviewed in Foster and Garcia ⁶¹ ). Down-modulation of CD4 and MHC-I as well as incorporation of Nef into virus particles require Nef association with the plasma membrane. ^38,50,64 Membrane localization is mediated by complex processes involving features within the N-terminal anchor region of Nef: N-terminal myristoylation and a cluster of basic amino acids. ^38,50

Our previously described LucR.T2A reporter virus strategy entailed expression of LucR and Nef from a bicistronic mRNA in which translation of Nef is linked in-frame to LucR with the T2A peptide, which results in ribosome skipping of the last peptide bond at the C-terminus of the T2A peptide that leaves the last T2A-encoded amino acid residue (Pro) to be added to the N-terminus of Nef. In addition, two additional residues (Ser-Arg) following the Pro are introduced from translation of a 6-nt sequence encoding an XbaI restriction site engineered in the vector's design. Thus, the LucR.T2A strategy may lead to impairment of this modified Nef's (“T2A-Nef”) functions and/or viral incorporation secondary to loss of the N-terminal myristoylation signal (i.e., N-terminal Gly) ^{36, 37} (Fig. 1A). We have previously shown that for LucR.T2A reporter viruses (NL-LucR.T2A, NL-LucR.T2A-BaL.ecto, and NL-LucR.T2A-SF162.ecto), levels of Nef expression in 293T cells following transfection of proviral DNA were somewhat lower than in parental nonreporter viruses (NL4-3, NL-BaL.ecto, and NL-SF162.ecto) ⁸ ; however, Nef function in T cells infected with these reporter viruses has not been analyzed.

Here, we further evaluated the expression and function of T2A-Nef in transfected 293T cells and in infected CD4⁺ T cell lines and PBMCs. We compared the LucR.T2A reporter viruses, NL-LucR.T2A and NL-LucR.T2A-BaL.ecto, with their parental nonreporter viruses, NL4-3 and NL-BaL.ecto, respectively. The R5-tropic virus expressing the BaL reference env was chosen as a stand-in for the large panel (n>120) of Env-IMC-LucR.T2A encoding env genes from heterologous HIV-1 strains ^4,8,11,17 (also Jones et al., unpublished observations). Additional nef mutations were analyzed (Fig. 1A) in both LucR.T2A reporter (NL-LucR.T2A-BaL.ecto) and nonreporter (NL-BaL.ecto) backbones, including those that abrogate either myristoylation (Nef^G2A), membrane association (Nef^KR), or expression (Nef^STOP) of Nef. As noted above, the T2A motif followed by an XbaI site added three extra amino acids (Pro-Ser-Arg) to the N-terminus of Nef (T2A-Nef). Therefore, two additional mutants were created, T2A^PSD-Nef and T2A^PSS-Nef, mutating the basic Arg (R) residue to either Asp (D) or Ser (S), in order to evaluate (1) whether the basic residue in T2A-Nef may, at least in part, functionally compensate for the loss of myristoylation through provision of an additional residue available for membrane association, and (2) whether removal of this additional charged residue would affect Nef function in the T2A context.

First, we compared Nef expression in 293T cells transfected with LucR.T2A reporter (NL-LucR.T2A and NL-LucR.T2A-BaL.ecto) and nonreporter (NL4-3 and NL-BaL.ecto) constructs by Western blot (Fig. 1B). Consistent with previous results, ⁸ T2A-mediated Nef expression was reduced compared to Nef expression from matched nonreporter constructs and the slightly higher molecular weight of T2A-Nef (Fig. 1B) was consistent with the three additional C-terminal amino acids (Fig. 1A). Replica blots probed with anti-p24 (Gag) mAb demonstrated proteolytic processing and similar levels of Gag expression (Fig. 1B). The Nef^G2A mutation did not adversely affect the Nef expression level in the context of NL-BaL.ecto virus; however, the Nef^KR mutation reproducibly reduced Nef expression, consistent with the findings of Bentham et al. ⁵⁰ None of the mutations analyzed in the LucR.T2A background had further adverse effect on the expression levels of Nef (Fig. 1B). As expected, Nef expression was undetectable for constructs containing Nef^STOP (Fig. 1B).

