Envelope-Dependent Restriction of Human Immunodeficiency Virus Type 1 Spreading in CD4⁺ T Lymphocytes: R5 but Not X4 Viruses Replicate in the Absence of T-Cell Receptor Restimulation

Abstract

The human immunodeficiency virus (HIV) replicates in activated CD4⁺ T lymphocytes. However, only CD4⁺ Th2 and Th0, but not Th1, CD4⁺ T-cell clones have been reported to efficiently support HIV-1 replication. This dichotomous pattern was further investigated in the present study in Th1, Th2, or Th0 cell lines derived from umbilical human cord blood and in T-cell clones obtained from the peripheral blood mononuclear cells (PBMC) of healthy adults. Both primary and laboratory-adapted HIV-1 strains with CCR5 as the exclusive entry coreceptor (R5 viruses) efficiently replicated in Th1, Th2, and Th0 cells. In sharp contrast, CXCR4-dependent (X4) viruses poorly replicated in both polarized and unpolarized CD4⁺ T cells, including adults’ PBMC infected several days after mitogenic stimulation. Unlike the X4 HIV-1_NL4-3, a chimera in which the env gene had been replaced with that of the R5 HIV-1_NL(AD8), efficiently replicated in both Th1 and Th2 cells. This X4-dependent restriction of HIV replication was not explained by either the absence of functional CXCR4 on the cell surface or by the inefficient viral entry and reverse transcription. T-cell receptor stimulation by anti-CD3 monoclonal antibodies fully rescued X4 HIV-1 replication in both Th1 and Th2 cells, whereas it did not alter the extent and kinetics of R5 HIV-1 spreading. Thus, R5 HIVs show a replicative advantage in comparison to X4 viruses in their ability to efficiently propagate among suboptimally activated T lymphocytes, regardless of their polarized or unpolarized functional profiles. This observation may help to explain the absolute predominance of R5 HIVs over X4 viruses observed after viral transmission and during early-stage disease.

CD4⁺ T lymphocytes are the main targets of human immunodeficiency virus (HIV) infection. Although infection of resting T lymphocytes can occur, it results in an abortive infection due to incomplete reverse transcription unless signals leading to T-cell activation intervene within hours or days of infection (30, 55). Among other possibilities, CD4⁺ T lymphocytes may undergo functionally distinct programs of activation and differentiation, such as those defined as Th1/Th2 pathways, in order to support either phagocyte-dependent (cellular) or -independent (humoral) immune responses, respectively (37). In regard to HIV infection, a shift from Th1- to Th2-mediated immune responses has been postulated as a crucial determinant of both the susceptibility of most individuals to infection and of the progression of disease (15) by analogy with the dichotomous roles played by polarized immune responses in parasite infections (48). This hypothesis has stimulated investigations on the susceptibility of polarized (i.e., Th1 versus Th2) CD4⁺ T cells to HIV infection. Of interest, a restricted pattern of HIV replication was observed in that Th2 and Th0, but not Th1, cell clones were found to be permissive for HIV replication (32). In these early studies, however, HIV was directly propagated by cocultivation of T-cell clones with irradiated peripheral blood mononuclear cells (PBMC) of HIV-infected individuals, without further characterization of the viral isolates. In this regard, it has been recently established that HIV strains differ in terms of their usage of chemokine receptors for entry into CD4⁺ cells (31). The two main viral coreceptors utilized by HIV-1 are CCR5 for viruses formerly known as macrophage-tropic or non-syncytium-inducing (NSI; now commonly referred to as R5 HIVs) and CXCR4 for T-lymphotropic, SI viruses (X4 HIVs) (4). R5 HIVs predominate soon after transmission and throughout the asymptomatic stage of disease, whereas X4 viruses emerge in approximately 50% of individuals infected by clade-B HIV-1 before the onset of AIDS (16, 44). The occurrence of a shift from NSI to SI is a determinant of accelerated disease progression independent from viremia (29). CC chemokine receptors other than CCR5, including CCR2b and CCR3, are more rarely used by SI viruses in association with CXCR4 (17, 24).

Interestingly, differential and even selective expression of chemokine receptors has been recently associated with Th1 and Th2 cells. Th1 cells preferentially express CCR5 and CXCR3, whereas Th2 cells are characterized by CCR3, CCR4, and CCR8 surface expression (8, 41, 42). CXCR4 is usually reported to be present on both Th1 and Th2 lymphocytes, although its expression is inducible by the Th2-associated cytokine interleukin-4 (IL-4) (22, 26, 52). Conversely, the Th1-associated cytokines IL-12 and gamma interferon (IFN-γ) have been reported to upregulate CCR5 and downregulate CXCR4 (22).

