|Home | About | Journals | Submit | Contact Us | Français|
CXCL12/stromal cell-derived factor 1 is a member of the CXC family of chemokines that plays an important role in hematopoiesis and signals through CXCR4 and CXCR7. Two splice variants of human CXCL12 (CXCL12α and CXCL12β) induce chemotaxis of CXCR4+ cells and inhibit X4 infection. Recent studies described four other novel splice variants of human CXCL12; however, their antiviral activities were not investigated. We constructed and expressed all of the CXCL12 splice variants in Escherichia coli. Recombinant proteins were purified through a His affinity column, and their biological properties were analyzed. All six CXCL12 variants induced chemotaxis of CXCR4+ and CXCR7+ cell lines. Enhancement of survival and replating capacity of human hematopoietic progenitor cells were observed with CXCL12α, CXCL12β, and CXCL12 but not with the other variants. CXCL12γ showed the greatest antiviral activity in X4 inhibition assays and the weakest chemotaxis activity through CXCR4. The order of potency in X4 inhibition assays was as follows: CXCL12γ > CXCL12β > CXCL12α > CXCL12θ > CXCL12 > CXCL12δ. The order of anti-human immunodeficiency virus (HIV) activity was associated with the number of BBXB motifs present in each variant; the most potent inhibitor was CXCL12γ, with five BBXB domains. The results suggest that the different C termini of CXCL12 variants may contain important molecular determinants for the observed differences in antiviral effects and other biological functions. These studies implicate CXCL12γ as a potent HIV-1 entry inhibitor with significantly reduced chemotaxis activity and small or absent effects on progenitor cell survival or replating capacity, providing important insight into the structure-function relationships of CXCL12.
Chemokines are small, structurally related chemoattractant cytokines, characterized by conserved cysteine residues. Based on the positions of the first N-terminal cysteines, chemokines fall into four subfamilies. The CC and CXC subfamilies have been well characterized. The CC subfamily includes RANTES (for “regulated on activation, normal T-cell expressed, and secreted”), MCP-1 (monocyte chemoattractant protein 1), and MIP-1β (macrophage inflammatory protein 1β). The prototype of the CXC subfamily is interleukin-8 (IL-8). The C chemokine (lymphotactine) and the CX3C chemokine (fractalkine) subfamilies have been identified more recently (reviewed in reference 28). Chemokines signal through G-protein-coupled seven-transmembrane-domain receptors and are primarily involved in immunosurveillance, activation, and recruitment of specific cell populations during disease (reviewed in reference 28).
Stromal cell-derived factor 1 (SDF-1)/CXCL12 is a member of the CXC chemokine family and is a key regulator of B-cell lymphopoiesis, hematopoietic stem cell mobilization, and leukocyte migration (reviewed in reference 28). CXCL12 has also been shown to block human immunodeficiency virus type 1 (HIV-1) infection (reviewed in reference 38). CXCL12 was originally thought to mediate these processes through the single receptor CXCR4 (6). However, later studies demonstrated that RDC-1/CXCR7 is also a receptor for CXCL12 (5, 10).
There are two known human splice variants of CXCL12: CXCL12α and CXCL12β (36). The genomic structure of the CXCL12 gene reveals that human CXCL12α and CXCL12β are encoded by a single gene and arise by alternative splicing. The cDNAs corresponding to CXCL12α and CXCL12β encode proteins of 89 and 93 amino acids, respectively. A third splice variant, classified as CXCL12γ, has been identified in rats (19). The human equivalent of CXCL12γ has recently been identified among other splice variants of CXCL12 (41). The novel human splice variants CXCL12γ, CXCL12, CXCL12δ, and CXCL12θ (also reported as CXCL12 ) are expressed through alternative splicing events that result in different exons being added to the same first three exons. Therefore, the six splice variants of CXCL12 are identical in the first 88 amino acid residues from the amino terminus (Fig. (Fig.11).
It has been demonstrated that CXCL12α and CXCL12β are expressed in numerous tissues, with the highest expression in liver, pancreas, and spleen (41). The mRNA encoding CXCL12γ was detectable in adult human heart tissue and weakly in several other tissues, while CXCL12δ, CXCL12, and CXCL12θ were detected in several human adult and fetal tissues, with the pancreas expressing the highest levels (41). Many questions regarding the biological activities of these novel CXCL12 variants remain to be investigated. Specifically, the survival-enhancing and replating capacities of hematopoietic progenitor cells and the chemotaxis activities of the CXCL12 variants have not been quantified. Additionally, the antiviral (anti-HIV) activities of the new CXCL12 variants have not been investigated.
