• Hematopoiesis;
  • Chemokines;
  • Chemokine receptors;
  • CXCR4;
  • CCR5;
  • Stromal derived factor-1 (SDF-1);
  • HIV


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In order to better define the role of HIV-related chemokines in human erythropoiesis we studied: A) the expression of chemokine receptors, both on human CD34+ cells which include erythroid progenitors and on more mature erythroid cells; B) the functionality of these receptors by calcium flux, chemotaxis assay and phosphorylation of mitogen-activated protein kinases (MAPK) p42/44 (ERK1/ERK2) and AKT, and finally C) the influence of chemokines on BFU-E formation. We found that HIV-related chemokine receptor CXCR4, but not CCR5, is detectable on human CD34+ BFU-E cells. CXCR4 surface expression decreased during erythroid maturation, although CXCR4 mRNA was still present in cells isolated from differentiated erythroid colonies. SDF-1, a CXCR4 ligand, induced calcium flux and phosphorylation of MAPK (p42/44) and AKT in CD34+KIT+ bone marrow mononuclear cells which contain BFU-E, as well as chemotactic activity of both human CD34+ BFU-E progenitors and erythroid cells isolated from day 2-6 BFU-E colonies. Responsiveness to SDF-1 decreased when the cells differentiated to the point of surface expression of the erythroid-specific marker Glycophorin-A. In contrast, the CCR5 ligands (macrophage inflammatory protein-1α [MIP-1α], MIP-1β, and RANTES) did not activate calcium flux, MAPK and AKT phosphorylation or chemotaxis of CD34+KIT+ cells or cells isolated from the BFU-E colonies. Interestingly, none of the chemokines tested in this study had any effect on BFU-E colony formation. In conclusion, only CXCR4 is functional, and its specific ligand SDF-1 may therefore play an important role in the homing and/or retention of early erythroid precursors in the bone marrow environment.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Patients infected with HIV-1 frequently exhibit a variety of hematological abnormalities, among the most prominent is anemia [1-6]. Anemia is very common in HIV-infected individuals, being diagnosed in approximately 10%-20% at initial presentation, and in 70%-80% of patients at some time during the course of their disease [3]. Anemia is likely multifactorial in origin and might result from: A) direct infection of erythroid progenitors by HIV; B) the inhibitory influence of HIV-related proteins, and C) the negative influence of the various inflammatory cytokines and chemokines elaborated in response to the infection itself [2, 3, 5].

The discovery that some chemokine receptors are coreceptors for HIV entry into cells offers new avenues for increasing our understanding of the mechanisms underlying HIV-1-associated marrow dysfunction [1, 4]. All HIV-1 strains studied to date use CCR5 (R5 strains), CXCR4 (X4) or both receptors (R5X4) to enter cells [1, 4]. It has been found that the CCR5 receptor binds macrophage inflammatory protein-1α (MIP-1α), MIP-1β, and RANTES, whereas the CXCR4 receptor binds stromal cell-derived factor-1 (SDF-1).

We previously reported that human erythroid progenitors and erythroblasts cannot be infected by HIV-1 envelope pseudotyped XR and R5 viruses and suggested that other indirect mechanisms are responsible for the pathogenesis of HIV-related anemia [7]. Since many different chemokines are elaborated in the body during HIV-1 infection [1, 2, 6] and some of these chemokines have been reported to inhibit erythropoiesis [8-10], we became interested in their potential role in the pathogenesis of HIV-related anemia. We approached this question from different points of view by: A) looking at the expression of chemokine receptors both on human CD34+ cells which include erythroid progenitors and on more mature erythroid cells; B) evaluating the functionality of the chemokine receptors in chemotaxis, mitogen-activated protein kinase (MAPK) p42/44 and AKT phosphorylation and calcium flux studies, and finally by C) studying the effect of selected chemokines on in vitro proliferation of human BFU-E progenitors.

We found that from all the chemokines tested, only SDF-1 induced calcium flux, MAPK p42/44, and AKT phosphorylation and chemotaxis in early human erythroid precursor cells. Therefore, as determined by the biological assays employed in this study, only CXCR4 appeared to be a functional chemokine receptor in cells of this lineage and SDF-1 may play an important role in the retention of early erythroid precursors in the bone marrow (BM) environment [11-17]. Finally, we did not observe any influence of chemokines on proliferation of human erythroid progenitor cells.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References


Light-density marrow mononuclear cells (MNC) were obtained from 15 consenting healthy donors and depleted of adherent cells and T lymphocytes (AT MNC) as described [18-21]. AT MNC were enriched for CD34+ cells by immunoaffinity selection with the murine monoclonal antibody (mAb) HPCA-1 (Becton-Dickinson; San Jose, CA) and magnetic beads according to the manufacturer's protocol (Dynal; Oslo, Norway).

Selection of CD34+KIT+ AT MNC by Fluorescence-Activated Cell Sorting (FACS)

The c-kit receptor positive (c-kit-R+) subset of CD34+ cells was isolated by FACS. Briefly, 2 × 107 human AT MNC were suspended in phosphate-buffered saline (PBS), supplemented with 5% bovine calf serum (BCS), and labeled for 30 min at 4°C with anti-Kit R mAb (clone #104D2) directly conjugated with Cy5 (Becton Dickinson; 20 μl/106 cells) and with an anti-CD34 mAb directly conjugated with phycoerythrin ([PE] anti-HPCA-2, Becton-Dickinson; 20 μl/106 cells). After the final incubation, cells were washed 3× in ice-cold PBS supplemented with 5% BCS, and then subjected to FACS using a FACS Star Plus II (Becton Dickinson).