In the present study, we have investigated the replicative capacity of R5 and X4 laboratory-adapted and primary HIV-1 strains in Th1, Th2, and Th0 cells obtained either from umbilical cord blood leukocytes (CB lines) differentiated under polarizing conditions or from T-cell clones derived from adults’ PBMC. R5 HIVs efficiently replicated, whereas X4 viruses did not in either Th1, Th2, or Th0 cells. These patterns were not attributable to inefficient entry of X4 viruses into polarized primary T lymphocytes that potently replicated in these cells upon restimulation by anti-CD3 monoclonal antibodies (MAbs).

MATERIALS AND METHODS

CB-derived Th1, Th2, and Th0 lines and primary T cell clones.

Human neonatal leukocytes were isolated from umbilical cord blood by use of a Ficoll-Hypaque (Pharmacia Biotech, Uppsala, Sweden) density gradient. Approximately 50% of these cells were CD4⁺ cells with a naive (CD45 RA⁺) phenotype, as reported previously (38). Th1 and Th2 differentiation was obtained by a stimulation of these cells with phytohemagglutinin (PHA; Wellcome, Beckenham, United Kingdom) and a combination of either anti-IL-4 Ab (PharMingen, San Diego, Calif.) plus IL-12 (Hoffman-La Roche, Inc., Nutley, N.J.) or anti-IL-12 antibody (gift of M. Gately, Hoffman-La Roche, Inc.) plus IL-4 (PharMingen) for 72 h, respectively, as described previously (38). Cell lines were washed and maintained at approximately 5 × 10⁵ cells/ml in RPMI 1640, 5% fetal calf serum, antibiotics, and 100 U of recombinant IL-2 (Hoffman-La Roche) per ml for between 15 and 21 days before HIV infection. These cells were fully differentiated along either Th1 or Th2 pathways, as confirmed by their selective patterns of cytokine expression upon restimulation. In some experiments, CB cells were stimulated by PHA in the absence of polarizing condition, giving origin to “neutral” cells expressing both IFN-γ and IL-4. The cells were maintained in IL-2-enriched medium before and after infection in the absence of restimulation unless otherwise specified.

In some experiments, Th1 or Th2 cell clones were obtained by limiting dilution of PBMC cultivated on irradiated allogenic PBMC, PHA, and IL-2 as described earlier (38). T-cell clones selectively expressing IFN-γ, but not IL-4, and vice versa, defined by cytofluorimetric analysis as specified below, were considered Th1 and Th2 clones, respectively, whereas coexpression of both cytokines was an indication of a Th0/neutral phenotype (37).

Phenotypic analysis of CB lines and clones for cell surface expression of CD4 (anti-CD4; Becton Dickinson, San Jose, Calif.), CXCR4 (with the 12G5 MAb from R&D Systems, Minneapolis, Minn.), and CD27 (M48 MAb; a kind gift of C. A. Smith, Immunex, Research and Development Corp., Seattle, Wash.) was performed before and at different time points after infection. Isotypic control antibody was purchased from Caltag Laboratories (Burlingame, Calif.). Cell staining was performed according to standard procedures (5); samples were acquired and analyzed with a FACScan apparatus (Becton Dickinson).

Analysis of cytokine production at the single-cell level.

Single-cell analysis of IFN-γ and IL-4 production was performed as previously described (40). Briefly, T-cell lines were collected 7 to 10 days after priming and washed, and 10⁶ cells were restimulated with phorbol myristate acetate (50 ng/ml; Sigma Chemical Corp., St. Louis, Mo.) and ionomycin (1 μg/ml) (Sigma) for 2 h at 37°C in complete medium. Brefeldin A (10 μg/ml) (Sigma) was added to the cultures, which were incubated for an additional 2 h. The cells were fixed with 2% paraformaldehyde and permeabilized with saponin. Fixed cells were stained with anti-huIFN-γ–fluorescein isothiocyanate (Pharmingen), anti-huIL-4–phycoerithrin (Pharmingen), and anti-huCD4–Quantum Red (Sigma), or anti-huCD8–Quantum Red (Sigma) after a protocol provided by the manufacturer. Samples were analyzed by use of a FACScan.

Chemotaxis.

Th1 and Th2 migration in response to recombinant stromal cell derived factor-1α (SDF-1α) (R&D Systems), a ligand of CXCR4 (3, 7, 36), was evaluated in a microchamber chemotactic assay as previously described (38). Briefly, the lower compartment of the chamber was filled with medium containing different concentrations of SDF-1α (ranging from 0.03 to 3 μg/ml), whereas Th1 or Th2 cells were seeded in the upper compartment at the concentration of 10⁶ cells/ml. The two compartments were separated by use of a 5-μm-pore-size polycarbonate filter (Nucleopore, Cabin John, Md.) and incubated for 90 min at 37°C in a humidified atmosphere containing 5% CO₂. After incubation, the filters were carefully removed, and the cells that migrated to the side of the filters facing the lower compartment of the chamber were stained with toluidine blue and counted by optical microscopy.