Previous studies have suggested that receptor activation (G-protein signaling) is not required for HIV-1 inhibition (2, 4, 20). The exact molecular determinants involved in chemokine activities are not well understood. Previous studies indicated that the two activities have distinct determinants that may overlap (14). Proost et al. have demonstrated that, in contrast to CD26/dipeptidyl-peptidase IV-processed RANTES3-68, truncated CXCL12α3-68 has diminished potency for inhibiting HIV-1 infection, indicating that the first three N-terminal residues may be required for the antiviral activity (33). However, Dettin et al. have demonstrated potent anti-HIV activity for synthetic peptides mapped to the C terminus of CXCL12β but observed no such activity with the analogous C terminus of CXCL12α (17). Other studies have demonstrated that the BBXB domain (N terminus) of CXCL12α is required for its optimal anti-HIV activity (39). The BBXB domain is present in most chemokines and is responsible for high-affinity binding to heparin sulfate (HS), chondroitin sulfate, and glycosaminoglycans present on chemokine receptors (34).
In this study we have expressed and purified large quantities of the six splice variants of CXCL12 and analyzed their properties in terms of antiviral and chemotaxis activities. We demonstrate that CXCL12γ is the most potent antiviral inhibitor, with no detectable enhancing activity for hematopoietic progenitor cell survival or replating capacity and the weakest chemotaxis activity. Our results suggest a strong association between the number of BBXB motifs and the antiviral activity of CXCL12.
All cell lines were obtained from the American Type Culture Collection (Manassas, VA). HeLa, MAGI-R5, and HeLa.T4.LTR cell lines were maintained in Dulbecco modified Eagle medium (DMEM) (Quality Biologicals, Gaithersburg, MD) containing 10% fetal bovine serum (FBS) (HyClone Laboratories, Logan, UT), 2 mM l-glutamine, and antibiotics. The CEM T-lymphoblast cell line was maintained in RPMI 1640 containing 10% FBS and antibiotics. The breast cancer cell line MCF-7 was maintained in charcoal-stripped modified Eagle medium supplemented with 10% FBS and penicillin/streptomycin.
Human peripheral blood mononuclear cells (PBMCs) were prepared by Ficoll-Hypaque fractionation of cell concentrates obtained from healthy donors seronegative for antibodies against hepatitis B virus and HIV. PBMCs were activated with 10 μg/ml phytohemagglutinin for 3 days in RPMI 1640 containing 10% FBS. Commercial CXCL12α was purchased from R&D Systems (Minneapolis, MN), and AMD3100 was a gift from AnorMed Corp. (Langley, British Columbia, Canada).
The mature peptide genes for all six CXCL12 splice variants were amplified by PCR. CXCL12γ, CXCL12θ, and CXCL12δ required multiple PCRs with reverse primers that extend the C-terminal end to include the entire DNA sequence. The final inserts were designed to have flanking 5′ Nde1 and 3′ Xho1 restriction sites. The reverse primer was also designed with the wild-type stop codon removed so that the histidine tag contained in the pET32a vector would be fused to the expressed protein. PCRs were carried out with the Faststart High Fidelity PCR System (Roche, Indianapolis, IN). The amplification protocol consisted of a 5-min denaturation at 94°C followed by 30 cycles of denaturation at 94°C for 1 min, annealing at 55°C for 1 min, and 72°C for 1 min. The amplified DNA strand was purified from an agarose gel with the PureLink Quick Gel extraction kit (Invitrogen, Carlsbad, CA). The amplified DNA was first cloned into the pGEM-T Easy plasmid (Promega, Madison, WI). The inserts were removed from pGEM-T Easy by digestion with Nde1 and Xho1 and ligated into the corresponding restriction sites of the bacterial expression plasmid pET32a (Novagen, Madison, WI). Positive clones were identified by digestion with the restriction enzyme HindIII and sequenced at the BBF DNA Sequencing Core Facility of Indiana University.
The BL21-Gold (DE3) pLysS strain of E. coli (Stratagene, La Jolla, CA) was transformed with the pET32a recombinant or pET32a empty vector plasmid. Single colonies were inoculated into 5 ml of LB broth containing 200 μg/ml ampicillin and 30 μg/ml chloramphenicol and shaken at 37°C overnight. After approximately 16 h of shaking, 5 ml of the inoculation was added to 1 liter of LB broth containing 200 μg/ml ampicillin and 30 μg/ml chloramphenicol. The 1-liter culture was shaken at 200 rpm and 37°C for approximately 2.5 h until an optical density at 600 nm (OD600) of 0.2 was achieved. Addition of 1 mM isopropyl-β-d-thiogalactopyranoside and shaking of the culture for 4 h at 200 rpm and 37°C induced protein expression. After being shaken, the cultures were aliquoted into 50-ml tubes and centrifuged at 3,500 × g and 4°C for 30 min. This step was repeated so that each 50-ml tube would contain a pellet from 100 ml of culture. The supernatants were discarded, and the pellets were allowed to air dry for 10 min and stored at −80°C until purification.