Ex Vivo Expansion of Human Erythroid Cells

CD34+ cells were expanded in a serum-free liquid system as described [19]. Briefly, CD34+ AT MNC were resuspended in Iscove's Dulbecco's modified Eagle's medium (104/ml; GIBCO BRL; Grand Island, NJ) supplemented with 25% of artificial serum containing 1% delipidated, deionized, and charcoal-treated bovine serum albumin (BSA), 270 μg/ml iron-saturated transferrin, insulin (20 μg/ml), and 2 mmol/l L-glutamine (all from Sigma; St. Louis, MO). BFU-E growth was stimulated with recombinant human (rHu) erythropoietin (EPO) (2 U/ml), and rHu kit ligand (KL) (10 ng/ml). Cytokines were from R&D Systems, Inc., Minneapolis, MN. Cultures were incubated at 37°C in a fully humidified atmosphere supplemented with 5% CO2. Under these conditions, approximately 100% of expanded cells were glycophorin A (GPA-A) positive, CD33, and gpIIa/IIIb-negative after 14 days [19]. In our studies we employed erythroid cells isolated from these cultures which were expanded from 2 to 14 days.

Cell Cultures

Briefly, CD34+ AT MNC cells (1 × 104) were cloned in 1 ml of Iscove's modified Dulbecco's medium (IMDM) (GIBCO BRL) containing 0.8% methylcellulose (Methocel MC; Fluka, Switzerland) and supplemented with 1% delipidated, deionized, charcoal-absorbed BSA (Sigma), 270 μg/ml iron-saturated transferrin (Sigma), 5.6 μg/ml cholesterol (Sigma), and 2 mmol/l L-glutamine as described [18]. rHu growth factors appropriate for growth of the colonies were added to the mixture which was then transferred to 3.5-cm plastic petri dishes and incubated (37°C, 95% air, 5% CO2 humidified atmosphere) for the appropriate times. Growth factors employed and utilized concentrations were as follows: EPO (5 U/ml) + KL (100 ng/ml) for optimal stimulation and EPO (1 U/ml) + KL (40 ng/ml) for suboptimal stimulation of BFU-E.

The various chemokines (SDF-1, MIP-1α, MIP-1β, RANTES, MIP-3α, interleukin (IL)-8, Gro-α, Gro-β, Gro-γ, NAP-2, ENA-78) in a dose of 100 ng/ml (R&D Systems Inc. and PeproTech Inc.; Rocky Hill, NY) were added as costimulators to some culture dishes. In all the experiments BFU-E colonies were counted with the aid of an inverted microscope on day 14.

FACS Analysis of Chemokine Receptor Expression

The expression of CCR1, CCR2, CCR3, CCR4, CCR5, CCR6, CCR8, CXCR2, CXCR3, and CXCR4 on expanded ex vivo erythroid cells from CD34+ BMMC under serum-free conditions and mature peripheral blood erythrocytes were evaluated by FACS as described [7]. mAbs against CCR1 (Clone #53504), CCR2 (Clone #48607), CCR3 (rat mAbs Clones #61856, 61832, 61837, 61834), CCR5 (Clone #45529), CCR6 (Clone #53113), CXCR2 (Clone #48301), CXCR3 (Clone #49801), and CXCR4 (Clone #44701) were obtained from R&D Systems. The CCR4 mAb was a kind gift from Dr. Patrick Gray (ICOS Corp.; Seattle, WA) and the rabbit polyclonal against CCR8 was a kind gift from Dr. Richard Horuk (Berlex; San Rafael, CA) to Dr. Robert Doms (University of Pennsylvania; Philadelphia, PA). Before we performed our studies, the specificity of these antibodies against their cognate receptors was confirmed on transfected 293T cells. Flow cytometric staining and analysis of the receptors were performed as described [7]. Briefly, the cells were stained in PBS (Ca and Mg-free) supplemented with 5% BCS (Hyclone; Logan, Utah). Primary mAbs were detected with secondary phycoerythrin or fluorescein isothiocyanate (FITC)-conjugated goat antimouse mAbs (Sigma) (1:100) or antirat antibodies. After the final wash, cells were fixed in 1% paraformaldehyde prior to FACS analysis using FACscan (Becton Dickinson). Similarly, we have stained expanded ex vivo human erythroid cells under serum-free conditions [19] and mature erythrocytes isolated from peripheral blood. In addition, we have also been looking for an expression of chemokine receptors on CD34+ BMMC. In brief, BMMC were double stained with chemokine receptors, and specific mAbs were detected with PE and anti-CD34 mAb directly conjugated with FITC anti-HPCA-1, (20 μl/106 cells; Becton Dickinson) as described [7]. Data analysis was performed using the Cell Quest (Becton Dickinson).

Chemotaxis Studies

All experiments were performed in triplicate on purified CD34+ or serum-free expanded erythroid cells. Briefly, after isolation the cells were resuspended in serum-free medium (106/ml) and equilibrated for 10 min at 37°C. In the meantime 600 μl/point of prewarmed serum-free medium containing appropriate ligand was added to the lower chamber of the Costar Transwell 24-well plate, 6.5 mm diameter, 5 μM pore filter (Costar Corning Co.; Cambridge, MA). Subsequently, 100 μl aliquots of the cell suspension were distributed to the upper chambers, and cultures were incubated at 37°C, 95% humidity, 5% CO2 for 4 h. After 4 h of incubation, the plates were evaluated under an inverted microscope and subsequently, cells from the lower chambers were collected and cell number was scored by FACscan (Becton Dickinson) as described [13]. Briefly, the cells were gated according to their forward scatter (FSC) and sideward scatter (SSC) parameters and counted during a 20-sec acquisition. Data are demonstrated as a percentage of the input number of the cells. If required, samples of cells from the lower chambers have been plated in vitro and stimulated to growth erythroid colonies as described above.

