|Home | About | Journals | Submit | Contact Us | Français|
Super-paramagnetic CD44 MicroBeads (Miltenyi) designed for the isolation of infectious HIV-1 from dilute or difficult biological samples dramatically enhance the infectivity of bound HIV virions, even if the original viral suspension is merely incubated with beads. Infection of the CEM T cell line with the NL4-3 virus clone or primary human CD4 T cells with X4- and R5-tropic clones and a clade C primary virus isolate all showed accelerated p24 production and larger fractions of infected target cells. Effects could be detected very early; incubation of virus with the CD44 MicroBeads promoted higher levels of viral integration within the first infection cycle. In summary, CD44 MicroBeads provide the means not only to concentrate dilute viral samples, but also to directly facilitate within days rather than weeks the in vitro expansion of patient isolates independent of coreceptor usage and the performance of HIV replication assays that require a large fraction of infected primary T cells.
Primary human CD4 T cells are notoriously inefficient targets for HIV-1 infection, making in vitro studies of rare or early infection events difficult to perform even in optimized in vitro culture systems. In addition, mutant or reporter variants of cloned HIV usually have reduced infectivity relative to the parental viral strain. Various strategies have been employed to overcome these difficulties such as pseudotyping virus for more efficient cell entry (Bartz and Vodicka, 1997), concentrating virus by ultracentrifugation, and infection by “spinoculation” (Forestell, 1996). Each of these methods has its drawbacks. The commonly used envelope protein for pseudotyping, VSV-G, utilizes acidic endosomal vesicles for virus internalization and uncoating, which bypasses the normal engagement of CD4 and coreceptor molecules and any associated signaling events (Aiken, 1997). Viral preparations can be concentrated by high speed centrifugation at the risk of destabilizing the oligomeric structure of HIV-1 envelope (Earl et al, 1990), or sedimented onto target cells with “spinoculation”; however, even infections initiated with very high multiplicity of infection (MOI) usually infect at best 30 – 40 % of primary T cells (Bartz and Vodicka, 1997, Noraz et al, 1997). Such high backgrounds of uninfected cells makes quantitative protein and molecular assays especially problematic.
A recently released product designed originally for the purification and concentration of infectious HIV-1 from limited amounts of dilute or difficult clinical specimens may provide a more physiological way to improve infection efficiency without altering viral genes or proteins. Based on a superparamagnetic microbead conjugated to an antibody recognizing CD44, the Miltenyi HIV VitalVirus reagent binds to CD44 molecules that are incorporated into the viral envelope as it buds from an infected or transfected cell (Tremblay et al, 1998). Although not all tissues express the CD44H isoform recognized by the CD44 MicroBeads, hematopoetic cell lineages (Dalchau et al, 1980, Flanagan et al, 1989, Lesley et al, 1993) as well as some cancer cells (Liu and Jiang, 2006) do express high levels of this isoform. CD44 functions as the hyaluronic acid receptor and has been associated with adhesion (Shimizu et al, 1989), activation signal transduction (Huet et al, 1989, Shimizu et al, 1989, Ponta et al, 2003, Hegde et al, 2008), lymphocyte homing (Berg et al, 1989, de la Hera et al, 1989, Ponta et al, 2003), and T cell maturation (Marquez et al, 1995, Patel et al, 1995, Ponta et al, 2003). Following MicroBead binding, virions remain attached for up to several days in culture even after manipulations on magnetic purification columns. Because some preliminary work indicated that HIV-Microbead mixtures produced enhanced infection of peripheral blood mononuclear cells (PBMC) separate from concentration effects (Miltenyi, personal communication), we decided to investigate this phenomenon in more detail using primary human CD4 T cells.
To be able to scale up the standard VitalVirus HIV Isolation Kit protocol (sample size 0.2–1 ml) and purify HIV from larger volumes of culture supernate, we sought initially to optimize reagent conditions for the ratio of beads to viral preparation volume and mixture incubation times. Sample volumes of 2 ml and 15 ml were tested at two bead ratios and incubation times. Controls for the incubation of virus without further manipulation were included for two bead ratios (2 ml samples only). Table 1 shows the range of viral infectivity and p24 protein from three experiments using the NL-EGFP clone, a GFP reporter variant of the X4 laboratory strain NL4-3, produced in the CEM T cell line. Unconcentrated samples, U1 and U2, were analyzed once with the 2 ml sample volume. Treatments of the smaller volume were the most effective for recovering p24 antigen as measured by ELISA and infectious units assayed on the reporter HeLa cell line P4R5. The best viral protein recovery obtained was approximately 60% of the input; this was most likely due to the presence of non-virion associated p24 in the original virus preparation. In two of the three assays, all viral samples incubated with CD44 MicroBeads yielded a higher infectious titer than could be accounted for by volume reduction alone during concentration. Such results implied that virus bound to CD44 MicroBeads had increased infectivity or accelerated sedimentation onto the adherent reporter cell targets.