Bacterial pellets for the CXCL12 splice variants were suspended in 5 ml of lysis buffer (20 mM sodium phosphate, 0.5 M NaCl [pH 7.8], 1% Triton X) per pellet and rotated at room temperature for 1 h. The lysates were then centrifuged at 14,000 × g for 20 min. The supernatants containing the soluble bacterial proteins were discarded. The pellets containing inclusion bodies were suspended in 5 ml of denaturing binding buffer (20 mM sodium phosphate, 0.5 M NaCl, 6 M guanidine-HCl [pH 7.8]) and rotated at room temperature for 2 h. The lysates were sonicated intermittently in 30-s bursts to fully suspend the pellets. The lysates were then clarified by centrifugation at 16,000 × g for 20 min at room temperature. The clarified supernatants were poured onto a 1-ml-bed-volume Ni-nitrilotriacetic acid (NTA) histidine-binding column (Sigma, St. Louis, MO) equilibrated with 10 column volumes of denaturing binding buffer. Both ends of the column were immediately capped, and the columns were rotated at 4°C overnight. The denatured proteins were renatured on the column by stepwise removal of the denaturant, as previously described (30). The columns were washed consecutively with 10 ml of 20 mM sodium phosphate, 0.5 M NaCl (pH 7.8), and 6, 4, 2, and 0 M urea. The columns were then washed with 10 ml of 20 mM sodium phosphate, 0.5 M NaCl (pH 7.8), and 20 and 40 mM imidazole to remove nonspecific bacterial contaminants. The proteins were eluted with 3 ml of elution buffer (20 mM sodium phosphate, 0.5 M NaCl, 0.5 M imidazole [pH 7.3]). The elution buffer was removed from the purified proteins on a PD-10 column equilibrated with phosphate-buffered saline (PBS). Fractions of 1 ml were collected and screened for OD600 readings. The peak fractions were pooled and quantified by the Bradford assay (Bio-Rad, Hercules, CA). The resulting purified recombinants were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and stored at −80°C.
Migration of CEM T lymphoblasts was quantified as previously described (32). Briefly, CEM T-lymphoblast cells were suspended in chemotaxis medium (Iscove's modified Dulbecco medium [IMDM] supplemented with 0.5% bovine serum albumin [BSA]) (Quality Biologicals, Gaithersburg, MD) at a concentration of 2 million cells per ml. The reagents were suspended in chemotaxis medium at the concentrations indicated in Fig. Fig.3.3. The suspended chemokines were first added at 600 μl to the wells of a 24-well plate. A 5-μm-pore-size transwell membrane was then added to each well (Costar; Corning Incorporated, Acton, MA). The cells were added to the inside of the transwell membrane at a volume of 100 μl (2 × 105 cells). The plates were then capped and incubated for 4 h at 37° and 6% CO2. After the incubation period, the transwell membranes were removed from the wells and the migrated cells were counted by flow cytometry. Results are reported as the percentage of cells that migrated compared to total cell counts in wells with no transwell membranes.
MCF-7 cells were used to analyze the CXCR7-mediated chemotactic activities of the CXCL12 isoforms. MCF-7 cells have previously been shown to express high levels of CXCR7 and no CXCR4 (10). The CXCL12 variants were suspended to the specified concentrations in chemotaxis medium (IMDM supplemented with 0.5% BSA). The chemokines were loaded into the wells of a 96-well chemotaxis plate (Neuro Probe, Inc., Gaithersburg, MD), and a 12-μm-pore-size filter was attached. MCF-7 cells were trypsinized and suspended at 4 × 106 cells per ml in chemotaxis medium. For AMD3100 assays, the AMD3100 was suspended at 100 nM with the cells. The cells were added to the tops of the filter so that 1 × 105 cells were added to each well. The plates were capped and incubated at 37°C and 5% CO2 overnight. Following incubation, the filters were cleaned with a cell scraper and PBS. EDTA (0.2 mg/ml) was used to clean any cells that were not scraped from the filters. The filters were removed, and the migrated cells were counted with a hemacytometer.