We next evaluated Nef expression and function, as measured by down-modulation of CD4 (at 24 h) and MHC-I (at 48 h), in infected J2574-R5 cells (Fig. 1C and D). These Jurkat-derived cells express GFP upon HIV-1 infection, which we exploited to normalize the number of infected cells analyzed by Western blot, and for gating on HIV-1-infected (GFP⁺) cells using flow cytometric analysis. The levels of wild-type Nef versus Nef mutants (i.e., Nef^G2A and Nef^KR) expressed from nonreporter viruses (NL-BaL.ecto) in J2574-R5 cells at 48 hpi (Fig. 1C) were similar to those observed in 293T cells following transfection of the respective proviral DNAs (Fig. 1B). Unexpectedly and importantly, the same level of T2A-Nef expression observed in the context of 293T cells transfected with proviral DNA (NL-LucR.T2A and NL-LucR.T2A-BaL.ecto) (Fig. 1B) was not exhibited upon productive infection of J2574-R5 cells with these LucR.T2A reporter viruses (Fig. 1C). Rather, infection of J2574-R5 cells with LucR.T2A reporter viruses resulted in severely reduced Nef levels compared to cells infected with matched, wild-type nef-encoding nonreporter viruses (Fig. 1C).

The effect of Nef on CD4 surface expression is best assessed early in infection since it will otherwise become increasingly obscured by down-modulation mediated by Vpu and Env. We assessed CD4 down-modulation at 24 hpi, when GFP expression in infected cells reached levels sufficient for reliable gating (Fig. 1D, insets). At this time point, infection with nonreporter viruses (NL4-3 and NL-BaL.ecto) encoding wild-type Nef resulted in nearly complete loss of surface CD4 expression. CD4 down-regulation was significantly impaired for NL-BaL.ecto encoding Nef^STOP virus (Fig. 1D, left upper panel). As expected, abrogating Nef myristoylation (Nef^G2A) did not affect Nef-dependent CD4 down-modulation, whereas abrogating Nef membrane association (Nef^KR) resulted in an intermediate phenotype (Fig. 1D, left upper panel), despite the low level of Nef^KR observed by Western blot analysis (Fig. 1C). However, and in line with the nearly complete absence of detectable T2A-Nef in J2574-R5 cells infected with LucR.T2A reporter viruses by Western blot (Fig. 1C), Nef-dependent down-regulation of CD4 was not observed with T2A-Nef, T2A-Nef^STOP, or with any of the T2A-Nef mutants (i.e., KR, PSD, or PSS) (Fig. 1D, left middle and lower panels). Interestingly, we observed an overall lesser degree of non-Nef-mediated (i.e., Vpu-mediated) CD4 down-regulation for LucR.T2A reporter viruses containing T2A-Nef^PSD and T2A-Nef^PSS (Fig. 1D, left lower panel); however, this may be related to very low infection levels, and thus possibly an earlier stage of the replication cycle, at the time of analysis.

We also assessed Nef effects on surface MHC-I in infected J2574-R5 cells at 48 hpi. As expected, infection with nonreporter viruses (NL4-3 and NL-BaL.ecto) expressing wild-type Nef resulted in lower MHC-I surface levels than infection with viruses lacking detectable Nef expression (i.e., Nef^STOP) (Fig. 1D, right upper and middle panels) and similar (only slightly lower) MHC-I surface levels as seen in mock-infected control cells (see Fig. 6, left upper panel). Down-modulation of MHC-I was either completely inhibited by abrogating Nef myristoylation (Nef^G2A) or impaired by mutations that affect Nef membrane association (Nef^KR) (Fig. 1D, right upper panel). Down-modulation of surface MHC-I expression was also not observed on J2574-R5 cells infected with LucR.T2A reporter viruses (NL-LucR.T2A and NL-LucR.T2A-BaL.ecto) encoding T2A-Nef or its mutants (Fig. 1D, right middle and lower panels). This result was again consistent with the very low or undetectable levels of Nef in these infected cells (Fig. 1C). Thus, surface expression of MHC-I in LucR.T2A reporter virus-infected cells was unlike that of nonreporter virus-infected cells expressing wild-type Nef; rather, it was similar to that detected on cells infected with nonreporter viruses with altered (Nef^G2A or Nef^KR) or abrogated (Nef^STOP) expression of Nef (Fig. 1D).

Derivation of novel LucR.IRES reporter viruses

Given that both Nef expression and function, as measured by down-modulation of CD4 and MHC-I, were impaired in the context of LucR.T2A reporter virus infection of Jurkat T cells (Fig. 1), as well as A3R5.7 cells and PBMCs (see below, Figs. 5 and 6), we investigated alternative reporter virus strategies that utilized an IRES bicistronic reporter system. We and others have found the performance of previous EMCV IRES HIV-1 reporter viruses (e.g., NLENG1i) ^{3,27 –29} to be suboptimal with respect to replication, stability of the EGFP reporter gene (personal observations), and supraphysiological overexpression of Nef. ⁸ Therefore, we examined several different IRES elements, including eight iterations of an EMCV IRES element similar to that found in the previous NLENG1i virus strategy. These elements were engineered in place of the T2A element within the LucR.T2A reporter virus (Fig. 2A).