RT-PCR.

Constitutive expression of the mRNAs for CXCR4, SDF-1, and IL-8 and for the housekeeping glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was determined by reverse transcriptase PCR (RT-PCR) analysis. Total RNA was extracted by the RNAzol method (Duotech, Milan, Italy). First, 500 ng of RNA was reverse transcribed in the presence of 1× RT buffer (GIBCO-BRL, Verviers, Belgium), an 800 mM concentration of each deoxynucleoside triphosphate (Pharmacia Biotech), 20 μg random examers (Promega, Madison, Wis.) per ml, 4 mM dithiothreitol (GIBCO-BRL), 16 U of RNA Guard (Promega), and 400 U of Moloney murine leukemia virus RT (GIBCO-BRL). The reaction mixture (50 μl) was incubated at 68°C for 5 min and then at 37°C for 60 min; it was next heated at 94°C for 5 min and then cooled on ice. Aliquots corresponding to 1/40 of the cDNA obtained were amplified in the presence of 0.4 mM primer pairs (PRIMM S.p.A., Milan, Italy), 200 mM concentrations of the deoxynucleoside triphosphates (Pharmacia Biotech), 1× PCR buffer, and 1.25 U of AmpliTaq Gold DNA polymerase (Perkin-Elmer Corp., Norwalk, Conn.) in a 50-μl reaction mixture. The mRNA of GAPDH was used as a control. The following primers were used for amplifying mRNAs: SDF-1 (kindly provided by S. Polo, San Raffaele Scientific Institute, Milano, Italy), sense 5′-ACG AAT TCG CGC CAT GAA CGC CAA GGT CGT-3′ and antisense 5′-CAG GAT CCT GCA AAC CTC AGG CCC GAT C-3′) (generating a 451-bp fragment) (2); CXCR4 (5), sense 5′-GCC AAC GTC AGT GAG GCA GAT G-3′ and antisense 5′-GAG GAT GAC TGT GGT CTT GAG G-3′ (a 209-bp fragment); IL-8 (49), sense 5′-GAT TTC TGC AGC TCT GTG TG-3′ and antisense 5′-ACA GAG CTC TCT TCC ATC AG-3′ (a 191-bp fragment); and GAPDH, sense 5′-CCA TGG AGA AGG CTG GGG-3′ and antisense 5′-CAA AGT TGT CAT GGA TGA CC-3′ (a 195-bp fragment). PCR products were analyzed by electrophoresis in a 2% agarose gel and then visualized by ethidium bromide staining.

HIV strains and chemokine coreceptor usage.

The HIV-1_BaL strain, a prototype for R5 macrophage-tropic viruses, and HIV-1_LAI/IIIB, a prototype of X4 viruses, were used. Three primary HIV-1 isolates (HIV-1_BON, HIV-1_BER, and HIV-1_CAR) were obtained from the cocultivation of PBMC of HIV-infected individuals with PHA-stimulated allogeneic PBMC of seronegative donors, without further passages. All of these viruses were characterized for their ability to replicate and induce syncytia in the MT-2 cell line (with HIV-1_LAI/IIIB and HIV-1_BON testing positive, i.e., SI viruses, and HIV-1_{BaL, BER, CAR} testing negative, i.e., NSI viruses, respectively), and for their ability to use selected chemokine receptors (including CXCR4, CCR2b, CCR3, and CCR5) together with CD4 in stably transfected U87 cells (kindly provided by Dan Littman, The Skirball Institute, N.Y.) (25). As expected, HIV-1_LAI/IIIB and HIV-1_BON used CXCR4, whereas HIV-1_BaL, HIV-1_BER, and HIV-1_CAR used CCR5 but not other tested coreceptors (R5 HIVs). Virus replication in MT-2 and transfected U87 cell lines and in cultures of Th1, Th2, and Th0 cells was tested by supernatant associated Mg²⁺-dependent RT activity (5).

HIV DNA quantitation by TaqMan.