To measure the effect of the MicroBeads on T cell infection independent of virus titer, a fixed number of CEM cells were infected with 0.05 infectious units (P4R5 cell infectivity assay) of NL-EGFP virus in a constant final culture volume. Figure 1 depicts a representative experiment in which the fraction of infected cells, as estimated by GFP expression, increased more rapidly if the virus preparation was exposed to MicroBeads under any condition including incubation without subsequent washing or concentration (Fig 1A). It should be noted that reporter virus expression of GFP underestimates the actual number of infected cells when compared to intracellular staining of p24 expression. All GFP positive cells express high levels of intracellular p24 (ICp24), but not all p24 positive cells express detectable GFP (data not shown). Secreted p24 production showed accelerated kinetics and higher peak levels if virus preparations had been exposed to MicroBeads, regardless of further purification (Fig 1B). Infection was so efficient in CEM cells with CD44 MicroBead treatment of virus that almost no viable cells were left to analyze after 7 days of infection. Primary CD4 cell cultures infected with Microbead-bound virus also displayed accelerated infection kinetics by GFP expression and p24 secretion with bead treated virus in three independent experiments (Fig 1C and D), although peak GFP expression appeared to plateau below 50% in the primary cells. Results similar to those with fresh viral preparations were also obtained when frozen and thawed viral aliquots were used for infection of CEM and primary CD4 cells; and the outcomes were not altered whether the MicroBead incubation occurred before or after the virus preparation was frozen (data not shown). Based on the results obtained from these initial experiments, all subsequent CD44 MicroBead interactions were performed with frozen/thawed virus preparations at the lowest ratio of 20 ul beads per ml virus, one hour prior to cell infection.
We next examined the ability of CD44 MicroBeads to enhance in vitro infections with both X4- and R5-tropic HIV in primary CD4 T cells. Standard results achieved by infection with the X4 clone, NL4-3, are shown in Figure 2A and B. At the peak of infection on day 7 of culture, approximately 30% of cells expressed ICp24 (Fig 2A, left) and 106 pg/ml of soluble p24 was produced (Fig 2B, left). In comparison, pretreatment of NL4-3 with CD44 MicroBeads increased the fraction of infected cells to an average of 80%, and did so 3 days sooner (Fig 2A, right). The proportion of cells expressing p24 declined thereafter, probably due to virus induced cytotoxicity. The kinetics of soluble p24 production was accelerated as well, reaching a maximal level by Day 4, equal to that of untreated virus infection at Day 7 (Fig 2B, right). To confirm that the enhancement effect of CD44 MicroBead treatment was independent of coreceptor tropism, preparations of the R5 HIV clone JR-CSF were used to infect primary CD4 cells. Without MicroBead treatment, JR-CSF infected up to 5.5% of cells (Fig 2C); however, 10 to 30% of cells showed ICp24 expression following infection with the MicroBead treated virus preparation. Levels of secreted p24 antigen increased earlier and reached a higher plateau following infection with treated virus (Fig 2E), similar to results seen with X4 NL-EGFP infection. As reported by other investigators (Blanco et al, 2001, Grivel et al, 2000, Lawson et al, 2004), infection with the R5-tropic clone of JR-CSF resulted in less cell death than infection with the X4-tropic clone of NL4-3. The JR-CSF infected cell cultures were maintained for 14 days, at which point cells with integrated HIV DNA and replication competent virus (infectious units per million cells, IUPM) were assayed. MicroBead treatment increased the number of integrated HIV-1 DNA copies from 5 to 35 fold (Fig 3A), reaching up to a million copies/200 ng total DNA (30,000 cell equivalents) in some cases, and increased IUPM from 3 to 105 fold (Fig 3B).
To test the general utility of the MicroBead reagent for expanding clinical virus isolates, infections were performed using a clade C, X4 primary isolate, 98IN017. An original aliquot of virus from the NIH AIDS Research and Reference Reagent Program had been expanded in PBMC culture and stored frozen at a relatively low titer of approximately 100,000 pg p24/ml. Infections with virus alone yielded variable results in cells from 3 different donors, with maximum infected cell fractions reaching less than 2% (Fig 2D) and soluble p24 production levels of 2200 to 8400 pg/ml after 14 days of culture (Fig 2F). Virus aliquots pretreated with MicroBeads exhibited improved infection capacity reaching maximal target cell infection rates of 3 to 13.5% (Fig 2D) and soluble p24 production of 25,000 to 56,000 pg/ml (Fig 2F) at Day 10. Consistent with the higher viral protein production, the number of integrated HIV copies and IUPM assayed at Day 14 also increased with virus incubation with MicroBeads. Untreated viral infections produced 4,400 to 50,200 integrants per 200 ng input DNA, while infections with CD44 MicroBead preincubated virus yielded 28,500 to 96,800 copies (Fig 3A). The frequencies of replication competent infected cells were more modest, increasing 2.5 or 5 fold with bead treatment (Fig 3B).