Four well-characterized IRES elements from the viruses, IAPV, CrPV, HCV, and EMCV (Fig. 2A and Supplementary Fig. S1), were chosen based on a number of characteristics, with particular attention given to the size and previously reported translational activity levels. The 5′-UTR IRES from HCV was of moderate size (324 nt) and activity. ⁶⁵ In contrast, the smaller IGR IRES elements from the insect viruses IAPV and CrPV have evolved to function in a bicistronic manner (as opposed to the 5′-UTR), ^66,67 similar to the design of our reporter virus technology concept. In addition, these were the smallest of the elements (219 and 190 nt, respectively) tested, and we predicted that they would have the least negative impact on reporter gene retention and stability.

The EMCV IRES is among the best described IRES elements. To overcome limitations observed in the context of NLENG1i-based HIV-1 strains, we engineered the EMCV IRES with eight different variations that have been described in the literature. The first variant was based on the “wild-type” (A)₆ (or “6A”) versus (A)₇ (or “7A”; found in pIRES-EGFP) bifurcation loop, which leads to known differences in translation efficacy. ⁴⁷ Given the relatively large size of the preferred “consensus” EMCV IRES found in a number of commercial plasmids, we derived both “full-length” and minimal, or “truncated” (TR), forms (576 and 435 nt, respectively). ⁴⁷ All four of these variants were configured to utilize the preferred native “ATG” in EMCV as the start codon for Nef. Lastly, each of the above variants was also designed to include a 29-nt spacer or “attenuator” ⁴⁹ at the 3′ end of the IRES (in place of the native EMCV “ATG” codon), just before the ATG codon of Nef; these are referred to as the “attenuated” (att) variants (as inclusion of the spacer is expected to result in reduced Nef expression compared to EMCV IRES without it). ⁴⁷ Overall, eight different EMCV IRES elements were created and designated as described in Materials and Methods (e.g., “6Ai” indicates the full-length EMCV IRES with the 6A bifurcation loop, and “att7ATRi” indicates the truncated EMCV IRES with the 7A bifurcation loop and attenuator spacer) (see also Supplementary Fig. S1).

Infectivity of 293T transfection-derived LucR.IRES reporter viruses

The 11 different 293T transfection-derived LucR.IRES reporter viruses showed infectious titers (IU/ml) and relative infectivity (IU/ng p24) similar to that of LucR.T2A reporter (NL-LucR.T2A-BaL.ecto) and nonreporter (NL-BaL.ecto) viruses in TZM-bl cells (data not shown). Furthermore, for each virus, linear responses of both cell-encoded LucF and virus-encoded LucR activities were observed up to an MOI of 0.5 following infection of TZM-bl cells (Supplementary Fig. S2A and B). As expected, infection with LucR.IRES reporter viruses resulted in nearly identical LucF values compared to cognate LucR.T2A reporter and nonreporter viruses, indicating similar cell-encoded LTR transactivation in the TZM-bl cell line (Supplementary Fig. S2A). Among all LucR.IRES reporter viruses, RLU values of LucR activity fell within a 3-fold boundary of RLU values obtained with the LucR.T2A reporter virus (Supplementary Fig. S2B). In A3R5.7 cells (which together with LucR.T2A reporter viruses have found application for sensitive detection of neutralizing immune responses ^9,10 ), linear responses of LucR activity were also observed for all LucR.IRES reporter viruses, which were once again similar to and within the 3-fold range of LucR.T2A reporter virus RLU values (Supplementary Fig. S2C).

Nef expression in 293T cells following transfection with LucR.IRES reporter proviral DNA

Expression of Nef in 293T cells transfected with the panel of LucR.IRES reporter constructs was compared to that from cognate LucR.T2A reporter (NL-LucR.T2A-BaL.ecto) and nonreporter (NL-BaL.ecto) constructs (Fig. 2B). We did not detect IAPVi-mediated and CrPVi-mediated expression of Nef. HCVi-mediated Nef expression was detectable but lower than T2A-mediated expression, which, as seen before (Fig. 1B and Edmonds et al. ⁸ ), was reduced compared to nonreporter virus (Fig. 2B, upper panel). As hypothesized, we observed a range of Nef expression levels from the eight EMCV IRES variants (Fig. 2B, upper panel). Supraphysiological levels of Nef expression were seen with the LucR.6Ai and LucR.7Ai reporter constructs, comparable to expression levels detected for the NLENG1i reporter construct (data not shown and Edmonds et al. ⁸ ); as expected, the “7A” variants generally demonstrated reduced expression compared to their “6A” counterparts (i.e., 6Ai vs. 7Ai, 6ATRi vs. 7ATRi, att6Ai vs. att7Ai, and att6ATRi vs. att7ATRi); the shorter “TR” variants resulted in lower Nef levels than the “full-length” versions (i.e., 6Ai vs. 6ATRi, 7Ai vs. 7ATRi, att6Ai vs. att6ATRi, and att7Ai vs. att7ATRi), and “attenuated” (att) variants resulted in lower Nef expression than their nonattenuated counterparts (i.e., att6Ai vs. 6Ai, att7Ai vs. 7Ai, att6ATRi vs. 6ATRi, and att7ATRi vs. 7ATRi). Furthermore, products of Nef proteolytic cleavage by HIV-1 protease, a process of unknown relevance that predominantly occurs in virions, ⁶⁸ are also detectable in provirally transfected 293T cells ⁶⁹ (and O. Fackler, personal communication), and become more prominent with increasing, and especially supraphysiological levels of Nef expression. As before, the proteolytic processing of Gag was normal and equal across samples, but supraphysiological Nef levels seem to have a moderate effect on Gag processing (Fig. 2B, lower panel). Equivalent expression levels of LucR were seen for the LucR.T2A and all LucR.IRES reporter constructs (Fig. 2B, middle panel).