Progenitor cell survival studies are linked with antiapoptotic effects (7, 27), and progenitor cell replating studies estimate the effects on self-renewal (9, 11). These studies for granulocyte-macrophage (CFU-GM) and multipotential (CFU-erythroid granulocyte-macrophage-megakaryocyte [CFU-GEMM]) progenitor cells were performed as previously reported (7, 27). The reagents and vendors the reagents were purchased from were also described previously (7, 27). In short, for cell survival studies, either low-density (LD) (density, <1.077 gm/cm3) or CD34+ magnetic bead-purified (purity, >90% CD34+) cord blood cells were plated at concentrations in 1 ml as indicated in the legend to Fig. Fig.44 in the absence or presence of different nanomolar concentrations of CXCL12 variants at time zero in semisolid methylcellulose culture medium. A combination of growth factors (1 U/ml human erythropoietin, 50 ng/ml human stem cell factor, 10 ng/ml human GM colony-stimulating factor, and 10 ng/ml human IL-3) was added to the plates at day 0 or at 24 (day 1) or 48 (day 2) h after initiation of the cultures. Cells were incubated at 5% CO2 and lowered (5%) O2 in a humidified atmosphere, and CFU-GM- and CFU-GEMM-derived colonies were scored 14 days after the addition of growth factors. For analysis of effects on the replating capacity of cord blood CFU-GEMM-derived colonies, such colonies formed in methylcellulose cultures (at the above cytokine concentrations in the absence or presence of CXCL12 isotypes) were removed as single colonies, and each single colony was replated into a separate secondary methylcellulose culture plate in the presence of the same cytokine concentration but in the absence of CXCL12 variants, as described previously (8, 9, 11) and in the legend to Fig. Fig.4.4. Secondary colonies were then scored to calculate the number of secondary replates per primary CFU-GEMM colony.
HIV Env-mediated fusion was quantified by a vaccinia virus-based reporter assay, as previously described (1). Briefly, MAGI-CCR5 cells (12) were infected by vaccinia virus encoding T7 RNA polymerase. Because MCF-7 cells are CD4−, they were coinfected with vCB-3 (vaccinia virus encoding CD4) and vaccinia virus encoding T7 RNA polymerase. Simultaneously, HeLa cells were coinfected by vaccinia virus encoding the β-galactosidase gene, driven by the T7 promoter, and vaccinia virus encoding either the X4-tropic Env LAV, the R5-tropic Env Ba-L, or the control Unc, an uncleaved HIV-1 Env that has its cleavage site mutated and therefore cannot engage in membrane fusion. We use Unc to measure the nonspecific background in the fusion assay. The target cells were plated in a 96-well plate at 1 × 105 cells per well and treated for 1 h with the recombinant chemokines at increasing concentrations. After 1 h of incubation at 37° and 6% CO2, the Env-expressing HeLa cells were mixed with the target cells at a 1:1 ratio. The cell mixtures were incubated for 2.5 h at 37° and 6% CO2. The extent of cell fusion was assayed by measuring β-galactosidase produced in fused cells. When PBMCs were used as target cells, they were infected with vaccinia virus encoding the β-galactosidase gene, driven by the T7 promoter, while the HeLa Env cells were coinfected with vaccinia virus encoding T7 polymerase.
The effects of the CXCL12α, CXCL12β, and CXCL12γ splice variants on HIV-1 infection were quantified with the indicator cell lines HeLa.T4.LTR (24) and MAGI-CCR5 (12). The evening prior to infection, 1 × 104 HeLa.T4.LTR or MAGI-CCR5 cells were plated in each well of a 96-well plate. The cells were first pretreated by chemokines suspended in DMEM supplemented with 10% FBS and 40 μg/ml dextran at the concentrations indicated in Fig. Fig.6.6. After 1 h of treatment, 100 μl of the X4-tropic HIV-1 strain IIIB or the R5-tropic strain Ba-L was added to each well. The cells were allowed to incubate with the HIV-1 for 2 h at 37° and 6% CO2. After the incubation period, the chemokines and HIV-1 were removed from the wells. The cells were then washed with 200 μl of PBS. Fresh DMEM, containing 10% FBS and freshly suspended chemokines, was added to each well, and the plates were incubated for 48 h at 37° and 6% CO2. Once the 48-h incubation was complete, the cells were lysed by the nonionic detergent NP-40. After transfer of 50 μl of lysate per well to a new plate, 50 μl of CPRG (chlorophenolred-β-d-galactopyranoside) was added to each well. β-Galactosidase was produced selectively in HIV-1-infected cells and quantified in nonionic detergent cell lysates by a 96-well spectrophotometer.
