6191.Full-Text.Pdf

Macrophage-Inflammatory Protein-1α Receptor Expression on Normal and Chronic Myeloid Leukemia CD34+ Cells

This information is current as Sian E. Nicholls, Guy Lucas, Gerard J. Graham, Nigel H. of September 26, 2021. Russell, Rachel Mottram, Anthony D. Whetton and Anne-Marie Buckle J Immunol 1999; 162:6191-6199; ; http://www.jimmunol.org/content/162/10/6191 Downloaded from

References This article cites 41 articles, 27 of which you can access for free at: http://www.jimmunol.org/content/162/10/6191.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 26, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Macrophage-Inﬂammatory Protein-1␣ Receptor Expression on Normal and Chronic Myeloid Leukemia CD34؉ Cells1

Sian E. Nicholls,* Guy Lucas,† Gerard J. Graham,‡ Nigel H. Russell,§ Rachel Mottram,* Anthony D. Whetton,* and Anne-Marie Buckle2*

We have assessed expression of MIP-1␣ binding sites on the surface of CD34؉ cells from normal bone marrow (NBM) and chronic myeloid leukemia (CML) peripheral blood. This study has highlighted a small subpopulation of CD34؉ (15.7 ؎ 6.2% in NBM and -in CML), which has specific macrophage-inflammatory protein-1␣ (MIP-1␣) cell surface binding sites. Further pheno 4% ؎ 9 typic characterization of the receptor-bearing cells has shown that they do not express the Thy-1 Ag, suggesting that they are committed progenitor cells rather than CD34؉ Thy؉ stem cells. However, more than 80% of methanol-fixed CD34؉ cells do bind MIP-1␣, suggesting that these cells may possess a pool of internal receptors, although we were unable to induce cell surface ؉ ؉ expression by cytokine stimulation. The percentage of these CD34 , MIP-1␣-R cells present in the CD34 compartment of NBM Downloaded from is significantly higher than in CML, implicating lack of binding sites as part of the mechanism for the loss of response to this chemokine seen in CML. Specific Ab to the MIP-1␣ receptor implicated in HIV infection, CCR5, revealed that very few CD34؉ cells expressed these receptors and that expression was confined to the CD34؉ Thy؊ progenitor population. Data presented in this work suggest that active binding sites for the stem cell growth inhibitor MIP-1␣ are not constitutively expressed on the surface of most resting primitive multipotent cells, and that these cells are not potential targets for HIV-1 infection through CCR5. The