Based on Nef expression levels most closely matching those of nonreporter or LucR.T2A reporter viruses, and avoiding supraphysiological Nef levels, we selected a subset of three LucR.IRES reporter viruses to evaluate further: LucR.HCVi (NL-LucR.HCVi-BaL.ecto), LucR.6ATRi (NL-LucR.6ATRi-BaL.ecto), and LucR.att6ATRi (NL-LucR.att6ATRi-BaL.ecto).

Replication kinetics and stability of LucR expression for LucR.IRES reporter viruses in PBMCs

The addition of exogenous sequence into the HIV-1 genome can have negative effects on viral replication, presenting as severely delayed replication kinetics and/or rapid selection of viruses defective for the expression of the exogenous reporter gene. ^{70 –72} Thus, we tested the selected subset of LucR.IRES reporter viruses for replication fitness by measuring p24 antigen production and LucR activity over time after infection at a low MOI of PHA-P-stimulated pooled donor PBMCs depleted of CD8⁺ cells (Fig. 3). Measurement of p24 antigen production (Fig. 3B) confirmed that the LucR.T2A reporter virus (NL-LucR.T2A-BaL.ecto) was replication competent with slightly delayed replication kinetics compared to the nonreporter virus (NL-BaL.ecto), in line with our previous results. ⁸ The three tested LucR.IRES reporter viruses were replication competent similar to the LucR.T2A virus (Fig. 3B). The kinetics of LucR activity (Fig. 3A), as a sensitive and robust measure of proviral gene expression and viral replication, were also highly similar in PBMCs infected with LucR.T2A reporter virus and all three LucR.IRES reporter viruses. In an independent experiment, inclusion of the protease inhibitor, idinavir (IDV), which restricts infection to a single round, validated that the ∼100-fold increase in LucR activity in the absence of IDV beyond day 2 postinfection was attributable to virus replication and spread (Fig. 3C).

OPEN IN VIEWER

FIG. 3.

Replication kinetics and stability of Renilla luciferase (LucR) expression for LucR.IRES reporter viruses in peripheral blood mononuclear cells (PBMCs). (A, B) Replication kinetics for a subset of LucR.IRES reporter viruses, LucR.HCVi (NL-LucR.HCVi-BaL.ecto), LucR.att6ATRi (NL-LucR.att6ATRi-BaL.ecto), and LucR.6ATRi (NL-LucR.6ATRi-BaL.ecto), was compared to cognate LucR.T2A reporter (NL-LucR.T2A-BaL.ecto) and nonreporter (NL-BaL.ecto) viruses. Replication kinetics was measured by LucR activity (A) and p24 antigen production (B) over time after infection at a low MOI (MOI of 0.05 TZM-bl IU, corresponding to an approximate MOI of 0.001 PBMC IU) of phytohemagglutinin (PHA)-P-stimulated pooled donor PBMCs depleted of CD8⁺ cells. In (B), the limit of p24 detection (100 pg/ml) is indicated. (C) PBMCs were infected with LucR.T2A and LucR.6ATRi reporter viruses, as above, but in the absence (solid lines) and presence (dashed lines) of the protease inhibitor, indinavir (IDV), and viral replication was monitored by LucR expression (left y-axis, black lines) and p24 (right y-axis, gray lines). p24 production for nonreporter virus is also shown for comparison. (D) To assess the stability of LucR expression from LucR.T2A and LucR.IRES reporter viral genomes over multiple cycles of virus replication, we collected culture supernatants from the infected PBMC cultures at days 5, 9, and 16. TZM-bl cells were then infected with the PBMC infection-derived, as well as the 293T transfection-derived (TD), virus supernatants, as detailed in Materials and Methods. The stability of LucR expression relative to infectivity was calculated as the ratio of virus-encoded LucR RLU to cell-encoded LucF RLU (which represents a measure of the overall infectivity of the respective virus stocks at each time point) and compared to the LucR:LucF ratio obtained with TD supernatants.