The predicted amino acid sequences of the four novel human CXCL12 splice variants are illustrated in Fig. Fig.1.1. We constructed bacterial expression vectors by PCR amplification of the cDNA for each splice variant, using pCDNA3/SDF-1β DNA as a template and identical forward and reverse primers with the alteration from CXCL12β at the carboxy-terminal end of the gene. Since the CXCL12γ, CXCL12δ, and CXCL12θ genes have variable lengths at the 3′ end, three reverse primers were designed for multiple separate nesting PCRs. The resulting PCR fragments were inserted into the Nde1 and Xho1 sites of the pET32a expression vector. The resulting recombinant vectors contained the mature peptide sequences with no signal peptide. In order to express the recombinant proteins in bacteria, an extra methionine was engineered at the amino-terminal end of all proteins. Also, the wild-type protein termination codons were removed in order to add additional leucine, glutamate, and six histidines to the carboxy-terminal end for purification purposes. The sequences of the cloned vectors were verified by the BBF DNA Sequencing Core Facility of the Indiana University School of Medicine.
Purification of the six CXCL12 splice variants was accomplished through the use of Ni-NTA+ His tag affinity chromatography columns. Each purification experiment yielded approximately 0.5 to 2.0 mg of protein. The relative concentrations of the proteins were measured by the Bradford assay, with BSA as the standard. The protein purities were verified with a Coomassie blue-stained SDS gel (Fig. (Fig.2A).2A). To verify our affinity purification procedure, a small fraction of the purified protein were fractionated by SDS-PAGE and blotted. The blot was probed with a polyclonal antibody to the histidine tag (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). The results of this analysis demonstrated that the purified proteins reacted with the anti-His antibodies and showed the expected molecular weights (Fig. (Fig.2B).2B). The apparent molecular weights of the purified CXCL12 variants were matched exactly to those obtained in the Coomassie blue-stained gel shown in Fig. Fig.2A.2A. These experiments verified the purity and molecular weights of our purified CXCL12 chemokine variants.
To verify the biological functions of the purified recombinant CXCL12 proteins, an in vitro chemotaxis assay was performed with the human CEM T-lymphoblast cell line. We used MC148R1 as a negative control, since it is well established that this chemokine-like protein does not induce chemotaxis of cells (15, 26). We expressed and affinity purified MC148R1 on His affinity columns in the same way as described for CXCL12 proteins (data not shown). As expected, the purified MC148R1 protein did not induce significant chemotaxis at any concentration (Fig. (Fig.3A).3A). In contrast, all CXCL12 variants induced low migration of cells at 20 nM concentrations and significantly increased migration at 100 nM (Fig. (Fig.3A).3A). An exception was CXCL12γ, which showed weak chemotaxis at 100 nM and was not significantly different at the highest concentration used. We also consistently observed significant chemotaxis activity for CXCL12δ at 500 nM concentrations. The strongest chemotaxis activity was consistently observed with CXCL12α and CXCL12. The results indicate that the CXCL12 variants have different potencies in chemotaxis assays, the weakest being CXCL12γ.
To determine whether chemotaxis occurred through CXCR4, we used AMD3100 to block CXCL12-induced cell migration. AMD3100 has previously been shown to antagonize CXCR4 (18, 35). The AMD3100 was added to the lower chamber along with the chemokine variant. The percent cell migration was measured in the absence or presence of AMD3100. We demonstrate that AMD3100 treatment reduced the chemotaxis activities of all affinity-purified CXCL12 proteins to background levels (Table (Table1).1). This is consistent with the staining profile we observed in the CEM T-cell line, where 95% of cells stain positively for CXCR4 and only 5% stain positively for CXCR7 (data not shown).
CXCR7-mediated chemotactic activity has previously been demonstrated with the A0.01 T-cell line transfected with CXCR7 (5). We utilized the MCF-7 cell line, which is CXCR7+ and has been demonstrated to be defective in CXCR4 expression by reverse transcription-PCR (10). We observed efficient migration of the CXCR4−/CXCR7+ MCF-7 cells with all CXCL12 variants. Some CXCL12 variants peaked at 20 nM (CXCL12δ and CXCL12θ), while others peaked at 100 nM (CXCL12α, CXCL12β, and CXCL12γ) and 500 nM (CXCL12) (Fig. (Fig.3B).3B). It has been previously demonstrated that CXCL12α binding to CXCR7-expressing MCF-7 cells is insensitive to AMD3100 (10). AMD3100 was added in a different set of experiments to verify the specificity of MCF-7 cell migration. No significant inhibiting effect for AMD3100 had been observed on chemokine-induced migration of MCF-7 cells with any CXCL12 variant (Table (Table2).2). In contrast, we observed some enhancement of CXCR7-mediated chemotaxis in the presence of AMD3100 (Table (Table2).2). The significance of this observation is currently under investigation. Together, the results suggest that the affinity-purified CXCL12 proteins are biologically functional in terms of inducing chemotaxis through CXCR4 and CXCR7 receptors.