Journal of Immunology, 1999, 162: 6191–6199. http://www.jimmunol.org/

acrophage-inflammatory protein-1␣ (MIP-1␣)3 is a accessory cells. Effects of MIP-1␣ on more primitive stem cell member of the chemokine family of cytokines that populations have also been described both in inhibition of stem M have been shown to elicit a wide variety of effects on cell proliferation (6) and in maintenance of the stem cell compart- cells of the immune system, including adhesion, chemoattraction, ment under conditions promoting differentiation (7–9). However, and stimulation of Ig synthesis, thereby defining chemokines as since stem cells have a requirement for stromal feeder cells, it is important mediators of the inflammatory response (1). not clear whether the action of MIP-1␣ is directly on the stem In addition to these shared chemokine activities, MIP-1␣ has cells, or an indirect effect via the stroma. by guest on September 26, 2021 been reported to act as an inhibitor of the proliferation of murine Chemokines are divided into four groups, C, CC, CXC, and hemopoietic stem cells both in vivo and in vitro (2, 3); indeed CX3C, based on the number and arrangement of conserved cys- MIP-1␣ was originally described as a murine stem cell inhibitor teines (10, 11). MIP-1␣ is a CC chemokine, for which multiple and has subsequently been shown to inhibit epidermal stem cells human G protein-coupled seven-membrane-spanning receptors ␣ (4). Although both inhibitory and stimulatory effects of MIP-1 on have been cloned, and termed CCR (12). These receptors have the proliferation of human stem cells have been reported according complex patterns of ligand binding, whereby they display both ␣ to the specific cytokine conditions of the assay (3), MIP-1 is specificity and promiscuity. For example, human chemokine re- known to inhibit the formation of methylcellulose colonies from ceptors CCR1 and CCR5 bind MIP-1␣, but also bind other mem- highly purified progenitors (5). Clonogenic studies suggest that bers of the CC group, such as MIP-1␤ and RANTES. Furthermore, ␣ MIP-1 is acting directly on progenitors rather than indirectly via CCR1 additionally binds monocyte-chemotactic protein 3 (12, 13), whereas CCR5 binds monocyte-chemotactic protein 2 (14). It has also been shown that cross-desensitization can occur among the *Leukemia Research Fund Cellular Development Unit, University of Manchester chemokine receptors, although the underlying mechanisms of re- Institute of Science and Technology (UMIST), Manchester, United Kingdom; †De- partment of Clinical Hematology, Manchester Royal Infirmary, Manchester, United ceptor internalization and signal transduction are not well charac- Kingdom; ‡The Beatson Institute for Cancer Research, Glasgow, United Kingdom; terized (15, 16). Recently, an additional high affinity receptor for § and Department of Haematology, Nottingham City Hospital, Nottingham, United ␣ Kingdom MIP-1 , known as D6, has been identified in the mouse (17) and Received for publication September 16, 1998. Accepted for publication March in humans (18). There is currently much interest in understanding 1, 1998. how specific receptor engagement relates to individual responses The costs of publication of this article were defrayed in part by the payment of page such as chemotaxis or growth inhibition, although there is now charges. This article must therefore be hereby marked advertisement in accordance emerging evidence in the murine system that some MIP-1␣ recep- with 18 U.S.C. Section 1734 solely to indicate this fact. tors such as CCR1 and CCR4 may be related to inflammation, 1 This work was funded by the Leukaemia Research Fund (U.K.). whereas others such as CCR5 and the novel receptor, D6, may be 2 Address correspondence and reprint requests to Dr. Anne-Marie Buckle, Leukaemia important in the control of proliferation (17). The recent finding Research Fund Cellular Development Unit, Department of Biomolecular Sciences, UMIST, Sackville Street, Manchester, U.K. M60 1QD. E-mail address: that several chemokine receptors may act with CD4 as coreceptors [email protected] for HIV-1 infection suggests that these proteins may play a role in 3 Abbreviations used in this paper: MIP, macrophage-inflammatory protein; APC, controlling the progression of AIDS (19). CCR5, CCR2b, and to a allophycocyanin; bMIP, biotinylated MIP; CML, chronic myeloid leukemia; HI, heat- inactivated; MNC, mononuclear cell; NBM, normal bone marrow; NCS, newborn calf lesser extent CCR3 have been shown to serve as coreceptors for serum; PI, propidium iodide; SAv, streptavidin. HIV infection of macrophage-tropic HIV-1 (20–22), whereas