The kinetics and levels of HIV DNA accumulation in Th1 and Th2 cells were determined by the TaqMan assay with an ABI 7700 Prism instrument, as recently described (2, 5). Briefly, DNase-treated HIV-1_LAI/IIIB or HIV-1_BaL stocks were normalized for a multiplicity of infection (MOI) of 0.4 and were incubated with 10⁶ Th1 or Th2 cells for 30 min at 37°C in 5% CO₂. Cells were then washed extensively and seeded in 24-well flat-bottomed tissue culture plastic plates (Falcon; Becton Dickinson Labware, Lincoln Park, N.J.). Cell aliquots corresponding to approximately 10⁶ cells were harvested at different time points and centrifuged at 12,000 rpm for 5 min, and the pellets were lysed in a proteinase K-containing buffer. The DNA was extracted by using phenol-chloroform and then ethanol precipitated; one-tenth of the DNA preparations (corresponding approximately to 10⁵ cells) was amplified with the following primers derived from the p24^gag region; their positions in the HIV-1_HXB2 clone (23) are indicated in parentheses: forward, 5′-ACA TCA AGC AGC CAT GCA AAT-3′ (positions 1368 to 1388); reverse, 5′-ATC TGG CCT GGT GCA ATA GG-3′ (positions 1472 to 1453); and probe FAM 5′-CAT CAA TGA GGA AGC TGCAGA ATG GGA TAG A-3′ (TAMRA) (positions 1401 to 1431). The thermal cycling conditions were 50°C for 2 min, 95°C for 12 min, 40 cycles of 95°C for 15 s, and 65°C for 1 min. The DNA extracted from serially diluted chronically infected ACH-2 T cells containing one copy of proviral DNA per cell (5) was used as an external standard. A linear distribution (r = 0.99) was obtained between 2 and 31,250 ACH-2 cells.

Infections.

Th1 or Th2 CB lines (5 × 10⁵ cells/ml) were infected 15 to 21 days after the initial activation by mitogens and polarizing agents with X4 and R5 viruses at an MOI of 0.4 in 48-well tissue culture plastic plates (Falcon) in the presence of complete medium containing 100 U of IL-2 (Hoffman-La Roche) per ml. The same experimental conditions were applied for infection of T-cell clones. Ca. 50% of the medium was removed every 2 to 3 days and stored at −20°C until tested for RT activity, and an equivalent volume of complete medium was added to the cultures. The kinetics of viral replication were measured by determining the RT activity on the supernatants collected and stored at −80°C. In some experiments, Th1 and Th2 CB lines were cultured in flat-bottomed 96-well plates coated with mitogenic (1 μg/ml) concentrations of the TR66 anti-CD3 MAb (kindly provided by P. Dellabona, San Raffaele Scientific Institute) or the control isotype antibody. HIV infection was done within 1 h of cell seeding.

RESULTS

R5 but not X4 HIV-1 strains efficiently replicate in Th1 and Th2 CB lines and primary T-cell clones.

CB lines were first stimulated with PHA and IL-2 under polarizing (IL-12 plus anti-IL-4 MAbs versus IL-4 plus anti-IL-12 MAbs for Th1 and Th2, respectively) conditions at least 15 days before HIV infection, as described previously (38). The resulting CB lines were either Th1 (IFN-γ⁺ IL-4⁻) or Th2 (IFN-γ⁻ IL-4⁺), respectively, as assessed by fluorescence-activated cell sorter (FACS) analysis of selective intracellular cytokine expression after cell restimulation (data not shown).

Th1 and Th2 CB lines, as well as polarized T-cell clones, were infected with the laboratory-adapted R5 HIV-1_BaL or the X4 HIV-1_LAI/IIIB strains and monitored for up to 25 days for virus production and cytopathicity. Efficient replication of HIV-1_BaL was observed without substantial differences either in the kinetics or in the peak levels of viral replication between CB lines and T-cell clones and regardless of their Th1 or Th2 phenotypes (Fig. 1). In sharp contrast, HIV-1_LAI/IIIB did not replicate or poorly replicated in both CB lines and T-cell clones without distinction between Th1 and Th2 cells (Fig. 1).

FIG. 1.

R5 but not X4 laboratory-adapted HIV-1 strains replicate in Th1 and Th2 CB lines and T-cell clones. All cells were infected with equal amounts of RT activities of laboratory-adapted BaL (R5) (●) and LAI/IIIB (X4) (○). Ten independent experiments were performed, with similar results.

In order to verify whether this dichotomous pattern could have been dependent on the use of laboratory-adapted viruses, similar experiments were performed with primary isolates obtained from infected individuals and characterized by MT-2 tropism and chemokine coreceptor usage. The patterns observed were fully supportive of those described above for the laboratory-adapted strains in that R5, but not the X4 viruses, efficiently propagated in CB lines and T-cell clones without substantial differences in terms of Th1 versus Th2 cells (Fig. 2).

FIG. 2.