In addition to improvements in infection, we observed that resting primary CD4 cells exposed to CD44 MicroBead treated virus tended to form microscopically visible clusters that remained up to a week in culture. Infected cells that were stimulated with αCD3 and αCD28, however, formed blasts and did not maintain as much cell-to-cell contact. To ascertain whether the faster infection kinetics observed were due to a more efficient initial infection or more efficient viral spread through enhanced cell-to-cell contact (Dimitrov et al, 1993), we performed quantitative HIV integrant analysis 24 hours after exposing cells to virus. MicroBead treatment of virus increased the amount of detectable HIV integration (copies per 200 ng genomic DNA) from 1.5 fold to as much as 90 fold (Fig 4). In some cases, even cells that had not been stimulated with αCD3 and αCD28 antibodies showed more HIV integration, if the virus preparation had been incubated with MicroBeads prior to infection. This suggested that MicroBeads bound to virions could affect early infection events, either through direct interaction with the CD44 molecules on cellular membranes or more rapid sedimentation onto the targeted T cells.
To address whether centrifugation steps during post infection cell washes were creating a “spinoculation” effect, CEM cell aliquots were exposed to NL-EGFP for 4 hrs in a standard limited volume, and then diluted with excess medium. Subsequent supernatant sampling for p24 production was performed in a manner that did not subject the continuing culture to centrifugal forces. The fraction of cells exhibiting productive infection by GFP expression and the levels of p24 secretion were at least as high as those seen in earlier experiments with MicroBead treated virus (data not shown). Thus, while virion – MicroBead complexes may settle more quickly onto cells than virus alone, the added g forces from centrifugation during cell washes did not appear to contribute to enhanced infection.
The cell clustering seen in both uninfected and infected cultures exposed to MicroBeads and the increased amounts of virus integration found in unstimulated cells exposed to MicroBead treated virus suggested a direct interaction between target CD4 T cells and CD44 MicroBeads. To separate any cell signaling events induced by viral proteins from potential MicroBead induced effects, NL4-3 preparations were generated to preclude virion binding to the MicroBeads. Virus was produced by plasmid transfections into either HeLa P4R5 cells expressing CD44H (hematopoietic isoform) or 293T cells, which do not bear CD44H and cannot bind the specific αCD44 antibody conjugated to the MicroBead reagent (clone DB105-2G12.1; data not shown). Because the infectious titer of HIV (IU/pg p24) produced in 293T cells is about 6 fold lower than that of virus produced in CEM or P4R5 cells (data not shown), we attempted to compensate for this factor in our experimental design. Therefore, the primary CD4 cell infections were performed using a relatively large final volume for each of the virus stocks, resulting in much lower cell and virus concentrations than our standard infection procedure with CEM or primary lymphocytes. As expected, virus produced from CD44H expressing cells and treated with the MicroBead reagent generated infections with markedly accelerated intracellular and secreted p24 production relative to parallel infections with untreated virus or MicroBead treated virus produced in 293T cells lacking CD44H (Fig 5A and B). MicroBead pretreatment of virus produced in CEM T cells resulted in infections with approximately 20 fold more infected cells (22 to 48% ICp24 positive) and secreted p24 (280,000 to 780,000 pg/ml). Infections with virus produced in P4R5 cells were less robust in general: however, CD44 MicroBead treatment caused an unexpectedly large magnitude of enhancement, with over 100 fold increase in infected cells (31 to 84% ICp24 positive) three days earlier (Day 4) than peak infection with untreated virus. In contrast, virus produced in 293T cells showed only a 2 fold change in infection efficiency with CD44 MicroBead pretreatment. Such results indicate that physical complexing of CD44 MicroBeads with virus is required for the observed enhancement of HIV infection. Despite the lower infectivity of 293T generated virus (as determined in a P4R5 cell assay), the untreated 293T virus preparations induced better infection (4 to 10% ICp24 positive cells and 71,000 to 166,000 pg/ml secreted p24) than did either the untreated P4R5 or CEM generated virus preparations under these diluted inoculum conditions.