Calcium Flux Studies by FACS

Sorted by FACS CD34+ KIT+, isolated by magnetic beads CD34+ and expanded ex vivo under serum-free conditions, erythroid cells were employed for Ca2+ flux studies. Cells were loaded with Indo-1AM (Molecular Probes; Eugene, OR) as described [22]. Subsequently, 2 × 105 sorted cells or 5 × 105 expanded cells were spun down at room temperature and resuspended in 1 ml of prewarmed loading medium (1 × Hank's balanced salt solution, 1 mM MgCl2, 1 mM CaCl2, 1% fetal bovine serum [FBS]), and 40 μl of 100 mM Probenecid + 10 μl of Indo-1AM mixture (50 μg of Indo-1 AM, 25 μl of Pluronic acid 127, 113 μl of FBS) were added. Cells were incubated for 30-45 min at 30°C and were then gently mixed every 10 min. After incubation, the cells were spun down at room temperature, supernatant was discarded and cells were resuspended in 300-500 μl of fresh loading medium. The cells were then equilibrated for 10 min at 37°C and analyzed by FACS StarPlus as described [22].

Calcium Flux Studies by Spectrophotofluorimeter

Calcium flux studies on expanded ex vivo CD34+ cells were also performed by a spectrophotofluorimeter, as previously described [21]. Briefly, the CD34+ cells were loaded with Fura-2/AM (Molecular Probes) for 30 min at 37°C. Fluorescence was recorded with an Aminco-Bowman Series-2 Luminescence Spectrophotofluorimeter (SLM Instruments, Inc.; Urbana IL). Cells (1 ml aliquots) were stirred continuously in a warmed holder during the period of changes in fluorescence recording. Fluorescence was monitored at 340 nm and 380 nm for excitation and 510 nm for emission. The data were recorded as the relative ratio of fluorescence excited at 340 nm and 380 nm.

Detection of MAPK and AKT Phosphorylation

1 × 106 of purified CD34+KIT+ cells were put in RPMI medium containing 0.5% BCS over night in order to make the cells quiescent. After 12 h, cells were transferred to the fresh RPMI without serum for an additional 2 h. After that time, cells were divided and stimulated with appropriate ligands for 1 min at 37°C, and subsequently they were spun down in order to stop the reaction. Supernatants were removed immediately and the cells were lysed in the lysing buffer (150 mM NaCl/50 mM Tris HCl/1% Triton-X-100 plus protease and phosphatase inhibitors) for 5 min on ice. Subsequently, the equal amounts of 2× loading buffer were added and the lysates were boiled for 3 min. The extracted proteins were separated by 15% SDS-PAGE gel and transferred to the nitrocellulose membrane—Hybond ECL (Amersham Life Sciences; Little Chalfont, England). Phosphorylation of 44/42 MAPK and AKT was detected by protein immunoblotting using monoclonal 44/42 phospho-specific MAPK or rabbit polyclonal AKT phospho-specific AKT antibodies (New England Biolabs; Beverly, MA) with horseradish peroxidase-conjugated goat anti-mouse IgG as secondary antibody (Santa Cruz Biotech; Santa Cruz, CA) as described [23]. The membranes were developed with ECL reagent (Amersham Life Sciences) and subsequently dried and exposed to film. Equal loading in the lanes was evaluated by stripping the blot and reprobing with an anti-MAPK antibody clone #9102 and an anti-AKT antibody clone #9272 (New England Biolabs).

Isolation of mRNA from BFU-E

Cells from expanded ex vivo BFU-E colonies growing in serum-free methylcellulose cultures were analyzed for CXCR4, CCR5, and β-actin expression by reverse transcriptase-polymerase chain reaction (RT-PCR). Briefly, colonies were visualized with the aid of an inverted microscope and plucked from serum-free cultures on day 6 through day 9 using a Pasteur pipette. Randomly selected colonies (∼20) from each culture dish were utilized for each assay. Colonies were resuspended in 10 ml of IMDM and incubated for 1 h at 37°C to dissolve the methylcellulose. The cells were washed 2× in PBS after which their mRNA was extracted as detailed below.


Cells from 20 colonies were lysed in 200 μl of RNAzol (Biotecx Labs; Houston, TX) + 22 μl of chloroform (Sigma). The aqueous phase was collected and mixed with 1 volume of isopropanol (Sigma). RNA was precipitated overnight at –20°C. The RNA pellet was washed in 75% ethanol and resuspended in 3×-autoclaved H2O. RT-PCR was carried out as described [7, 18, 20]. Briefly, mRNA (0.5 μg) was reverse-transcribed with 500 U of Moloney murine leukemia virus reverse transcriptase and 50 pmol of an oligodeoxynucleotide (ODN) primer complementary to the 3′ end of the following sequence of CXCR4 (5′-CAA GGA AGC TGT TGG CTG AAA-3′) or CCR5 (5′-GAA AAT GAG AGC TGC AGG TGT-3′), according to the reported cDNA sequences as described [7]. The resulting cDNA fragments were amplified using 5 U of Thermus aquaticus (Taq) polymerase and primers specific for the 5′ end of CXCR-4 (5′-CGA GGC AAG TGA CGC CGA GGG GCT G-3′) or for the 5′ end CCR5 (5′-GCT GTC CAC ATC TCG TTC TCG-3′). β-actin mRNAs were amplified simultaneously using specific primers as described respectively [18]. Amplified products (10 μl) were electrophoresed on a 2% agarose gel, and transferred to a nylon filter and documented photographically. Specificity of the amplified products was further confirmed by Southern blotting as described [18].

Statistical Analysis

Arithmetic means and standard deviations were calculated on a MacIntosh computer using Instat 1.14 (GraphPad; San Diego, CA) software. Experiments were performed on cells obtained from 15 healthy donors. Cell cultures were carried out in quadruplicate. Data were analyzed using the Student's t-test for unpaired samples. Statistical significance was defined as p < 0.05.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Expression of Chemokine Receptors on Human CD34+ Cells Which Include Erythroid Progenitors and on More Mature Erythroid Cells

The earliest erythroid progenitors (BFU-E) are found among CD34+ cells [24]. We therefore examined these cells for the presence of chemokine receptors. In our previous studies, we demonstrated that ∼50% of CD34+ cells express the CXCR4 protein by FACS [4, 7]. At the same time, the CCR5 protein was undetectable on the surface [4, 7]; however, at the same time we demonstrated in these cells a presence of CCR5 intracellularly [25].