R5 but not X4 primary HIV-1 isolates replicate in Th1 and Th2 CB lines and T-cell clones. HIV isolates were obtained from infected individuals and characterized for MT-2 cell tropism and chemokine receptor usage, avoiding further in vitro passages. This experiment is representative of four independently performed. ●, BER (R5); ○, BON (X4).

Cell surface expression of different markers including CD4, CXCR4, and CD27 (a molecule belonging to the TNF receptor superfamily [6]) was monitored at different time points after infection of both Th1 and Th2 cells with HIV-1_LAI/IIIB or HIV-1_BaL by MAb staining and cytofluorimetric analysis. None of these molecules was found to be substantially altered with respect to the levels of expression observed in parallel uninfected cells (Table 1 and data not shown). However, increased levels of expression on a per-cell basis (mean fluorescence intensity) of CXCR4 were noted in Th2 versus Th1 cells, in agreement with recent reports (references 22 and 26 and data not shown).

TABLE 1.

Expression of CXCR4 on Th1 and Th2 CB lines before and after infection by X4 or R5 HIV

Time postinfection (h)	% Cells expressing CXCR4a
	Th1			Th2
	Nil	X4	R5	Nil	X4	R5
6	26	18	20	21	25	18
48	25	18	18	23	23	24
144	17	17	13	24	23	22

^aCells were stained with the anti-CXCR4 12.G5 MAb (R&D Systems) according to standard procedures (5). Samples were acquired and analyzed with a FACScan apparatus (Becton Dickinson). Nil, uninfected CB lines. 

Th0 cell clones and neutral CB lines support the replication of R5 HIV but not of X4 viruses.

A number of independent Th0-cell clones and neutral CB lines stimulated and maintained for at least 15 days in IL-2-enriched medium in the absence of polarizing conditions were infected by R5 or X4 HIV strains. As observed with Th1 and Th2 cells, HIV-1_BaL efficiently replicated in Th0 cell clones (Fig. 3). Unlike HIV-1_BaL, neither HIV-1_LAI/IIIB nor the primary X4 HIV-1_BON isolate replicated in unpolarized, neutral CB lines (Fig. 3).

FIG. 3.

R5 but not X4 HIVs replicate in unpolarized T cells. (A) Three independent cell clones from a single donor were infected by equal amounts of virus. Solid symbols, BAL (R5); open symbols, LAI/IIIB (X4). (B) CB cells were stimulated by PHA and IL-2 without Th1 or Th2 polarizing stimuli and maintained in IL-2-enriched medium. When restimulated with PMA and ionomycin, these cells coexpressed both IFN-γ and IL-4. These cells were infected with the laboratory-adapted strains LAI/IIIB (○) and BaL (●) and with the primary X4 BON virus (□).

Thus, R5 HIVs possess a greater replicative potential compared to X4 viruses both in CB lines and T-cell clones regardless of their polarized or unpolarized profile of cytokine expression, according to the Th1 and Th2 differentiation pathways.

The R5 versus X4 HIV replication dichotomy emerges in long-term cell cultures.

In order to verify whether the observed patterns of permissive versus restricted replication of HIV were influenced by the use of neonatal cells (CB lines) or were the consequence of T-cell cloning, total PBMC were obtained from normal volunteers. These cells were immediately stimulated by PHA and infected by an equivalent MOI (MOI = 0.4) of HIV-1_LAI/IIIB (X4) or HIV-1_BaL (R5) viruses in IL-2-containing medium after 3 (standard PHA blasts), 7, or 11 days from mitogenic stimulation. Both X4 and R5 HIVs productively infected blasts at 3 days with similar levels of RT activity production at peak infection, which indeed occurred earlier with HIV-1_LAI/IIIB (day 14) compared to HIV-1_BaL (day 21) (Fig. 4). Productive replication of both viruses was also observed with the cells infected 7 days after mitogenic stimulation, whereas a dichotomous pattern of R5 permissiveness versus X4 restriction was observed in blasts maintained in IL-2-enriched medium for 11 days prior to infection. Of note, R5 viruses replicated in these long-term lymphocyte cultures with an efficiency comparable to and even superior to that observed in blasts at 3 days in terms of the kinetics of infection (Fig. 4).

FIG. 4.

Long-term culture of PHA-activated PBMC (PHA blasts) results in loss of X4 virus replication and persistent R5 production. PBMC were freshly isolated from an individual and stimulated with PHA. At 3 days after stimulation, cells were washed and resuspended in an IL-2-enriched medium. Cells were immediately infected with either LAI/IIIB or BaL or were left uninfected in culture for an additional 4 or 8 days, respectively, at which time points they were infected with the same MOI of both viruses.