Next, we examined whether LucR expression from the viral genome was stably maintained over multiple cycles of virus replication, essentially as previously described. ⁸ For this, 293T transfection-derived virus and supernatants collected at days 5, 9, and 16 from the infected PBMC cultures described above were used to infect TZM-bl cells as detailed in Materials and Methods. Since activity of cell-encoded LucF represents a measure of the overall infectivity of the respective virus stocks at each time point, we calculated the ratio of virus-encoded LucR RLU to LucF RLU for PBMC supernatants and compared it to the LucR:LucF ratio obtained with transfection-derived virus. As illustrated for a representative experiment in Fig. 3D, nearly full retention of LucR activity was observed over the 16-day course of replication of the LucR.HCVi reporter virus. Reduction of LucR activity was seen for LucR.6ATRi and LucR.att6ATRi reporter viruses as of day 16, similar to what was observed for the LucR.T2A reporter virus (NL-LucR.T2A-BaL.ecto) here and previously, ⁸ as well as with two additional LucR.T2A reporter virus constructs with different env genes. ¹²

Nef expression and function in J2574-R5 T cells infected with LucR.IRES reporter viruses

Given that the results for replication kinetics and LucR reporter gene stability in PBMCs were similar for LucR.T2A virus as compared to the subset of three LucR.IRES viruses, we further compared all of them with respect to Nef virion incorporation and Nef expression and function in T cells. In addition, since we had observed the most physiological Nef levels in 293T cells for the LucR.6ATRi reporter virus, we also introduced into this construct the functional nef mutations described above (see Fig. 1A) that abrogate myristoylation (Nef^G2A), membrane association (Nef^KR), or expression (Nef^STOP) of Nef protein to use as controls.

Nef levels in 293T cells following transfection with LucR.T2A (NL-LucR.T2A-BaL.ecto), LucR.HCVi (NL-LucR.HCVi-BaL.ecto), LucR.6ATRi (NL-LucR.6ATRi-BaL.ecto), and LucR.att6ATRi (NL-LucR.att6ATRi-BaL.ecto) reporter constructs (Fig. 4A) were consistent with earlier results (Figs. 1B and 2B). The nef mutations, when analyzed in the context of the LucR.6ATRi reporter backbone, had effects on expression (Fig. 4A) similar to those observed when analyzed in the nonreporter (NL-BaL.ecto) background (Fig. 1B). Incorporation of unaltered Nef and T2A-Nef into virions was reduced disproportionally for those virus constructs that resulted in lower-than-wild-type cellular Nef expression levels (Fig. 4A). In this regard, Nef was substantially decreased in the LucR.T2A reporter virions compared with nonreporter virions, and Nef was absent or significantly decreased in virions for LucR.HCVi and LucR.att6ATRi reporter viruses, respectively (Fig. 4A).

OPEN IN VIEWER

FIG. 4.

Similar expression and function of Nef in J2574-R5 cells infected with LucR.6ATRi reporter or nonreporter viruses. (A) Western blot evaluation of Nef levels (with anti-Nef mAb clone EH1) in 293T cells (top) and virions (bottom) 48 hpi with LucR.T2A reporter (NL-LucR.T2A-BaL.ecto), LucR.HCVi reporter (NL-LucR.HCVi-BaL.ecto), LucR.att6ATRi reporter (NL-LucR.att6ATRi-BaL.ecto), LucR.6ATRi reporter (NL-LucR.6ATRi-BaL.ecto), and nonreporter (NL-BaL.ecto) proviral DNA. Mutations that abrogate either myristoylation (Nef^G2A), basic residue-mediated membrane association (Nef^KR), or expression (Nef^STOP) of Nef were incorporated into the LucR.6ATRi reporter background and were also included. Unaltered (wild-type) Nef is indicated as “Nef.” Replica blots were prepared in parallel and probed with anti-p24 (Gag) to demonstrate equivalent transfection efficiencies and virion production. (B, C) Evaluation of expression (B) and function (C) of Nef in J2574-R5 after infection with 293T transfection-derived reporter and nonreporter viruses prepared in (A). As in Fig. 1, samples were normalized to the number of infected cells before analysis by Western blot. (B) Western blot evaluation of Nef levels in J2574-R5 cells 48 hpi. Replica blots were prepared in parallel and probed with anti-p24 (Gag) to demonstrate loading of equivalent numbers of infected cells. (C) Flow cytometric evaluation of Nef function, as measured by CD4 (left panels) and MHC-I (right panels) down-modulation on infected (GFP⁺) J2574-R5 cells at 24 and 48 hpi, respectively. Levels of CD4 and MHC-I down-regulation for NL-LucR.HCVi-BaL.ecto, NL-LucR.att6ATRi-BaL.ecto, and NL-LucR.6ATRi-BaL.ecto reporter viruses were compared to nonreporter virus (NL-BaL.ecto), containing wild-type Nef (filled histograms), and either NL-BaL.ecto or NL-LucR.6ATRi-BaL.ecto containing the abrogated Nef expression mutation (Nef^STOP) (dashed histogram). Color images available online at www.liebertpub.com/aid