CXCL12α enhances the survival of hematopoietic progenitor cells subjected to delayed addition of growth factors (7, 27), effects completely blocked by AMD3100 (23). The results demonstrate that only the commercial CXCL12α and the recombinant CXCL12α, CXCL12β, and CXCL12 variants we produced were active in enhancing the survival of human cord blood CFU-GEMM (Fig. (Fig.4A)4A) and CFU-GM (Fig. (Fig.4B)4B) subjected to delayed addition of growth factors, with the CXCL12α and CXCL12β variants being active at 10 and 20 nM concentrations and the CXCL12 variants active at 20 nM. The other variants of CXCL12 were not active in this assay at concentrations up to 20 nM.
An important function of hematopoietic stem cells is the capacity to self-renew or make more of themselves (9). While self-renewal is generally confined to stem cells, there is evidence from the in vitro replating capacity of immature subsets of hematopoietic progenitors that hematopoietic progenitor cells such as CFU-GEMM may have some limited capacity for self-renewal (9, 11). Recently, we found that CXCL12α enhanced the replating capacity of murine bone marrow and human cord blood CFU-GEMM (8). As shown in Fig. Fig.4C,4C, at concentrations of 20 nM, CXCL12α, CXCL12β, and CXCL12, but not the other CXCL12 isotypes, significantly enhanced the number of secondary colonies per replated human cord blood CFU-GEMM colony.
Because CXCL12 has been previously demonstrated to inhibit X4 HIV-1 infection (6, 31), we set up a fusion assay to examine the effects of the CXCL12 variant proteins on HIV-1 Env-mediated cell fusion. The assay measures fusion between effector cells expressing HIV-1 Env and target cells expressing CD4, CXCR4, and CCR5. The affinity-purified proteins were preincubated with target cells before the addition of effector cells. A dose-response analysis was performed for each of the CXCL12 variants preincubated with MAGI-R5 cells (Fig. 5A and B) or human PBMCs (Fig. (Fig.5C).5C). The results show that the CXCL12 protein variants caused a dose-dependent inhibition of X4 Env fusion (Fig. 5A and C) but had no significant effect on R5 Env fusion (Fig. (Fig.5B).5B). The R5-mediated fusion served as a negative control, as previous studies had demonstrated that CXCL12 does not bind CCR5 and does not inhibit R5 infection (6, 31).
We consistently observed different inhibition profiles with the six CXCL12 variant proteins. Table Table33 shows the 50% inhibitory concentrations (IC50s) obtained with these chemokines. The values represent average IC50s obtained in at least four different experiments. We consistently observed more potent inhibition with CXCL12γ (IC50, ~50 nM). The weakest activity in the X4 Env fusion assay was always obtained with CXCL12δ (IC50, ~1,350 nM). The results suggest that the observed differences in inhibition of X4 Env fusion by the CXCL12 variant proteins may correlate with the different amino acid structures at their carboxyl termini.
To determine the blocking efficiency of the CXCL12 variants on CXCR7-mediated Env fusion, we utilized the CXCR4−/CXCR7+ MCF-7 cell line as the target in an HIV-1 Env-mediated fusion assay. All CXCL12 chemokine variants showed dose-dependent inhibition of Env-mediated fusion (Fig. (Fig.5D).5D). The most effective was CXCL12β, with an IC50 of ~200 nM, and the least effective were CXCL12 and CXCL12δ, with IC50s of >500 nM. CXCL12γ and CXCL12θ had IC50s of 250 and 350 nM, respectively. Overall, the chemokine inhibition profile of CXCR7-mediated fusion was essentially similar to the CXCR4-mediated profile (CXCL12δ and CXCL12 being the weakest), except that CXCL12γ had approximately five-times-weaker activity in the CXCR7-mediated fusion assay (Fig. (Fig.5D5D).