Therefore, neither CB cells, T-cell cloning, nor polarizing conditions are critical factors in the emergence of the dichotomous R5 versus X4 infection profiles observed with primary CD4⁺ T lymphocytes.

X4 HIV capacity of productively infecting long-term cultured T lymphocytes is Env restricted.

Sequence variations involving the hypervariable V3 region of gp120 Env are responsible for the interaction with both CD4 and the chemokine receptors (31). Therefore, experiments were conducted with viruses derived from infectious molecular clones in which the 3′ portion of the HIV-1_NL4-3 genome, including env, was derived from the X4 HIV-1_LAI/IIIB, whereas the 5′ portion was HIV-1_NY5 (1). In particular, a macrophage-tropic clone of HIV-1_NL4-3 (HIV-1_NL(AD8), kindly donated by Eric Freed, National Institutes of Health) was generated by replacing the 1.7-kb fragment containing env (encompassing gp120 env and part of gp41 env) of HIV-1_LAI/IIIB with the homologous portion of the macrophage-tropic, R5 HIV-1_ADA strain (18, 21). HIV-1_NL(AD8) but not HIV-1_NL4-3 efficiently replicated in both Th1 and Th2 lymphocytes (Fig. 5). Therefore, the inability to replicate typically observed with X4 HIVs is probably dependent upon the recognition of different chemokine entry coreceptors (CXCR4 versus CCR5).

FIG. 5.

A virus chimeric for R5 Env acquires the ability to efficiently replicate in Th1 or Th2 cells. Th1 and Th2 CB lines were infected with comparable amounts of NL4-3 (X4) (Th1 [○] and Th2 [□] or its chimera NL(AD8) (Th1 [●] and Th2 [▴]). Five independent experiments provided similar results.

The restriction of X4 HIV in T cells is not explained by either the lack of functional CXCR4 or by the expression of endogenous SDF-1.

In order to verify whether the X4 restricted patterns could be explained simply by the loss of CXCR4 expression in long-term cultures, both Th1 and Th2 CB lines or T-cell clones were monitored by FACS analysis at different points during their differentiation, both in the presence and in the absence of X4 and R5 HIVs (Table 1). Comparable levels of CXCR4-expressing cells were observed among differentiated Th1 and Th2 CB lines, which were not altered by an ongoing infection with either X4 or R5 HIVs (Table 1). Similar conclusions were drawn in terms of the density of CD4 and CXCR4 on a per-cell basis at different times throughout the culture, although Th2 cells frequently expressed higher levels of this chemokine receptor, as reported previously (22, 26; data not shown).

We next investigated whether Th1 or Th2 CB lines expressed SDF-1, which may potentially interfere with binding of X4 virus to the chemokine receptor. However, no evidence of expression of this chemokine was obtained by RT-PCR (Fig. 6A), whereas both CXCR4 and the CXC chemokine IL-8 were found comparatively expressed in Th1 and Th2 CB lines (Fig. 6A).

FIG. 6.

Expression of functional CXCR4 molecules on Th1 and Th2 CB lines. (A) Comparable expressions of CXCR4 and IL-8, but lack of SDF-1 expression, were observed in CB lines of Th1 or Th2 phenotype. (B) Chemotaxis of Th1 and Th2 cells to exogenous SDF-1. Both types of cells show a comparable concentration-dependent response to the chemokine.

Finally, both Th1 and Th2 cells comparably migrated in response to exogenously added SDF-1α in a concentration-dependent manner (Fig. 6B), thus indicating that both cell types expressed functional CXCR4 molecules on their cell surface.

In conclusion, the inability of X4 viruses to efficiently propagate in Th1 and Th2 cells was not explained by the lack of surface expression of CXCR4.

Comparable entry and reverse transcription of R5 and X4 HIVs in Th1 and Th2 cells.

The availability of functional CXCR4 on the cell surface of both Th1 and Th2 cells did not support the hypothesis that the lack of X4 HIV replication could be ascribed to an impaired viral entry. In order to investigate whether the dichotomous pattern described here was related to a differential ability of R5 versus X4 HIVs to enter and reverse transcribe their genomes into DNA, we compared the abilities of X4 and R5 HIVs to infect Th1 and Th2 cells by quantitative real-time PCR in a TaqMan assay (5). In particular, the determination of the number of HIV DNA copies synthesized within 24 h provides a correlate of a single round of HIV replication and, therefore, it could indicate whether X4 virus entry was impaired. However, similar numbers of R5 and X4 HIV DNA copies were observed early and at up to 72 h after infection (Fig. 7), implying that entry and reverse transcription were accomplished by both types of viruses with comparable efficiency. In contrast, a 10- to 100-fold-superior accumulation of R5 versus X4 DNA was observed after 72 h of infection (Fig. 7), supporting the divergent accumulation of RT activity in supernatants of X4-infected versus R5-infected cell cultures.