In this paper, we extended these observations by looking for the presence of eight other chemokine receptor proteins on human CD34+ cells. We found that in addition to CXCR4, a small percentage of CD34+ cells (∼5%-10%) express the CCR6 receptor. At the same time, we were unable to demonstrate the presence of CCR1, CCR2, CCR3, CCR4, CCR5, CCR8, CXCR2, or CXCR3 receptors on the same cells using flow cytometry. Figure 1 demonstrates the coexpression of selected chemokine receptors (CCR1, CCR3, CCR5, CCR6, and CXCR4) and CD34 antigen on human BMMC in the lymphocyte gate. We then phenotyped more differentiated erythroid cells for expression of these chemokine receptors. First, we looked for the presence of major HIV coreceptors (CXCR4, CCR5) on human erythroid cells grown in serum-free liquid cultures, (Fig. 2A-D). Figure 2B shows that the CXCR4 protein was still detectable on ∼15% of human erythroid cells cultured ex vivo for four days and became undetectable on more differentiated erythroid cells which were cultured for six days (Fig. 2D). By employing a more sensitive RT-PCR assay, we were still able to detect the presence of CXCR4 mRNA in cells isolated from day 11 erythroid colonies (Fig. 3). Similarly, while the CCR5 protein was not detected by FACS on human CD34+ cells (Fig. 1D), or cultured erythroid cells (Fig. 2A, C), CCR5 mRNA was easily detected by RT-PCR (Fig. 3). Therefore, we postulated that the apparent lack of expression of major chemokine receptors on the differentiated erythroid cells was the result of limited FACS detection sensitivity.

thumbnail image

Figure Figure 1.. Expression of chemokine receptors on BM CD34+BMMC.BMMC were isolated from BM aspirates of healthy donors by Ficoll-gradient centrifugation and stained with mAbs against CD34 antigen (FITC) and chemokine receptors (PE). (A) Forward and side scatter analysis of BMMC. Lymphocyte region is defined by R1. (B-F) Analysis of cells dual-labeled with FITC-anti-CD34 mAb and (B) PE-anti-CCR1 mAb; (C) PE-anti-CCR3 mAb; (D) PE-anti-CCR5 mAb; (E) PE-anti-CCR6 mAb, and (F) PE-anti-CXCR4 mAb. Data from at least three different donors were analyzed with similar results. Data from a representative donor is presented.

Download figure to PowerPoint

thumbnail image

Figure Figure 2.. FACS analysis of CCR5 (A, C) and CXCR4 (B, D) expression in erythroid cells expanded from CD34+cells under serum-free conditions for four days (upper panel) and six days (lower panel).The isotype negative controls are overlaid (bold line), and M1 represents a positive population. Data from four different donors were analyzed. Histograms from a representative donor are presented.

Download figure to PowerPoint

thumbnail image

Figure Figure 3.. RT-PCR analysis of CXCR4 (lanes 1, 2) and CCR5 (lanes 3, 4) mRNA expression in erythroid cells isolated from day 11 BFU-E colony.Lanes 1, 3 = RT-PCR for the presence of CXCR4 and CCR5 mRNAs; lanes 4, 5 = PCR control for a potential “contamination” of genomic DNA by RNA; lane 5 = mRNA integrity control (RT-PCR for β-actin); lanes 6, 7, 8 = negative RT-PCR control reactions (H20 instead of mRNA). Specificity of the RT-PCR products shown was confirmed by Southern blotting (data not shown).

Download figure to PowerPoint

We also examined erythroid cells for expression of eight other chemokine receptors after 6, 9, and 11 days of ex vivo expansion cultures. We noticed that at day 6 these cells express CCR6 and CXCR2, but not CCR1, CCR2, CCR3, CCR4, CCR8, and CXCR3 proteins (Fig. 4). We also found that CCR6 was still present on ∼20% erythroid cells, expanded for 11 days (not shown).

thumbnail image

Figure Figure 4.. Expression of chemokine receptors on serum-free expanded human erythroid cells.The red line represents the isotype control antibody. Cells were expanded from CD34+BMMC for six days in the presence of EPO + KL. Data from three different donors were analyzed. Histograms from a representative donor are presented.

Download figure to PowerPoint

Finally, in order to complete the phenotype of the entire spectrum of erythroid cells, we examined human peripheral blood erythrocytes for expression of chemokine receptors, but none were found (not shown).

Functional Studies on Chemokine Receptor-Chemokine Axes

Flow cytometric detection of chemokine receptors on cell surfaces is sensitive to the level of ∼200-500 molecules. If a protein is expressed at a lower level, a definitive statement about expression cannot be made. Further, the absence of a particular protein as assessed by FACS does not allow us to conclude that this receptor is not present on the cell surface. In fact, mRNAs for many chemokine receptors have been found in early human erythroid cells [9, 10]. We assumed that various chemokine receptors that were not detectable by FACS might still be functional even though they are expressed at low levels on the cell surface. To examine this possibility we performed functional studies, looking specifically for the ability of selected chemokines to activate chemotaxis in CD34+ progenitors and differentiating erythroid cells, induce calcium flux and phosphorylate MAPK p42/44 (ERK-1, ERK-2) and AKT.