To determine whether contact between CD44 MicroBeads alone and target T cells can induce cellular changes, flow cytometric phenotype analysis was performed on Days 1, 4, and 7 following infection to measure expression of surface activation markers, CD25 (IL2 receptor α chain) and CD69, a very early marker of T cell receptor induced activation (Testi et al, 1989). Despite high levels of CD44 on the surface of T cells, modulation of cell membrane activation markers after exposure to CD44 MicroBeads was detected in only a minor subpopulation of the resting CD4 target cells. The largest alteration of CD25 and CD69 expression was seen on Day 1 after MicroBead treatment (Fig 5C and D). CD25 expression in cell controls without bead treatment ranged from 0.6% to approximately 6% on Day 1; exposure to MicroBeads resulted in slight increases of 1.3 to 2.5 fold, the magnitude of which varied between cell donors, but not with the presence of virus (Fig 5C). CD69 expression was detected on a maximum of 3% of unstimulated CD4 T cells, one day following exposure to virus mixed with MicroBeads (Fig 5D). CD69 expression increased 3 to 13 fold after 1 day of culture from 0.1 to 0.8% in untreated cells and 1.0 to 2.6% in cells exposed to beads alone or in combination with virus. If data from both the uninfected and infected cell samples were pooled, bead treatment induced significant changes in both CD25 and CD69 expression at this early time point (p = 0.0005, Wilcoxon Matched Pairs Signed Rank Test). In contrast, stimulation with combined TCR/CD28 cross-linking caused virtually all cells to express CD69 after one day, and CD25 by Day 4. No additional effects of MicroBead treatment could be detected in this setting (data not shown).
To further investigate the mechanism(s) of enhanced infection by CD44 MicroBeads, we compared infections achieved in CEM cells by NL-EGFP virus preparations that were pretreated with one of the following: CD44 MicroBeads, free αCD44 antibody, biotinylated αCD44 antibody complexed with recombinant streptavidin, and CD45 MicroBeads. Only the multivalent reagents (biotinylated αCD44 + streptavidin and CD45 MicroBeads) enhanced infection relative to the positive control of CD44 MicroBead treatment, although with 1–2 day delayed kinetics (Fig 6A and B). When tested in the single-cycle viral integrant assay, virus treated with CD44 MicroBeads yielded the greatest number of integrants followed by CD45 MicroBeads and biotinylated-streptavidin complexed αCD44 (Table 2). Use of free αCD44 treatment produced similar levels of integrants as virus alone (< the 5 copy detection limit). FACS analysis of CD25 expression in unstimulated cells, one day after exposure to αCD44 or αCD45 reagents alone, virus alone, or virus preincubated with αCD44 or αCD45, revealed little difference from uninfected, untreated cells. However, incubation with CD45 MicroBeads resulted in a mild increase in CD69 expression, from less than 1% in untreated cells to about 5% following CD45 bead exposure in both uninfected and infected cells (Fig 6C). None of these observed changes in CD25 or CD69 expression were significant (p = 0.1 by Wilcoxon Matched Pair Ranked Sign Test).
Using a T cell line and primary CD4 lymphocytes, this study confirms and expands upon a prior anecdotal observation of increased HIV infectivity with virus preparations treated by the Miltenyi VitalVirus HIV Isolation Kit. Specifically, we demonstrate accelerated infection kinetics, increased peak levels of infected cells and virus production, independence of virus clade and coreceptor tropism, and the requirement for CD44H antigen inclusion in the viral envelope.
We have endeavored to understand how CD44 MicroBead treatment improves HIV-1 infection of T cells by using comparisons with non-bead associated αCD44 antibody reagents and MicroBeads conjugated to αCD45 antibody (CD45 MicroBeads). Our results demonstrate that maximal enhancement of infection occurs when MicroBeads are conjugated to the αCD44H monoclonal antibody. Surprisingly, CD45 MicroBeads appeared just as effective in accelerating HIV infection as multivalent αCD44 streptavidin complexes and almost as effective as CD44 MicroBeads, even though it is known that CD45 molecules are excluded from budding virions (Tremblay et al, 1998, Nguyen et al, 2003, Chertova et al, 2006). The significance of these findings is difficult to assess, however, given the 5 fold range of antibody concentrations among the CD44 reagents tested, and the 10 to 50 fold higher antibody concentration of the CD45 MicroBead reagent. The increased multivalency of αCD44 antibody, present on the surface of MicroBeads and in the streptavidin-αCD44-biotin, is correlated with the effectiveness of HIV infection enhancement. Thus, for reagents containing αCD44, it is possible to envision that complexing of αCD44 with HIV, as well as delivery of virions to target T cells, is more efficient when virus particles are associated with more αCD44 antibody, but not necessarily with a paramagnetic particle. In the case CD45 MicroBead effects, it is well known that primary T lymphocytes express high levels of membrane CD45 (Schwinzer, 1989, Thomas, 1989), and modulation of CD45 protein tyrosine phosphatase activity (Clark and Ledbetter, 1989) can negatively or positively affect signaling pathways dependant on the receptor or endpoint assayed (Thomas, 1995). Elucidation of the mechanism(s) involved in CD45 MicroBead interaction with T cells will require additional work that is outside the scope of our current research.