To determine the effects of the CXCL12 protein variants on HIV-1 infection, we used the indicator cell lines HeLa.T4-LTR-lacZ and MAGI-CCR5, which have the β-galactosidase gene driven by the long terminal repeat (LTR) promoter integrated into the cellular DNA under G418 selection (24). In these experiments, the chemokines were preincubated with the target cells for 1 h and then infected with the X4 virus stock. To clearly represent the results of these experiments, we show the inhibition profiles of only three CXCL12 variants: CXCL12α, CXCL12β, and CXCL12γ (Fig. (Fig.6A).6A). The specificity of X4 inhibition by these CXCL12 chemokine variants was confirmed by examining their effects on R5 infection. None of the CXCL12 variants had any significant effect on R5 infection of MAGI-CCR5 cells (Fig. (Fig.6B).6B). Table Table33 shows the IC50 values calculated from at least three different X4 infectivity assays performed with the CXCL12 proteins. The results from these infection assays confirm that the most potent inhibitor among the CXCL12 variants is CXCL12γ (IC50, ~20 nM). We consistently observed that the IC50s required in the HIV-1 infection assays were much lower than those required for blocking of X4 Env-mediated fusion (Table (Table3).3). Interestingly, all CXCL12 variants enhanced X4 infection at low concentrations (5 nM). These results confirm our findings in the X4 Env-mediated assay and demonstrate that CXCL12γ is the most potent inhibitor of X4 infection.
The ability of chemokines to block HIV-1 infection (6, 13, 31) has stimulated interest in the mechanism of their antiviral activity and potential use as therapeutic agents. This study reports significant differences in chemotaxis and antiviral activities for the newly reported CXCL12 variants. The E. coli-expressed and affinity-purified CXCL12 proteins were biologically functional in terms of inducing chemotaxis through CXCR4 and CXCR7. The observed chemotaxis was blocked by AMD3100 with CEM T cells but not with MCF-7 cells. This provided evidence that the chemotaxis observed in CEM T cells was mediated through CXCR4. The chemotaxis activity of CXCR7+ cells has been demonstrated only in human cell lines (5, 10). Since AMD3100 had no inhibitory effect on chemotaxis observed with the MCF-7 cells, we conclude that migration of these cells was CXCR7 mediated.
Our findings are consistent with previous data in the literature demonstrating that CXCL12β is at least a twofold-more-potent HIV inhibitor than is CXCL12α (14, 21). We show for the first time that the newly identified CXCL12γ is the most potent of the CXCL12 variants in inhibition of X4 infection. CXCL12γ consistently showed more than threefold-higher antiviral activity than did CXCL12β. The chemotaxis activity of CXCL12γ was the weakest among the six splice variants. Additionally, hematopoietic progenitor cell survival and replating assays demonstrated no significant activity for CXCL12γ at concentrations up to 20 nM. Because the first 68 residues of the six variants are completely homologous, it is proposed that the nonhomologous C-terminal residues may contain critical determinants for the observed potent antiviral activity. The observed enhancement of survival and replating capacity of hematopoietic progenitor cells with CXCL12α, CXCL12β, and CXCL12 but not with the other three variants may implicate a critical role for the additional C-terminal amino acid residues present in the three new CXCL12 isoforms, CXCL12γ, CXCL12θ, and CXCL12δ, that were functional in the chemotaxis assays. These results may suggest the involvement of different molecular determinants in the different biological functions exhibited by the CXCL12 variants.
Most, if not all, chemokines bind to HS, a glycosaminoglycan (GAG) found ubiquitously at the cell surface. The current model argues that binding to HS either enhances the local concentration of chemokines in the vicinity of the G-protein-coupled receptor or provides a haptotactic gradient of the protein along cell surfaces (22). Previous studies reported that the BBXB motif of RANTES is the principal site for HS binding and controls receptor selectivity (34). Mutation of the first BBXB domain (residues 44 to 47) of RANTES abrogated 80% of HS binding capacity, whereas mutation of the second domain (residues 55 to 59) had no effect (34). The first-domain mutant showed an 80-fold reduction in affinity for CCR1 despite normal binding to CCR5 (34).
The BBXB domain located between residues 24 to 27 is present in the N-terminal region of all CXCL12 splice variants. It has been previously determined that CXCL12α associates with HS through the BBXB domain (24KHLK27) (3). Mutation of the 24KHLK27 sequence in CXCL12α produced a protein that is unable to bind HS and has reduced ability to inhibit X4-tropic HIV-1 infection but had no significant effect on chemotaxis or receptor activation (3, 39). It should be noted that the GAG binding sites are not restricted to the BBXB domains. The binding sites of three of the chemokines studied do have this motif: CXCL12, MIP-1α, and MIP-1β. Mutagenesis studies with the CXC chemokines IL-8 and PF-4 (which do not contain BBXB motifs) led to the conclusion that CXC chemokines may bind heparin through a cluster of positively charged residues located at the carboxyl-terminal amphipathic α helix (29, 37, 40).