FIG. 7.

Lack of X4 HIV spreading among Th1 and Th2 CB lines. Comparable levels of HIV DNA from X4 and R5 HIV are retrotranscribed in both cell types up to 72 h postinfection, after which only R5 viruses efficiently propagated in the cell cultures.

These results, together with the presence of CXCR4 on the surface of Th1 and Th2 T cells, strongly indicate that the restriction of X4 HIV in these cells is likely the consequence of a deficient viral spreading rather than a blockade of viral entry.

Selective activation of X4 HIV-1 replication in Th1 and Th2 cells by restimulation with anti-CD3 MAb.

The observation that both Th1 and Th2 cells were readily infected by both R5 and X4 viruses and that persistent levels of HIV DNA were demonstrable up to 6 days after infection suggested that X4 viruses could latently infect these cells. Therefore, we investigated whether activation of HIV replication could be obtained by cell stimulation with mitogenic concentrations of anti-CD3 MAb, as previously reported in other systems (9, 10, 12, 46). Anti-CD3 MAb did not substantially modify the replication pattern of R5 HIV-1, although a modest acceleration of the kinetics of infection was observed in both Th1 and Th2 cells (Fig. 8, upper panels). In sharp contrast, a robust increase of HIV replication was observed in X4-infected cells stimulated with anti-CD3 MAb compared to unstimulated cultures. This effect was specific in that it was not observed in the presence of a control isotype MAb (not shown). A kinetic analysis of anti-CD3 MAb stimulation revealed that enhancement of X4 HIV replication was inducible for up to 8 days after infection (not shown), a finding compatible with a state of latent infection in the absence of T-cell activation.

FIG. 8.

Anti-CD3 MAb reactivation of X4 virus replication. CB lines were infected and were either left unstimulated (○) or were stimulated with mitogenic concentrations of anti-CD3 mAb (●). No substantial effects were observed upon infection by R5 viruses, whereas a potent upregulation of viral replication occurred in X4-infected cells.

DISCUSSION

In the present study we have observed that R5, but not X4, HIVs efficiently replicate in Th1 and Th2 differentiated T lymphocytes. Similar patterns were observed in unpolarized T cells, including Th0 cell clones, neutral CB lines, and PHA blasts infected after long-term cultivation. The lack of X4 HIV replication was not sustained by the absence of CXCR4 from the cell surface or by endogenous expression of its ligand SDF-1. Indeed, infection by both R5 and X4 HIVs in Th1 and Th2 cells occurred with comparable efficiency, as determined by the quantitation of HIV DNA copies, for up to 72 h postinfection, at which time the two infection patterns diverged and only R5 viruses efficiently propagated throughout the culture. Restimulation of Th1 and Th2 cells with anti-CD3 MAb, however, strongly induced X4 HIV replication, whereas it minimally affected R5 virus production.

The strict dependency of HIV replication from T-cell activation, obtained either via T-cell receptor or through cytokine stimulation, has been substantiated in several model systems (9, 10, 12, 27, 46). With particular regard to the polarized pathways of differentiation and activation of CD4⁺ T lymphocytes, a preferential replication of HIV in Th0 and Th2 versus Th1 cell clones was noted in one early study (32) and was subsequently confirmed by at least some investigators (19), although no information on the mechanism of viral restriction was provided. It should be underscored, however, that the original observation was obtained by infecting polarized T-cell clones with uncharacterized viruses, i.e., by means of cocultivation with irradiated PBMC of HIV infected individuals in advanced stages of disease (32).

The recent discovery of the fine mechanism of HIV infection by utilization of different chemokine receptors as entry cofactors together with CD4 has provided new opportunities to investigate host-virus interactions. In this regard, several investigators have recently studied the susceptibility of both T cells, including Th1 and Th2 lymphocytes, and macrophages to infection by different strains of HIV in relation to their usage of chemokine receptors. Among these studies, CXCR4 was shown to be upregulated by IL-4 and downregulated by the Th1-associated cytokines IL-12 and IFN-γ (22, 26, 52). Recently, IL-12 and IL-4 have been shown to favor R5 and X4 HIV replication in activated PBMC, respectively (50). Previously, a superior capacity of macrophage-tropic viruses to replicate in CD4⁺ T cells, irrespective of their Th1 versus Th2 phenotypes, was noted (51). However, this study did not address whether the described patterns of viral replication were either due to the acute effects of cytokine stimulation or to an intrinsic functional divergence between Th1 and Th2 cells. Other investigators have not observed substantial differences between Th1 versus Th2 T lymphocytes in their ability to sustain HIV replication (33, 34). Of note is the fact that in all of these studies, HIV infection was initiated only after a few days of cell stimulation by either PHA and/or cytokines, again raising the question of the respective roles of acute cell activation versus the state of polarization of T lymphocytes on viral replication. For these reasons, we decided to investigate the replicative capacities of Th1, Th2, and Th0/neutral cells after at least 15 days from PHA stimulation in the presence or absence of polarizing conditions. Striking differences were consistently observed with both laboratory-adapted and primary HIVs, in that only HIVs using CCR5 but not those dependent upon CXCR4 efficiently replicated in Th1, Th2, or Th0 cells.