Another experimental finding, which begs further explanation, is the increased expression of cell surface activation molecules seen in a small fraction of unstimulated cells following exposure to complexed αCD44 antibody, CD44 MicroBeads or CD45 MicroBeads. Because CD44 and CD45 are both highly expressed on primary T cells, it is possible that exposure to these reagents induces receptor cross-linking and initiates a minimal cell activation event. It is important to note that any changes in functional status of cells following interaction with these MicroBead reagents are dependent primarily on cell surface expression of antigen and the degree of cross-linking induced by the cognate monoclonal antibody complexed to the MicroBeads, and not to factors inherent to the MicroBeads themselves (Grützkau and Radbruch, 2008; Schmitz, 2008). In addition, the interaction of multivalent antibodies with cell surface antigens during prolonged periods at 37°C (as performed here) is likely to cause aggregation and internalization of membrane receptors, resulting in modulation of some gene expression. Never the less, none of the complexed αCD44 or αCD45 reagents induced changes in CD25 or CD69 expression that were significantly greater than those observed in untreated cell cultures or during infections with virus preparations unable to complex with CD44 MicroBeads (non-CD44H bearing virions). Therefore, any contribution of CD44 Microbeads to enhancement of viral infection through direct effects on cellular CD44 function would appear to be minor.
A more likely scenario to explain the effect of MicroBeads on HIV infection is a model in which CD44 MicroBead binding simultaneously to both virions and target cells contributes to the increase in apparent viral infectivity (Figure 7). Incubation of CD44 MicroBeads with virions that contain cellular CD44H in their envelope allows direct virus binding to beads. Although MicroBeads are 50 nm in diameter and do not sediment in aqueous solution, CD44 MicroBeads attached to HIV in multimeric complexes may display very different characteristics. The added mass of the MicroBead along with the highly multivalent character of CD44 MicroBead-virion interaction may help sediment virus onto target cells so effectively over the incubation period that subsequent centrifugations to remove unbound virus do not create any additional spinoculation effect. As a second contributing factor, αCD44 antibodies conjugated to the MicroBeads could physically bridge virus to target cells through engagement of unoccupied antibody binding sites with CD44H antigen expressed on the cell membrane. Given the high affinity of gp120 binding to the CD4 receptor (4–8 nM for HXB2 and YU2 gp120; Stricher et el, 2008), and the CCR5 coreceptor for R5 virus (JR-FL; Doranz et al, 1999), the ~1 nM affinity of typical antibody-antigen interactions (O’Shannessy et al, 1994, Katsamba et al, 2006) is not likely to provide any enhancement of initial contact between virions and cell surface molecules. However, due to the large number of antibody molecules conjugated to CD44 MicroBeads, interaction with the highly expressed CD44 molecule on target cells could be expected to generate high avidity binding and stabilize virus-cell contact. αCD44 mediated tethering may prolong CD4-gp120 engagement, allow coreceptor reorganization and binding, and facilitate the coupling of other envelope embedded cellular proteins with their CD4 cell counterparts, some of which are known to increase virus infectivity (Arthur et al, 1992, Arthur et al, 1995, Chan et al, 1995, Cantin et al, 2005). While virus associated ICAM-1 binding to cellular LFA-1 is an example of a high affinity interaction enhancing infectivity 10 fold (Fortin et al, 1997, Rizzuto and Sodroski, 1997) or even 100 fold (Fortin et al, 1998), the relatively weak interaction between HLA-DR in virion envelope and cellular CD4 can also promote virus entry, resulting in two-fold faster infection kinetics (Cantin et al, 1997). CD44 MicroBead interactions with HIV may create similar types of phenomena.
In addition, CD44 engagement by certain antibodies is known to induce adhesion and activation signal transduction (Huet et al, 1989, Shimizu et al, 1989, Ponta et al, 2003). In the work reported here, we observed that cells exposed to CD44 MicroBeads (or multimeric αCD44) formed small clusters independent of the presence of virus in the complex. Such clustering would be expected to facilitate far more efficient viral transmission through direct cell-to-cell contact, in contrast to infection achieved with soluble virus (Dimitrov et al, 1993). The clustering effect induced by MicroBeads may be the direct result of CD44’s adhesion function, or an indirect effect of signal transduction by CD44 activation of other adhesion molecules, such as LFA-1. It has been reported that binding of RANTES to CD44 can lead to p44/p42 MAPK signaling and improved HIV infection (Roscic-Mrkic et al, 2003), and a similar mechanism may play a role in our observations. While not statistically significant, the increase in both viral integration and CD69 expression following cell exposure to viral preparations with CD44 or CD45 MicroBeads suggests that engagement of CD44 and CD45 on cell membranes may cause an effect that could, at least partially, explain observed increases in viral infectivity.