Functional Analysis of Chemokine Receptors on CD34+ Cells by Chemotaxis Assay

As demonstrated above, ∼50% of human CD34+ cells coexpress X4 HIV-binding CXCR4 coreceptor, while ∼5%-10% coexpress the CCR6 protein. In addition, human CD34+ cells express at least CCR5 mRNA for the R5 HIV cell surface coreceptor [4]. The biological significance of CXCR4, CCR6 expression and potential presence of CCR5 receptor was investigated by employing various functional studies. First, we asked whether any of the chemokines show a chemotactic activity against human CD34+ cells. We found that only SDF-1 (CXCR4 ligand), but not MIP-1α, MIP-1β, RANTES (CCR5 ligand) or MIP-3α (CCR6 ligand) were able to attract human CD34+ cells (Fig. 5). It has been reported that ∼25%-45% of human BM CD34+ cells show chemotaxis to SDF-1 [11, 12, 16]. We found in our studies that ∼ 40% of the human CD34+ cells migrated to SDF-1 in the chemotaxis assay (Fig. 5). In addition, we noticed that among CD34+ cells, which migrated in chemotaxis assay to SDF-1, BFU-E were present, which formed erythroid colonies after replating to the semisolid methylcellulose cultures. We calculated that 47% ± 12% of input BFU-E migrated during 4 h in the chemotaxis assay to SDF-1 (300 ng/ml).

thumbnail image

Figure Figure 5.. Activation of chemotaxis by various chemokines in the human erythroid cells, expanded under serum-free conditions from CD34+BMMC.Chemotaxis was evaluated at days 3, 6, and 9 (x axis). The percentage of the cells which responded by chemotaxis to different chemokines is shown on y axis (to the left), and the percentage of GPA-A+cells is shown on y axis (to the right). The data are pooled from the three independent experiments performed on cells from the three different donors.

Download figure to PowerPoint

Moreover, since it has been recently reported that MIP-1α, MIP-1β and RANTES activate chemotaxis in malignant hematopoietic progenitor CTS cells first after addition of KL [26], we have supplemented medium containing chemokines with this growth factor. The addition of KL, however, did not influence chemotaxis of human CD34+ cells to MIP-1α, MIP-1β and RANTES (not shown).

Functional Analysis of Chemokine Receptors on Erythroid Cells Expanded Ex Vivo by Chemotaxis Assays

Next, we studied chemotaxis of the differentiated erythroid progenitor/precursor cells. We found that human early erythroid cells showed chemotaxis to SDF-1 after three and six days of culture. It is also worth mentioning that early erythroid cells from day 6 expansion cultures, collected from the lower chambers after chemotaxis assay, formed small colony-forming units-erythroid (CFU-E) colonies after replating in the methylcellulose cultures (not shown). We noticed that the chemotactic activity of early erythroid cells cultured ex vivo decreased after they started to differentiate and express GPA-A on their surface (Fig. 5). Accordingly, by day 9, all the cultured cells were GPA-A+ and no longer migrated in response to SDF-1 (Fig. 5).

Surprisingly, even though early erythroid cells express CCR6 and CXCR2 proteins as well as CCR5 mRNA, the ligands which bind to those receptors (MIP-3α, IL-8, MIP-1α, MIP-1β, RANTES) did not activate chemotaxis in those cells (Fig. 5).

Functional Analysis of Chemokine Receptors on CD34+ Cells by Calcium Flux Studies and Phosphorylation of MAPK (ERK1 and ERK2) and AKT

In order to learn more about the biological role of chemokine receptor expression on human CD34+ cells, we performed calcium flux studies. Since CD34 antigen marks early lymphoid cells as well as myeloid cells, we FACS sorted a population of CD34+KIT+ cells, which is enriched in BFU-E [26], to identify responding cells by lineage. Accordingly, CD34+KIT+ cells were loaded with Indo-1AM, and calcium flux was evaluated by FACS. We found that SDF-1 (Fig. 6A) but not MIP-1α, MIP-1β (Fig. 6B), or RANTES and MIP-3α (not shown) strongly induced calcium flux in this population of early stem/progenitor cells. As demonstrated in Figure 6A, SDF-1 induced calcium flux in almost 100% of CD34+KIT+ cells. Similarly, we found that SDF-1, but not MIP-1α, MIP-1β, MIP-3α, or MCP-3, were able to phosphorylate ERK1 and ERK2 and AKT in human CD34+KIT+ cells (Fig. 7). This strongly supports our observation that, in contrast to CCR5, CCR6 and CCR2, CXCR4 receptor protein is not only detectable but functional on human CD34+ KIT+ cells.

thumbnail image

Figure Figure 6.. FACS study on calcium flux in CD34+KIT+cells.Cells were loaded with Indo-1AM and subsequently stimulated by SDF-1 (A) and MIP-1α, MIP-1β, and RANTES (B). A representative experiment is shown of three independent repeats with similar results.

Download figure to PowerPoint

thumbnail image

Figure Figure 7.. Western blot data on phosphorylation of MAPK (ERK1 - p44 and ERK2 - p42) (panel A) and AKT (panel B) in CD34+KIT+BMMC cells.Sorted by FACS, human CD34+KIT+BMMC were starved for 24 h in serum-depleted medium, and subsequently nonstimulated (lane 1) or stimulated for 1 min with MIP-1α (lane 2), MIP-1β (lane 3), MIP-3α (lane 4), MCP-1 (lane 5), and SDF-1 (lane 6). Equal loading in the lanes was evaluated by stripping the blot and reprobing with an anti-MAPK antibody (ERK1/ERK2) and an anti-AKT antibody (lower parts of panels A and B, respectively). A representative experiment is shown of three independent repeats with similar results.

Download figure to PowerPoint

Functional Analysis of the Chemokine Receptors on Erythroid Cells Expanded Ex Vivo by the Calcium Flux Studies

The ability of different chemokines to induce calcium flux, both in the CD34+ cells stimulated for two days with KL + EPO for erythropoietic expansion (Fig. 8), and in more differentiated erythroid cells that had been expanded ex vivo for three to six days, was also evaluated (Fig. 9). Accordingly, CD34+ cells were isolated by magnetic beads and cultured serum-free for two days with KL + EPO (Fig. 8), and calcium flux was first evaluated by spectrophotofluorimeter. We found that SDF-1 (Fig. 8B) but not MIP-1α (Fig. 8A), MIP-1β, RANTES, and MIP-3α (not shown) were able to induce calcium flux in this population of cells.

thumbnail image

Figure Figure 8.. Representative spectrophotofluorimetrical calcium flux study in human CD34+cells expanded serum-free for two days toward erythroid lineage.CD34+cells were isolated by magnetic beads, loaded with Fura-2/AM and subsequently stimulated by MIP-1α (A) and SDF-1 (B). Three different experiments were performed with similar results.