In contrast, 6ATRi-mediated expression of Nef led to similar (to slightly reduced) levels in virions, as compared with the nonreporter virus (Fig. 4A). Abrogation of Nef myristoylation (Nef^G2A) alone did not seem to affect Nef incorporation into virions produced from 293T cells, while the Nef^KR mutant resulted in impaired packaging (Fig. 4A). Probing with anti-p24 (Gag) mAb demonstrated that the proteolytic processing of Gag was normal and that similar amounts of each of the viruses had been produced and analyzed (Fig. 4A). In summary, the Nef virion incorporation phenotype of LucR.6ATRi reporter virus was essentially the same as that of wild-type nonreporter virus.

Similar to earlier results (Fig. 1C), the level of Nef detected in J2574-R5 cells infected with the 293T transfection-derived LucR.T2A reporter virus was significantly less compared to that detected for nonreporter virus (NL-BaL.ecto) (Fig. 4B). Interestingly, in contrast to 293T cell lysates, Nef was nearly absent in J2574-R5 cells infected with the LucR.HCVi and LucR.att6ATRi reporter viruses. On the other hand, in J2574-R5 cells infected with the LucR.6ATRi reporter virus, Nef was expressed at levels similar to wild-type nonreporter virus, with the Nef^G2A mutation imparting only moderate effects on levels of Nef, as would be expected (Fig. 4B). Of note, in cells infected with the LucR.6ATRi reporter virus, the degree of MHC-I down-modulation (Fig. 4C) mediated by Nef essentially reached levels obtained following infection with nonreporter virus, as predicted based on the expression level of this unaltered Nef. As expected, in cells infected with LucR.6ATRi reporter viruses harboring nef functional mutations, MHC-I down-modulation was abrogated (Nef^G2A or Nef^STOP) or impaired (Nef^KR) (Fig. 4C). Consistent with the near absence of detectable Nef by Western blot, we did not observe down-modulation of MHC-I expression in J2574-R5 cells infected with the LucR.HCVi reporter virus (Fig. 4C).

Despite very low Nef levels in cells infected with the LucR.att6ATRi reporter virus (Fig. 4B), MHC-I expression was partially down-modulated (Fig. 4C). The effect of Nef expressed from the various LucR.IRES proviruses on CD4 down-regulation was analyzed at 24 hpi (Fig. 4C), as described above. As seen for MHC-I, in cells infected with the LucR.6ATRi reporter virus, the magnitude of CD4 down-modulation (Fig. 4C) mediated by Nef was similar to that observed following infection with nonreporter virus. We attribute the small amount of remaining surface CD4 expression to the observation that induced GFP levels in cells infected with the LucR.6ATRi reporter virus lagged slightly behind those of nonreporter virus (NL-BaL.ecto)-infected cells (data not shown), possibly indicating a slightly earlier stage of infection for some reporter virus-infected cells, and thus a lesser degree of non-Nef-mediated (e.g., Vpu-mediated) CD4 down-modulation. By 48 hpi, a difference in GFP levels in infected cells was no longer observed.

Taken together, our findings suggest that the absence of Nef-mediated CD4 and MHC-I down-regulation in cells infected with LucR.T2A reporter viruses may be a consequence of low Nef levels in combination with defective Nef function (i.e., the presumed lack of myristoylation). Importantly, results obtained with the LucR.6ATRi reporter virus indicated that this novel strategy succeeded in overcoming the limitation of LucR.T2A reporter constructs with regard to Nef function while preserving several other important features (e.g., replication capacity, reporter stability, and sensitive detection of infection), which have made the LucR.T2A reporter viruses such valuable tools for a number of applications.

Nef expression and function in A3R5.7 and PBMCs infected with LucR.IRES reporter viruses

To further investigate the utility of the LucR.6ATRi reporter approach, we proceeded with analyzing Nef function after infection of A3R5.7 cells (a representative CCR5⁺ T cell line chosen for its role in HIV-1 NAb neutralization assays) ^{9 –11,17} and PBMCs (as a model of primary HIV-1 target cells). We infected both cell types with a subset of 293T transfection-derived LucR.T2A and LucR.6ATRi reporter viruses with and without functional nef mutations and assessed Nef expression and MHC-I down-modulation at 96 hpi (Figs. 5 and 6). Staining for intracellular HIV-1 p24 allowed us to quantify the number of infected cells by flow cytometry in order to normalize samples for Western blot and to gate on HIV-infected (p24⁺) cells for flow cytometric analysis of MHC-I surface expression.