The ability to accelerate infection of CD4 T cells has practical utility for both the production of virus stocks and the improvement of viral assay systems. Shortening the culture time necessary to reach high viral titers decreases the consumption of tissue culture reagents during virus expansion. Production of high titer virus preparations can facilitate infection of a larger proportion of primary cell targets and enable individual cell analysis, as has already been reported for infection with HIV-1 concentrated by CD44 Microbeads (Mitchell et al, 2008). Most importantly, application of CD44 MicroBeads to the in vitro expansion of primary clinical virus isolates may allow for genomic and phenotypic characterizations that would otherwise not be possible. Additionally, tissue culture based assays that measure viral products may be enhanced by the inclusion of CD44 MicroBeads, which can shorten the time necessary to cross a detection threshold, or potentially increase the sensitivity of an assay by acting as a biological signal amplifier.
We have demonstrated improved HIV infection of primary CD4 T cells with the use of CD44 MicroBeads. How well these MicroBead reagents can perform to enhance infection of macrophage or microglial cells remains to be determined. The amount of CD44 expressed on macrophages could be as high as that on circulating lymphocytes (Haynes et al, 1989, Koller et al, 1996, Lee et al, 2001), but the adherent growth properties of macrophages would preclude MicroBead formation of cell clusters for increased cell-to-cell transmission. Still, the ability of virion-bound MicroBeads to settle onto cells more rapidly and interact with cell membrane CD44 molecules may provide a noticeable improvement to current infection methods. In conclusion, CD44 MicroBeads provide a practical means to enhance HIV infection simply by binding to CD44H-containing virions, in addition to their ability to purify and concentrate infectious virus from solution.
293T cells (Dubridge et al, 1987) were a kind gift from Dr. Nathaniel Landau and maintained in Eagle’s MEM with 2X vitamins and nonessential amino acids (Quality Biologicals, Gaithersburg, MD) supplemented with 2 mM glutamine (Invitrogen, Carlsbad, CA) and 10% fetal bovine serum (FBS, Gemini Bio-Products, West Sacramento, CA). P4R5 MAGI cells were obtained from the NIH AIDS Research and Reference Reagent Program (Germantown, MD); they are HeLa P4 cells stably transduced to express human CD4 and CCR5, and with β-galactosidase under transcriptional control of the HIV-1 LTR (Charneau et al, 1994). They were maintained in Dulbecco’s MEM supplemented with 2 mM glutamine, and 100 U/ml penicillin/100 ug/ml streptomycin (Invitrogen, Carlsbad, CA), and 10% FBS (Gemini Bio-Products, West Sacramento, CA) (D10) with 1 ug/ml puromycin (Calbiochem, Gibbstown, NJ). CEM cells (American Type Culture Collection, Manassas, VA) were cultured in RPMI 1640 (Invitrogen, Carlsbad, CA) with 10% FBS from Cambrex (Walkersville, MD) (R10).
RosetteSep Human CD4+ T Cell Enrichment Kit was purchased from StemCell Technologies (Vancouver, BC, Canada). Prescreened, pooled human AB serum for primary CD4 T cell culture was purchased from Omega Scientific (Tarzana, CA). Dulbecco’s Phosphate Buffered Saline without calcium or magnesium (PBS) was from Mediatech (Herndon, VA). Formaldehyde for flow cytometry samples was purchased as a 10% solution from Polysciences of Warrington, PA. Potassium ferricyanide was from Mallinkrodt Baker (Phillipsburg, NJ) and 5-bromo-4-chloro-3-indolyl β-D-galactopyranoside (X-gal) was acquired from Apex Bioresearch Products (San Diego, CA). Recombinant human IL-2 (Lahm and Stein, 1985) (rIL-2) was obtained from the NIH AIDS Research and Reference Reagent Program. Recombinant human IL-15 (rIL-15) and human interferon β1a (IFNβ) were purchased from R&D Systems (Minneapolis, MN). HIV VitalVirus kits containing CD44 MicroBeads, buffers and magnetic columns, biotinylated αCD44 complexed with streptavidin, free αCD44, and CD45 MicroBeads were generous gifts from Miltenyi Biotec, Bergisch Gladbach, Germany. FuGene6 lipid mediated transfection reagent was from Roche Applied Science (Indianapolis, IN). Sigma Chemical Company (St. Louis, MO) was the source for Histopaque-1077 and all other chemicals.
Goat anti-mouse IgG + M (H + L) was from Kirkegaard & Perry Laboratories (Gaithersburg, MD). Anti-CD3, αCD28, APC-αCD25, and APC-Cy7-αCD69 were purchased from BD Biosciences (San Jose, CA). PE-αHIV-1 p24/gag clone KC-57 was obtained from Beckman Coulter (Carlsbad, CA).
Blood from healthy donors was collected by venipuncture into polypropylene syringes containing sodium heparin according to institutional review board approved protocols. Buffy coats were incubated with 1/10 volume of RosetteSep CD4 T cell enrichment cocktail at room temperature for 20 min, diluted with one volume of PBS containing 2% human AB serum (2% HAB/PBS), then CD4 T cells were separated from the remainder of PBMC and red cell rosettes by density gradient centrifugation. CD4 cells were washed twice with 2% HAB/PBS and resuspended in RPMI 1640 medium supplemented with 2 mM glutamine, 100 U/ml penicillin/100 ug/ml streptomycin, and 5% human AB serum (R5HAB). Cells were infected overnight, or held overnight in R5HAB growth medium prior to a 6 hr infection the following day.