Download figure to PowerPoint

thumbnail image

Figure Figure 9.. Calcium flux in human day-3 (A), day-4 (B) and day-6 (C) serum-free expanded erythropoietic cells.Cells were loaded with Indo-1AM and subsequently stimulated by SDF-1. Calcium flux was evaluated by FACS. A representative experiment is shown of three independent repeats with similar results.

Download figure to PowerPoint

Subsequently, calcium flux was examined in cultured human erythroid cells by FACS (Fig. 9). We chose this technique because FACS, in contrast to spectrophotofluorimetry, not only allows for detection of a calcium flux itself but also allows for an estimate of the percentage of the cells which respond in this assay. We noticed that a subpopulation of erythroid cells, which was expanded ex vivo for three (Fig. 9A) and four days (Fig. 9B) responded to stimulation with SDF-1. To our surprise ex vivo-expanded erythroid cells did not show calcium flux after stimulation with MIP-1α, MIP-1β, RANTES, IL-8, and MIP-3α (not shown) even though the corresponding chemokine receptors were expressed. Simultaneously, we did not observe any influence of those chemokines on MAPK and AKT phosphorylation in the same population of cells (not shown).

Influence of HIV-Related and Other Chemokines on BFU-E Growth Under Serum-Free Conditions

Data from the literature on the influence of chemokines on human hematopoiesis are inconsistent [6, 8-10, 27-32] and in our hands neither MIP-1α nor IL-8 inhibited human erythropoiesis in vitro [33]. We therefore reexamined these observations focusing on those chemokines, for which the receptors are expressed at a high level on the BFU-E (CXCR4) or on more differentiated erythroid cells (CCR6, CXCR2). Human CD34+ cells were isolated by magnetic beads and cultured serum-free in the presence or absence of different chemokines (Table 1). Transforming growth factor-β1, a potent known inhibitor of the erythropoietic colony formation, was added to some cultures [34].

Table Table 1.. The effect of various chemokines (100 ng/ml) and TGF-β1 (5 ng/ml) on BFU-E formation by human CD34+ cells
BFU-E growth costimulated with:n BFU-E colonies (±SD)/104 cells plated
  1. a

    Isolated by magnetic beads CD34+ cells were plated serum-free and stimulated with EPO (5 U/ml) + KL (100 ng/ml). Each data entry constitutes four independent clonogenic assays from six different donors.

  2. b

    *p < 0.0001 in comparison to (—).

(—)219 ± 34
SDF-1222 ± 51
MIP-1α215 ± 49
MIP-1β243 ± 66
RANTES219 ± 57
MIP-3α211 ± 43
IL-8228 ± 39
Gro-α217 ± 52
Gro-β220 ± 37
Gro-γ208 ± 28
NAP-2231 ± 53
ENA-78214 ± 28
TGF-β143 ± 22*

As shown in Table 1, SDF-1 (CXCR4 ligand), MIP-3α (CCR6 ligand) and IL-8, Gro-α, -β, -γ, NAP-2, ENA-78 (CXCR2 ligands) all failed to affect human BFU-E growth in vitro. Similarly, MIP-1α, MIP-1β, and RANTES (CCR5 ligands) were without apparent effect. After addition of chemokines, the number of colonies, their hemoglobinization, and expression of GPA-A remain unchanged. At the same time, however, the growth of the erythroid colonies was strongly inhibited after an addition of TGF-β1 to the cultures. Similar data also were obtained when BFU-E growth was stimulated with a suboptimal dose of EPO + KL (not shown).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The role of HIV-related chemokines in human erythropoiesis is important for two reasons. First, HIV-related chemokines are elaborated in the serum of AIDS patients and thus may influence human erythropoiesis [1, 5]. Second, the therapeutic targeting of HIV-related chemokine receptor-chemokine axes, e.g., by employing receptor agonists or blocking mAbs [1], could potentially affect the proliferation of human erythroid progenitors in HIV-infected patients.

We found that CXCR4 is highly expressed on human CD34+ cells enriched for BFU-E. In this study, we also phenotyped human CD34+ cells for the expression of eight other chemokine receptors using FACS. We found that a small population of CD34+ cells (5%-10%) expresses the CCR6 protein, which binds MIP-3α. Other receptors were undetectable. Hence, CXCR4 remains the most highly expressed member of this receptor family on human CD34+ cells. This observation supports other reports demonstrating that CXCR4 plays an important and privileged function in homing and trafficking of human CD34+ progenitor/stem cells [11, 13, 15, 17, 35]. Our failure to detect CCR1, CCR3, and CCR5 proteins on CD34+ cells is intriguing because all these receptors bind chemokines which interfere with R5 HIV-1 entry into the cells (MIP-1α, MIP-1β, RANTES) [10]. In addition, CCR1 and CCR5 bind MIP-1α, which has been found to inhibit proliferation of CD34+ cells [8].

To validate these observations on a population of pure erythroid cells, we phenotyped human erythroid cells expanded ex vivo as well as mature erythrocytes. We found that the expression of CXCR4 varied with the state of erythroid maturation and could not be detected on BFU-E-derived cells expanded for six days. Nevertheless, these more differentiated erythroid cells still contained CXCR4 mRNA. In contrast, human erythroid cells expressed CCR6 protein, and to a lesser degree CXCR2, after six days of culture. Expression of CCR6 and CXCR2 decreased as the erythroid cells matured. Finally, we looked for the presence of chemokine receptors on mature human peripheral blood erythrocytes but were unable to detect any of the chemokine receptors for which we tested.