OPEN IN VIEWER

FIG. 5.

Nef expression and function are preserved in A3R5.7 cells and PBMCs infected with LucR.6ATRi reporter virus. (A–C) Evaluation of expression (A, B) and function (C) of Nef in the T cell line A3R5.7 and PBMCs after infection with 293T transfection-derived (TD) LucR.T2A reporter (NL-LucR.T2A-BaL.ecto), LucR.6ATRi reporter (NL-LucR.6ATRi-BaL.ecto), and nonreporter (NL-BaL.ecto) viruses. Mutations that abrogate Nef expression (Nef^STOP) were incorporated into each background. An abrogated Nef myristoylation mutant (Nef^G2A) was also included in the NL-BaL.ecto background. Unaltered nonmutated (wild-type) Nef is indicated as “Nef.” A3R5.7 cells or PBMCs infected with HIV-1 were identified by intracellular p24 staining in order to normalize the number of infected cells analyzed by Western blot, and for gating on HIV-1-infected (p24⁺) cells for flow cytometric analysis. (A, B) Western blot evaluation of Nef (lower panels, with anti-Nef mAb clone EH1) and LucR (upper panel) levels in A3R5.7 cells (A) or PBMCs (B) 96 hpi with TD LucR.T2A or LucR.6ATRi reporter and nonreporter viruses. For the evaluation of Nef, original and digitally enhanced images are presented in parallel to illustrate significantly reduced (A) or absent (B) Nef levels (see closed arrowheads, ▲). Replica blots were prepared in parallel and probed with anti-p24 (Gag) (middle panel) to demonstrate loading of equivalent numbers of infected cells. (C) Flow cytometric evaluation of Nef function, as measured by MHC-I down-modulation on infected (p24⁺) A3R5.7 cells (left panels) or PBMCs (right panels) 96 hpi. Levels of MHC-I down-regulation for NL-LucR.T2A-BaL.ecto and NL-LucR.6ATRi-BaL.ecto were compared to both nonreporter virus (NL-BaL.ecto), containing wild-type Nef (filled histograms), and either NL-LucR.T2A-BaL.ecto or NL-LucR.6ATRi-BaL.ecto containing the abrogated Nef expression mutation (Nef^STOP) (dashed histogram). Additionally, levels of MHC-I down-regulation for nonreporter virus encoding the Nef^G2A mutation were compared to nonreporter virus containing wild-type Nef (filled histograms) and the Nef^STOP mutation (dashed histogram).

OPEN IN VIEWER

FIG. 6.

Compared to virus-naive cells, MHC-I levels increase or decrease after infection of T cells with viruses harboring abrogated or functional Nef expression, respectively. The level of MHC-I expression (HLA-A/B/C) was analyzed by flow cytometry in T cell lines, J2573-R5 and A3R5.7, and pooled donor PBMCs depleted of CD8⁺ cells with 293T transfection-derived nonreporter (NL-BaL.ecto) and LucR.6ATRi reporter (NL-LucR.6ATRi-BaL.ecto) viruses encoding either unaltered functional Nef (solid histogram) or Nef^STOP (dashed histogram). Cells from virus-naive (mock-infected) cultures were included to provide normal MHC-I levels (filled histograms). A3R5.7 cells or PBMCs productively infected with HIV-1 were identified by intracellular p24 staining at 96 hpi (as in Fig. 5), and productively infected J2574-R5 cells were identified by GFP expression at 48 hpi (as in Fig. 1). Data from a representative experiment are shown.

As seen in J2574-R5-infected cells (Fig. 4B), the levels of Nef in A3R5.7 cells and PBMCs infected with LucR.6ATRi reporter virus (NL-LucR.6ATRi-BaL.ecto) were similar to nonreporter virus (NL-BaL.ecto) (Fig. 5A and B). However, Nef expression from the LucR.T2A reporter virus (NL-LucR.T2A-BaL.ecto) was significantly reduced in A3R5.7 (Fig. 5A, see arrowhead) and undetectable in PBMCs (Fig. 5B, see arrowhead). The consistent lack of detectable Nef expression mediated by the T2A molecular strategy in infected T cells is noteworthy, but further investigation was outside the scope of the current study.

In agreement with Nef expression data, and illustrated in Fig. 5C, we did not observe down-regulation of MHC-I expression on A3R5.7 cells or PBMCs that were infected with LucR.T2A reporter viruses (Fig. 5C, middle panels) or with any of the viruses with abrogated Nef (Nef^STOP) expression. Importantly, in both A3R5.7 cells and PBMCs infected with the LucR.6ATRi reporter virus, MHC-I expression levels were similar to those in cells infected with nonreporter virus encoding wild-type Nef (Fig. 5C), as we had seen in J2574-R5 cells (Fig. 4C).