Viruses X4 viruses NL4-3 and NL-EGFP (a kind gift from Dr. Naoki Yamamoto, Tokyo Medical & Dental University), were grown from starter viral stocks in CEM cells for 13 to 17 days. At harvest, cells and large debris were removed by centrifugation and viral aliquots were stored at −80°C. Alternatively, plasmids pNL4-3 (Adachi et al, 1986, pNL-EGFP (Miura et al, 2003), and pJR-CSF (Koyanagi et al, 1987) were transfected into 293T cells with FuGene6, according to the manufacturer’s protocol. Cell-free supernatants were harvested two days later. X4 tropic viruses were expanded in CEM cells while PHA-stimulated healthy control PBMC were infected to grow JR-CSF R5 viral stocks. The clade C X4 primary patient isolate 98IN017 (NIH AIDS Research and Reference Reagent Program) was cultured initially in PHA activated healthy control PBMC then expanded with heterologous healthy control CD4 T cells. For comparison of CD44H–lacking and -containing virus preparations, P4R5 cells were transfected in parallel with 293T cells and supernatants filtered through a 0.45 um pore before storage at −80°C. For all cultures, secreted p24 was measured by ELISA (Alliance HIV-1 p24 kit, Perkin Elmer, Waltham, MA) according to manufacturer’s instructions. Infectious virus titer was measured by limiting dilution in CEM cells and TCID50 calculated by the Reed-Muench Accumulative Method (Reed and Muench, 1938) or by the P4R5 MAGI LTR β-galactosidase, blue cell reporter cell assay for infectious units (Day et al, 2006).
CD44 MicroBead pretreatment involved combining fresh or thawed virus preparations with CD44 MicroBeads at specified ratios and times at 4°C with gentle rotation or rocking. Treated virus was then either added directly to cells or loaded onto Miltenyi magnetic columns and allowed to enter columns by gravity flow. Bound virus was washed and eluted according to manufacturer’s instructions into 0.5 or 1 ml R10 medium to achieve desired sample volume reduction. Aliquots were stored at −80°C; p24 and infectious titer were assayed by ELISA or dilution onto P4R5 cells, respectively.
CEM infections were performed by resuspension of pelleted cells in a small fixed volume of virus solution in R10 medium. GFP expression levels of the NL-EGFP reporter virus were monitored by FACS analysis; secreted p24 was assayed in culture supernatant by ELISA.
Primary CD4 T cells were incubated with viral preparations either overnight or for 6 hr in the single infection cycle assay, then washed 4 times with 2% HAB/PBS before resuspension in R5HAB growth medium. Where noted, mock infections were performed with CD44 MicroBeads which had been incubated for 1 hr at 4°C in growth medium alone, then added to cells for 6 hr. Unstimulated cells were kept in 14 ml polypropylene round bottom tubes in R5HAB medium. Stimulated cells were cultured in microplates with flat-bottom wells coated with αCD3 and αCD28 followed by the addition of cytokines as described previously (Spina et al, 1997, Spina et al, 2000). Cell and supernatant aliquots were removed for FACS analysis and p24 ELISA at specified days or upon addition of fresh medium. Integrated HIV DNA analysis and IUPM titer were performed on Day 14, where noted.
Infections with free or biotin-streptavidin complexed αCD44 (same clone as in MicroBead reagent) or CD45 MicroBead virus treatment were performed as described above. Mock infections were also performed with αCD44 or αCD45 reagents alone. Final concentrations of CD44 or CD45 antibody were 0.2, 0.8, 1 and 10 ug/ml for CD44 MicroBeads, biotin-streptavidin complexed αCD44, free αCD44 and CD45 MicroBeads, respectively.
Aliquots of 5×105 cells were taken from culture, washed with PBS and surface markers were stained with APC-Cy7 labeled (3 ul) or APC labeled (5 ul) antibodies in 200 ul of PBS containing 2% FBS and 0.1% sodium azide (FACS Staining Buffer) for 30 min at 4°C protected from light. Cells were washed twice with 1 ml FACS Staining Buffer and fixed with 0.5% formaldehyde in PBS. GFP quantification was performed on washed cells, which were immediately fixed. Intracellular p24 analysis was performed by a modification of the method of Bartz and Vodicka (1997). Briefly, cells were treated at 4°C with BD CytoFix/CytoPerm for 20 min, washed twice with BD Perm Wash solution, then stained with 50 ul of 1/100 dilution of PE conjugated αp24 for 30 min at 4°C. Cells were washed and resuspended in 0.5% formaldehyde in PBS. Multiparamter data were acquired on a BD FACSCanto and analyzed with FACSDiva software (BD Biosciences).