In functional assays, we found that SDF-1 (CXCR4 ligand), but not MIP-1α, MIP-1β, RANTES (CCR5 ligand), IL-8 (CXCR2 ligand) or MIP-3α (CCR6 ligand) induced calcium flux and phosphorylated MAPK ERK-1 and ERK-2 and AKT in a FACS-purified CD34+KIT+ population. Moreover, only SDF-1-induced calcium flux in human erythroid cells expanded for two to four days. The response to SDF-1 disappeared when the cells began to differentiate and express GPA-A. In contrast, even though more differentiated erythroid cells expressed CCR6 and CXCR2 receptors, their corresponding ligands (MIP-3α and IL-8) did not activate calcium flux in those cells. We also confirmed that of all the chemokines tested, only SDF-1 is a strong chemotactic factor for human CD34+ cells [11, 13] as well as for human CD34+ BFU-E and for the more differentiated CD34 CFU-E progenitor cells. Of note, none of the chemokines evaluated in this study affected cloning efficiency or maturation of human BFU-E. These data are in apparent conflict with other reports which found that MIP-1α inhibits human erythroid colony formation [8, 10].

The contradictory responses to MIP-1α reported here might be explained by differences in MIP-1α protein purification, differences in the cloning system itself and in the formation of inactive dimers [29]. This latter possibility is unlikely because the same preparations of MIP-1α-activated calcium flux in our study and chemotaxis in certain human hematopoietic cell lines (not shown). This suggests that the source of MIP-1α we utilized in our experiments was biologically active. In addition, the reported inhibitory effects of MIP-1α on hematopoiesis are often unclear [6]. Mice with MIP-1α knock-out do not show any hematopoietic defects [28]. In some reports MIP-1α was found to be only inhibitory if added at a very low dose [29], whereas in others was reported to inhibit cell growth if added at a very high dose and repetitively added doses [31].

Finally, after we recently demonstrated that human BM CD34+ cells secrete MIP-1α endogenously [25], we concluded that the role of this chemokine in regulating human hematopoiesis is more complex than we initially thought [6, 8, 29]. In light of our recent data we conclude that human BM CD34+ cells are exposed in vivo to low levels of secreted MIP-1α and other chemokines [25].