Our results thus suggest that the acquisition of CXCR4 usage imposes restrictions on the ability of HIV to spread in suboptimally stimulated CD4⁺ T lymphocytes in comparison to R5 viruses. In contrast to the latter, X4 HIVs failed to replicate in all types of primary T lymphocytes investigated, including Th1, Th2, and Th0/neutral CB lines, T-cell clones, and long-term cultivated PHA blasts. In this respect, we did not observe substantial differences in terms of CXCR4 expression among fully differentiated Th1 and Th2 CB lines at the time of infection. In addition, the comparable levels of X4 versus R5 HIV DNA accumulation observed in Th1 and Th2 CB lines indicates that entry and reverse transcription of X4 viruses, like that of R5 HIVs, was not substantially different between Th1 and Th2 cells. This interpretation is also supported by the absence of endogenous SDF-1 expression and by the comparable chemotactic response to this chemokine of both Th1 and Th2 cells. X4 HIVs lose their spreading ability as a function of the time interval from mitogenic stimulation and establish a latent form of infection. Cell restimulation, as here obtained by using anti-CD3 MAb, however, powerfully activated X4 HIV spreading in these CD4⁺ T lymphocytes. Of interest, this pattern of Env-dependent restriction occurring at a post-entry level has been previously associated with macrophage infection by T-cell-line-adapted X4 HIV (45) or SIV_mac239 (35), as well as by testing the infectivity of different simian immunodeficiency virus isolates in MAGI cell lines expressing different chemokine coreceptors (11).

The observed patterns of in vitro viral replication resemble events occurring during the natural history of HIV infection in most individuals, in whom R5 HIVs almost invariably predominate after viral transmission and during the asymptomatic stage of disease, whereas X4 viruses emerge in more advanced stages of the disease. The observation that R5 viruses require less-stringent conditions of T-cell activation, together with their ability to replicate in macrophages (47), provides a potential correlation of their superior capacity to engraft after viral transmission compared to X4 HIVs. The observation that individuals homozygous for the Δ32 deletion in the CCR5 gene are virtually protected by HIV infection (39, 43) emphasizes the importance of the viral usage of this coreceptor for infecting most individuals. In addition, a relative defect of CXCR4 but not of CCR5 has been reported in dendritic cells obtained from the cervical mucosa, suggesting an important contribution of these cells in virus selection during sexual transmission (56).

The propensity of X4 viruses to latently infect suboptimally activated CD4⁺ T cells acquires relevance in light of the demonstrated existence in vivo of a pool of latently infected resting memory cells which is established soon after primary HIV infection (14, 20, 54). Of interest, reactivation of HIV replication in these cells has been accomplished by stimulation with either cytokines or the anti-CD3 MAb (13); in our study this last condition was capable of triggering X4 HIV replication in both Th1 and Th2 cells. Furthermore, enhancement of X4 HIV replication by anti-CD3 MAb has been independently demonstrated in activated PBMC (10), whereas in vivo administration of this antibody has resulted in increased levels of viremia (9).

The nature of the observed X4 restriction in our primary T-lymphocyte cultures remains elusive, although it appears associated with the selective ability of the HIV-1 Env to interact with either CXCR4 or CCR5. Having ruled out a defective entry and/or reverse transcription as the most direct correlates of the observed dichotomous patterns of infection, the most likely explanation for this is the triggering of a differential signaling cascade by R5 versus X4 Env molecules, ultimately resulting either in a latent (X4) or in a productive (R5) form of infection. In support of this hypothesis, only trimeric R5 Env complexes have been shown to activate Ca²⁺ and chemotaxis of T lymphocytes, whereas X4 complexes did not (53). Of note is the fact that Ca²⁺ was previously shown to either trigger or potentiate HIV expression (28). Studies are in progress in order to verify whether this or other signaling cascades are involved in the differential abilities of R5 and X4 viruses to replicate in CD4⁺ T lymphocytes.