Interestingly, we observed in PBMCs, as well as in the two T cell lines tested, that infection with “nef-minus” (Nef^STOP) or functionally Nef-defective (Nef^G2A or Nef^KR) viruses reproducibly led to an increase in MHC-I levels compared to “mock-infected” cells in a parallel culture [with the increase in mean fluorescence intensity (MFI) ranging from 1.3- to 1.8-fold]. Infection by cognate viruses expressing fully functional Nef (i.e., NL-BaL.ecto and NL-LucR.6ATRi-BaL.ecto) led to a modest degree of down-regulation of MHC-I compared to mock-infected cells (to a range of 0.56- to 0.75-fold). Thus, the effect of Nef on MHC-I was most apparent when comparing MFI between cells productively infected with virus encoding wild-type Nef versus virus encoding Nef^STOP (Fig. 6). Of note, when we gated on uninfected (i.e., p24⁻ or GFP⁻) cells in a virus-exposed culture harboring productively infected cells, MHC-I levels were elevated compared to virus-naive or “mock-infected” cultures, irrespective of the nef genotype (data not shown).

Comparison of LucR.T2A and LucR.6ATRi reporter viruses in the A3R5/Env-IMC-LucR neutralization assay

We previously developed a large panel of LucR.T2A reporter viruses with env genes from seven clades ¹⁷ (also Jones et al., unpublished observations) that enables a number of immune monitoring approaches (e.g., of vaccine sera ¹¹ ) including the recently formally validated A3R5/Env-IMC-LucR neutralization assay. ^9,10 While this, or other T cell-based-NAb assays, does not require functional Nef, we wanted to ensure that the new LucR.6ATRi reporter virus (also collectively referred to as “Env-IMC-LucR.6ATRi”) performed similarly (“noninferior”) to its counterpart, the LucR.T2A reporter virus (i.e., “Env-IMC-LucR.T2A”). We recently conducted a carefully controlled neutralization tier phenotyping study of over 120 Env-IMC-LucR.T2A viruses (Jones et al., unpublished observations), including NL-LucR.T2A-BaL.ecto, which allowed us to test the performance of NL-LucR.6ATRi-BaL.ecto in parallel. Both viruses were similarly neutralized by the eight different bNAbs tested, according to IC₅₀ and IC₈₀ data (Fig. 7A). Neutralization curves for NL-LucR.T2A-BaL.ecto and NL-LucR.6ATRi-BaL.ecto essentially overlapped. The curve-based (Fig. 7A) and point-based (data not shown) IC₅₀ and IC₈₀ values for each respective bNAb/virus combination were closely matched. Importantly, the fold-difference between NL-LucR.T2A-BaL.ecto and NL-LucR.6ATRi-BaL.ecto IC₅₀ values for each bNAb was very low (Fig. 7B, 2-fold range indicated), and well within the 3-fold range that was previously defined as acceptable for interexperimental variation of IC₅₀ values for the same virus in the A3R5/Env-IMC-LucR neutralization assay. ¹⁰

OPEN IN VIEWER

FIG. 7.

Equivalent performance of Env-IMC-LucR.T2A and Env-IMC-LucR.6ATRi in the A3R5/Env-IMC-LucR neutralization assay. Inhibition of NL-LucR.T2A-BaL.ecto and NL-LucR.6ATRi-BaL.ecto by the IgG1b12 (b12), 2G12, 2F5, 4E10, CH31, CH01, PG9, and PG16 bNAbs was determined on A3R5.7 cells, following the recently validated neutralization assay protocol, and utilizing reduction of LucR activity as a readout to determine 50% and 80% inhibitory concentrations (IC₅₀ and IC₈₀, respectively). (A) Curve-based IC₅₀ and IC₈₀ values obtained for each of the bNAbs against LucR.T2A and LucR.6ATRi reporter viruses. Dashed lines indicate the starting concentrations of each bNAb (5 μg/ml for PG9 and PG16; 25 μg/ml for all others), i.e., maximum achievable IC₅₀ or IC₈₀ concentration for the assay. (B) Fold-difference (ratio) between NL-LucR.T2A-BaL.ecto and NL-LucR.IRES-BaL.ecto (T2A:6ATRi) IC₅₀ (or IC₈₀) values for each bNAb. Dashed lines indicate a 2-fold range of IC₅₀ values (note: a 3-fold range for interassay and intraassay variation was previously demonstrated to be acceptable and indicative of equivalent IC₅₀ values in the validated A3R5/Env-IMC-LucR neutralization assay ¹⁰ ).