Based on the assays carried out in this study (chemotaxis, calcium flux, phosphorylation of MAPK and AKT), we concluded that only the CXCR4 receptor is functional on the earliest human erythroid cells. It appears to play an important role in regulating chemotaxis and perhaps in homing and retention of erythroid progenitors in the BM microenvironment. Therefore, the perturbation of the SDF-1/CXCR4 axis which occurs during HIV infection or when induced by therapeutic agents designed to block HIV cell entry (e.g., ALX40) [1, 36] may affect the homing and development of erythroid cells. Since we did not observe any effect of chemokines on erythroid colony formation by human CD34+ cells, we conclude that the influence of chemokines on human hematopoiesis is more complicated than initially thought. Further studies are needed to elucidate their role in human hematopoiesis, including examining their autocrine effect on individual hematopoietic lineages [25] as well as their indirect influence as mediated by interactions with hematopoietic accessory cells.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We would like to thank R&D Systems for providing some of the chemokine receptor antibodies used in this study. This work was supported by a grant NIH R01 HL61796-01 to Dr. MZR and NIH DK52558-A2 to Dr. AMG.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Cairns JS, D'Souza MP. Chemokines and HIV-1 second receptors: the therapeutic connection. Nat Med 1998; 4:563568.
  • 2
    Fauci AS. Host factors and the pathogenesis of HIV-induced disease. Nature 1996; 384:529534.
  • 3
    Hoxie JA. Hematologic manifestations of AIDS. In HoffmanR, BenzEJ, ShatillSJ et al., eds. Hematology Basic Principles and Practice. New York: Churchill Livingstone, 1995; 21712200.
  • 4
    Lee B, Doranz BJ, Ratajczak MZ, et al. An intricate web: chemokine receptors, HIV-1 and hematopoiesis. Stem Cells 1998; 16:7988.
  • 5
    Moses A, Nelson J, Bagby GC. The influence of human immunodeficiency virus-1 on hematopoiesis. Blood 1998; 91:14791495.
  • 6
    Rollins BJ. Chemokines. Blood 1997; 90:909928.
  • 7
    Lee B, Ratajczak J, Doms RW, et al. Coreceptor/chemokine receptor expression on human hematopoietic cells: biological implications for HIV-1 infection. Blood 1999; 93:11451156.
  • 8
    Broxmeyer HE, Sherry B, Cooper S, et al. Comparative analysis of the human macrophage inflammatory protein family of cytokines (chemokines) on proliferation of human myeloid progenitor cells. J Immunol 1993; 150:34483458.
  • 9
    Su S, Mukaida N, Wang J, et al. Inhibition of immature erythroid progenitor cell proliferation by macrophage inflammatory protein-1α by interacting mainly with a C-C chemokine receptor, CCR1. Blood 1997; 90:605611.
  • 10
    de Wynter EA, Durig J, Cross MA, et al. Differential response of CD34+ cells isolated from cord blood and bone marrow to MIP-1α and the expression of MIP-1α receptors on these immature cells. Stem Cells 1998; 16:349356.
  • 11
    Aiuti A, Webb IJ, Bleul C, et al. The chemokine SDF-1 is a chemoattractant for human CD34+ hematopoietic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood. J Exp Med 1997; 185:111120.
  • 12
    Bleul CC, Fuhlbridge RC, Casanovas JM, et al. A highly efficacious lymphocyte chemoattractant, stromal cell-derived factor 1 (SDF-1). J Exp Med 1996; 184:11011109.
  • 13
    Kim CH, Broxmeyer HE. In vitro behavior of hematopoietic progenitor cells under the influence of chemoattractants: stromal cell-derived factor-1, steel factor, and the bone marrow environment. Blood 1998; 91:100110.
  • 14
    Ma Q, Jones D, Borghesani PR, et al. Impaired B-lymphopoiesis, myelopoiesis, and derailed cerebellar neuron migration in CXCR4 and SDF-1 deficient mice. Proc Natl Acad Sci USA 1998; 95:94489453.
  • 15
    Mohle R, Bautz F, Rafii S, et al. The chemokine receptor CXCR4 is expressed on CD34+ hematopoietic progenitors and leukemic cells and mediates transendothelial migration induced by stromal cell-derived factor-1. Blood 1998; 91:45234530.
  • 16
    Nagasawa T, Hirota S, Tachibana K, et al. Defects of B-cell lymphopoiesis and bone-marrow myelopoiesis in mice lacking the CXC chemokine PBSF/SDF-1. Nature 1998; 382:635638.
  • 17
    Zou Y, Kottmann AH, Kurod M, et al. Function of the chemokine receptor CXCR4 in haematopoiesis and in cerebellar development. Nature 1998; 393:595599.
  • 18
    Ratajczak J, Zhang Q, Pertusini E, et al. The role of insulin (INS) and insulin like growth factor-I (IGF-I) in regulating human erythropoiesis. Studies in vitro under serum free conditions—comparison to other cytokines and growth factors. Leukemia 1998; 12:371381.
  • 19
    Ratajczak J, Marlicz W, Machalinski B, et al. An improved serum free system for cloning human “pure” erythroid colonies. The role of the different growth factors and cytokines on BFU-E formation by the bone marrow and cord blood CD34+ cells. Folia Histochem Cytobiol 1998; 36:5560.
  • 20
    Ratajczak MZ, Plecher ChH, Marlicz W, et al. CD34+, Kit+, Rhodamine123low phenotype identifies a marrow cell population highly enriched for human hematopoietic stem cells. Leukemia 1998; 12:942950.
  • 21
    Kowalska MA, Ratajczak J, Hoxie J, et al. Platelet and megakaryocytes express the HIV co-receptor CXCR4 on their surface but do not respond to stromal derived factor (SDF)-1. Br J Haematol 1999; 104:220229.
  • 22
    Abe R, Vandenberghe P, Craighead N, et al. Distinct signal transduction in mouse CD4+ and CD8+ splenic T cells after CD28 receptor ligation. J Immunol 1995; 154:985997.
  • 23
    Rocca JDG, Biesen T, Daaka Y, et al. Ras-dependent mitogen-activated protein kinase activation by G protein-coupled receptors. Convergence of Gi- and Gq- mediated pathways on calcium/calmodulin, Pyk2, and Src kinase. J Biol Chem 1998; 272:1912519132.
  • 24
    De Jong MO, Wagemaker G, Wognum AW. Separation of myeloid and erythroid progenitors based on expression of CD34 and c-kit. Blood 1995; 86:40764085.
  • 25
    Majka M, Rozmyslowicz T, Lee B, et al. Bone marrow CD34+ cells and megakaryoblasts secrete β-chemokines; implications for infectability by M-tropic human immunodeficiency virus (R5 HIV). J Clin Invest 1999; 104:17391749.
  • 26
    Dutt P, Wang JF, Groopman JE. Stromal cell-derived factor-1α and stem cell factor/kit ligand share signaling pathways in hemopoietic progenitors: a potential mechanism for cooperative induction of chemotaxis. J Immunol 1998; 161:36523658.
  • 27
    Cashman JD, Eaves CJ, Sarris AH, et al. MCP-1, not MIP-1α, is the endogenous chemokine that cooperates with TGF-β to inhibit the cycling of primitive normal but not leukemic (CML) progenitors in long-term human marrow cultures. Blood 1998; 92:23382344.
  • 28
    Cook DN, Beck MA, Coffman TM, et al. Requirement of MIP-1α for an inflammatory response to viral infection. Science 1995; 269:15831585.
  • 29
    Mantel C, Kim YJ, Cooper S, et al. Polymerization of murine macrophage inflammatory protein 1α inactivates its myelosuppressive effects in vitro: the active form is monomer. Proc Natl Acad Sci USA 1993; 90:22322236.
  • 30
    Miller JS, McCullar V, Verfaille CM. Ex vivo culture of CD34+/Lin/DR cells in stroma-derived soluble factors, interleukin-3, and macrophage inflammatory protein-1α maintains not only myeloid but also lymphoid progenitors in a novel switch culture assay. Blood 1998; 91:45164522.
  • 31
    Patel VP, Kreider BL, Li Y, et al. Molecular and functional chracterization of two novel human C-C chemokines as inhibitors of two distinct classes of myeloid progenitors. J Exp Med 1997; 185:11631172.
  • 32
    Sanchez X, Cousins-Hodges B, Aguilar T, et al. Activation of HIV-1 coreceptor (CXCR4) mediates myelosuppression. J Biol Chem 1997; 272:2752927531.
  • 33
    Gewirtz AM, Zhang J, Ratajczak J, et al. Chemokine regulation of human megakaryocytopoiesis. Blood 1995; 86:25592567.
  • 34
    Zauli G, Vitale M, Gibellini D, et al. Inhibition of purified CD34+ hematopoietic progenitor cells by human immunodeficiency virus 1 or gp 120 mediated by endogenous transforming growth factor β1. J Exp Med 1996; 183:99108.
  • 35
    Wang JF, Liu ZY, Groopman JF. The β-chemokine receptor CXCR4 is expressed on the megakaryocytic lineage from progenitor to platelets and modulates migration and adhesion. Blood 1998; 92:756764.
  • 36
    Doranz BJ, Govit-Ferbas K, Sharron MP, et al. A small molecule inhibitor directed against the chemokine receptor CXCR4 prevents its use as an HIV-1 coreceptor. J Exp Med 1997; 186:13951400.