CXCR4 functions as a coreceptor for T cell-trophic HIV strains CCR5 were kindly provided by Dr. Charles Mackay (Leukosite, Cam- (23). The detection of CD4 expression on some CD34ϩ cells (24) bridge, MA) and were used in conjunction with FITC anti-mouse (Phar- has raised the possibility that these cells may be a target for HIV-1 Mingen). infection. Although it is not clear whether primitive stem cells do Labeling for MIP-1␣ binding sites indeed constitute a potential reservoir for HIV-1, the detection by PCR of the CXCR4 receptor suggests that they may be a target for T MNC, prelabeled with anti-CD34 APC and anti-Thy-1 PE, were assayed for the presence of MIP-1␣ binding sites using the method provided with cell-trophic strains (25), although no conclusions could be drawn from a Fluorokine kit purchased from R&D Systems. Briefly, cells were incu- the PCR data generated in this study about CCR5 expression (25). bated with a biotinylated human rMIP-1␣ at 4°C for 60 min. The cells were Since cancer chemotherapy is dose limited by the damage in- then incubated (without washing) with streptavidin-fluorescein (SAv- flicted on proliferating cells within the bone marrow, members of FITC). The kit includes a negative control Ab (a soybean trypsin inhibitor protein that has been biotinylated to the same extent as the MIP-1␣) and a the chemokine family are under consideration in strategies aimed MIP-1␣-blocking Ab (a polyclonal goat IgG anti-human MIP-1␣ Ab). The at the chemoprotection and mobilization of normal bone marrow competable binding of the bMIP-1␣ was assessed using a 50-fold excess of (NBM) stem and progenitor cells (26). There is now evidence in unlabeled BB-10010 (British Biotech Phamaceuticals, Oxford, U.K.), diseases such as CML (6, 27, 28), acute myelogenous leukemia which is a well-characterized mutant of MIP-1␣ with a single amino acid ␣ (29), and some forms of cancer (30) that chemokines may not inhibit substitution (33), or a 50-fold excess of the R&D Systems MIP-1 of the same composition as that which is biotinylated in the Fluorokine kit. Cells proliferation of leukemic or cancer cells. The possibility that growth were incubated at 4°C for 60 min in bMIP-1␣ premixed with unlabeled inhibition is selective for normal cells presents chemokines as partic- MIP-1␣, then labeled with SAv-FITC. All samples were labeled with 200 ularly attractive candidates for chemoprotective therapies. ␮g/ml propidium iodide (PI) (Molecular Probes, Eugene, OR), for dead ␣ In this study, we have further defined the role of MIP-1␣ and its cell exclusion, at the end of the bMIP-1 labeling. Using this assay, only Downloaded from cells expressing a specific MIP-1␣ receptor can be identified, and the in- receptors in the inhibition of stem cell proliferation by analyzing tensity of the staining is directly proportional to the receptor density. the binding of biotinylated MIP-1␣ (bMIP-1␣) to human CD34- positive bone marrow stem cells. The use of biotinylated cytokines Fixation and acid washing of samples to study cytokine receptor expression pattern on rare populations In some experiments, cells were fixed in 70% cold methanol overnight. of human stem cells has been successfully demonstrated by Wog- Cells were then washed and resuspended in HBSS ϩ 5% NCS (HI) for 30