Isolated primary CD4 cells were held overnight in medium, infected for 6 hr, washed, and stimulated overnight. Cells were harvested for FACS analysis of surface activation markers CD25 and CD69 and genomic DNA prepared with the Qiagen Blood and Cell Culture DNA Mini kit according to kit instructions (Qiagen, Valencia, CA). This kit isolates 20 – 150 kb DNA fragments, obviating the need to remove partially reverse transcribed or unintegrated circular viral cDNA prior to integrant analysis. DNA was quantified by UV spectrophotometry at 260 nm with correction for absorbance at 320 nm. Up to 200 ng of infected cell genomic DNA were assayed by TaqMan real-time PCR for HIV-1 NL4-3 gag with the following primers (Integrated DNA Technologies, Coralville, IA): forward 5′-AAAAGAGACCATCAATGAGGAAGC – 3′, reverse 5′ – TGGTGCAATAGGCCCTGC – 3′, probe 5′-6-FAM- CAGAATGGGATAGATTGCATCCAGTGCA -BHQ-1 - 3′. For quantification of clade C X4 isolate 98IN017 and R5 clone JR-CSF integrants, pol primer sequences (Rousseau et al, 2004) forward pol-15 5′ – TACAGTGCAGGGGAAAGAATA – 3′, reverse pol-4 5′ – CTGCCCCTTCACCTTTCC – 3′, and probe Polp 1 5′ - 6-FAM-TTTCGGGTTTATTACAGGGACAGCAG-BHQ-1 - 3′ were used instead. Duplicate 25 ul PCR reactions containing TaqMan Universal Mastermix (Applied Biosystems, Inc., Foster City, IA), 900 nM each forward and reverse primer, and 200 nM probe were incubated at 50°C × 2 min and 95°C × 10 min, followed by 45 cycles of 95°C × 15 sec and 60°C × 1 min on an ABI Prism 7700 Sequence Detector (Applied Biosystems, Inc.). Sequence Detector software was used to calculate Ct values, and copy number standard curves were derived from dilutions of NcoI linearized pNL4-3 plasmid in 200 ng uninfected CD4 T cell genomic DNA. The detection limit for each assay was defined as the lowest amount of standard that could be amplified in both reactions with Ct values distinct from the next highest concentration.
Replication competent infectious titer was measured with a modification of the AIDS Clinical Trials Group Lab Manual HIV Quantitative PBMC culture protocol revised version 2.0 May 2004 (Lathey et al, 1994) based on maximum likelihood estimation of infectious unit concentration (Myers et al, 1994). Day 14 infected cells were washed twice with 2% HAB/PBS, counted, and plated in duplicate wells of 96-well plates coated with αCD3 and αCD28 antibodies. A total of 105 cells were plated neat and at five serial 1:5 dilutions with autologous uninfected Day 14 cells in a final volume of 200 ul per well. The dilution series was also plated in uncoated wells (unstimulated) to control for carry over of free virions. After 7 days, conditioned medium was taken and assayed for soluble p24 with the Perkin Elmer Alliance HIV-1 p24 ELISA kit. Positive wells were defined as having at least 50 pg/ml p24 above input. Outcomes were compared to a table of maximum likelihood estimates and goodness of fit P values received from Don Brambilla of New England Research Institutes, Inc. Results were normalized to infectious units per million and valid assays had P values greater than 0.2.
Prism version 4 (GraphPad Software, Inc., San Diego, CA) was used for statistical analysis. Wilcoxon Two Sample Test was used to determine effect of virus on FACS expression data before pooling CD44 positive and negative viruses or all samples for CD44 MicroBead effect on cells. FACS and HIV integrant data were tested by Wilcoxon Matched Pairs Signed Rank test for significance at p < 0.05 of bead or reagent treatment.
This work was supported by funding from a VA Merit Review award (CAS), and the services provided by the Flow Cytometry (Judy Nordberg, Neal Sekiya), Genomics (Steffney Rought), and Viral Pathogenesis (Nancy Keating, Sherry Rostami, Deya Collier) Cores of the San Diego Center for AIDS Research (AI036214), and the UCSD General Clinical Research Center (National Center for Research Resources, M01RR 000827). The following reagents were obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: P4R5 MAGI cells (Dr. Nathaniel Landau), human rIL-2 (Dr. Maurice Gately, Hoffmann-La Roche, Inc.), pNL4-3 (Dr. Malcolm Martin), HIV-1 98IN017 (Cat # 5441, Dr. Robert Bollinger and the UNAIDS Network for HIV Isolation and Characterization). We sincerely thank Miltenyi Biotec for the generous provision of reagents, and Drs. John Guatelli, Mary Lewinski and Davey Smith for their insightful discussions.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.