The fact that CXCL12β is an at least twofold-more-potent inhibitor than CXCL12α (14; also this study) suggests that the increased anti-HIV-1 activity may be due to the presence of a second BBXB domain in its C terminus. Our results demonstrate a strong association between the number of BBXB domains and the antiviral potency of a CXCL12 variant. CXCL12γ (with five BBXB domains) consistently showed the greatest potency, compared to CXCL12α (one BBXB domain) or CXCL12β (two BBXB domains). CXCL12δ is the weakest splice variant in terms of anti-HIV-1 activity. It is interesting to note that CXCL12δ has the largest C terminus, with 50 extra amino acids compared to the prototypic CXCL12α. The extra amino acids in CXCL12δ do not contain any BBXB domains and are mostly nonpolar amino acids. In contrast, the C-terminal end of CXCL12γ contains four BBXB domains in addition to the conserved BBXB domain found in all CXCL12 splice variants. We hypothesize that the four BBXB domains in the C terminus of CXCL12γ may play a critical role in the overall structure of the protein. Whether the increase in antiviral potency is strictly due to the extra BBXB domains requires further investigation.
It is possible that structural determinants other than the BBXB domains are involved in the observed weak chemotaxis activity of CXCL12γ. The extra amino acid residues present at the C termini of the new CXCL12 variants might have created a different protein structure with weaker chemotactic properties. It is important to note that there are 19 basic amino acids at the additional C terminus of CXCL12γ; 7 of them are not contained within the BBXB domains. Previous studies reported that mutation of the first BBXB domain of CXCL12α and RANTES resulted in reduced antiviral activity but had no significant effects on chemotaxis (3, 34, 39). We also observed that CXCL12δ had weaker chemotaxis and HIV-1-inhibitory activities than did CXCL12α, even though both splice variants contain only one BBXB domain. Together, the data suggest that GAG binding may play a role in antiviral activity but not in chemotactic activities. The C-terminal regions of the CXCL12 splice variants may contain different structural determinants that affect their biological activities. Previous studies reported that removal of the C-terminal lysine reduced the chemotaxis activity of CXCL12α (16). Interestingly, we found that CXCL12 exhibited chemotaxis activity equal to that of CXCL12α, even though the C-terminal end of CXCL12 has the lysine residue removed and replaced with an asparagine followed by a cysteine. Further work is required to determine the mechanism behind these differences.
We consistently observed significant enhancement of X4 infection at low CXCL12 concentrations, with CXCL12α being the most potent enhancer. The mechanism for this enhancement is unknown. This phenomenon has previously been reported with the CC chemokines RANTES, MIP-1α, and MIP-1β (25) but not with CXCL12. Kinter et al. demonstrated that the enhancement occurred during the early stages of X4 infection and was associated with increased cell surface expression of CD4 and CXCR4 (25). It is important to note that such enhancement was observed at low chemokine concentrations: 10 nM in the study of Kinter et al. (25) and 5 nM in this study. We never observed such enhancement in our Env-mediated fusion assays. This could be due to the nature of the experimental setup, as fusion assay results are analyzed after 2.5 h of incubation of the partner cells, whereas infection is analyzed 48 h following the addition of chemokine.
CXCL12 has previously been proposed as an antiviral agent for decreasing virus load and preventing the emergence of the syncytium-inducing viruses, which are characteristic of the late stages of AIDS (31). A modified version of CXCL12γ with no chemotaxis activity might be a beneficial therapeutic agent for HIV+ individuals. The fact that CXCL12γ is found mainly in the heart, with no detectable expression in peripheral blood lymphocytes, makes it an attractive molecule for future investigation. Further analysis of the molecular determinants involved in the antiviral activities of the new CXCL12 variants will provide important insights into the structure-function relationships of the CXC chemokines.
We thank Zainab VanHorn-Ali for excellent technical assistance. Special thanks go to Lokesh Agrawal and Ronald Wek for helpful discussions. We thank Harikrishna Nakshatri for providing the MCF-7 cell line, Young-June Kim for providing the CEM T-lymphoblast cell line, and H. Lee Tiffany and Philip Murphy for helpful discussions.
This study was supported by NIH grant RO1 A152019-01 to G.A. and in part by grants RO1 HL56416 and RO1 HL67384 and a project of PO1 HL53586 to H.E.B. J.Q. was supported by a scholarship from the Chinese Scholarship Council, Beijing, China.
Published ahead of print on 16 May 2007.