num et al. (31, 32). We have now adopted a similar approach using min before Ab labeling. The acid wash procedure consisted of two washes http://www.jimmunol.org/ bMIP-1␣ in a multiparameter staining protocol that enables us to in HBSS ϩ 5% NCS (HI), resuspension of the cells in PBS, pH 3, for 1 min ␣ exactly at 4°C, then further washing before resuspending in HBSS ϩ 5% examine the MIP-1 -binding ability of specific human hemopoi- NCS (HI). All of the above cell preparations were then kept at 4°C for Ab etic progenitor cell populations. This approach not only has the labeling. advantage of identifying small subpopulations of MIP-1␣ receptor-bearing progenitor cells, but also is not restricted to identifying Flow cytometry any one class of receptor and will therefore detect all members of Cells were visualized using a dual laser FACSVantage (Becton Dickinson) the known CCR family that bind MIP-1␣, as well as novel MIP-1␣ flow cytometer with excitation of FITC, PE, and PI from an Argon laser, binding sites. This is of particular significance in the case of he- and of APC using a HeNe laser. Data were collected and analyzed using mopoietic stem cells, since there may be novel receptors on these CellQuest software. by guest on September 26, 2021 cells yet to be identified. In addition, we have used Ab to the Immunocytochemistry specific receptor CCR5 that is implicated in the control of stem cell ϩ Cytospin preparations of THP-1 cells were made using a Shandon Cytospin proliferation and in the infection of CD34 cells by HIV-1 to ϫ ϩ ϩ at 500 g for 5 min. Slides of CD34 cells from NBM and CML pe- define CCR5 expression patterns on CD34 cells. ripheral blood were prepared by sorting 3000 cells directly onto micro- scope slides using the flow cytometer. Slides were air dried before fixation Materials and Methods for 10 min in methanol (prechilled to Ϫ20°C), then allowed to dry at room Isolation of mononuclear cells from clinical samples temperature. Mononuclear cells (MNC) were prepared from samples of NBM from con- Results senting donors and peripheral blood from patients with CML at presenta- ␣ tion using Ficoll-Paque density-gradient separation (d ϭ 1.077) (Pharma- Validation for specific binding of bMIP-1 to viable human cia, Uppsala, Sweden). MNC were washed and kept in HBSS hemopoietic cells supplemented with 5% heat-inactivated newborn calf serum (HBSS ϩ 5% NCS (HI)) (Life Technologies, Paisley, Scotland) at 4°C for all further cell The human erythroleukemic TF-1 cell line and the myelomono- manipulations. cytic cell line THP-1 have both been reported to bind MIP-1␣ (34). To determine whether the specific binding of bMIP-1␣ could be Cell lines quantified using viable cell populations, we investigated the bind- The TF-1 cell line was maintained in RPMI 1640 (Life Technologies) ϩ ing of bMIP-1␣ to TF-1 and THP-1 cells. The histogram plots A ϩ 10% FCS 2 ng/ml GM-CSF (R&D Systems, Minneapolis, MN). The and B in Fig. 1 show that there was a significant level of bMIP-1␣ THP-1 cell line was maintained in RPMI 1640 ϩ 10% FCS. HEK293 cells transfected with human CCR5 and the parent HEK293 cells were main- binding over that seen with the control biotinylated protein for tained in DMEM (Life Technologies) and 10% FCS, and were passaged both cell lines. Furthermore, anti-human MIP-1␣ mixed with using trypsin EDTA (Life Technologies). bMIP-1␣ before usage with TF-1 or THP-1 cells totally abrogated Ab staining the specific binding observed. The upward shift in fluorescence shown in TF-1 cells that bind bMIP-1␣ indicates that the majority MNC were incubated for 20 min at 4°C with Ab to CD34 conjugated to PE, of the homogeneous cell population binds bMIP-1␣. As a further or allophycocyanin (APC) (Becton Dickinson, Oxford, U.K.). In some experiments, cells were additionally stained with Ab to Thy-1 conjugated to control, a competable binding experiment was performed in which PE (PharMingen, San Diego, CA). Negative controls using isotype- a 50-fold excess of either BB-10010 or human rMIP-1␣ was used. matched Abs conjugated to APC and to PE were routinely performed. After The binding of bMIP-1␣ to TF-1 cells was competed by both staining, cells were washed and resuspended in HBSS ϩ 5% NCS for forms of MIP-1␣, confirming the specificity of the binding assay staining with bMIP-1␣ (see below). Rabbit polyclonal anti-human CCR-5 sera and preimmune control sera were kind gifts from Jane McKeating at (see Fig. 1C). It is important to note that the binding assay is the University of Reading, and were used in conjunction with FITC sheep performed at 4°C, thus minimizing intracellular uptake of labeled anti-rabbit (The Binding Site, Birmingham, U.K.). mAbs to CCR3 and receptors. The ability of bMIP-1␣ to detect CCR5 receptors was The Journal of Immunology 6193 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 1. Validation of the Fluorokine assay using TF-1 and THP-1 cell lines. TF-1 and THP-1 cells were thoroughly washed to remove medium, serum, and growth factors. Cells were labeled with bMIP-1␣, as described in Materials and Methods. Histogram plots A and B show the level of bMIP-1␣ binding to TF-1 and THP-1 cells, respectively (solid line), as compared with the negative control (dotted line). Speciﬁcity was demonstrated using an anti-MIP-1␣ Ab (dashed line). These results using the TF-1 cell line are also presented in the dot plots I, II, and III. The upward shift in FL-1 (bMIP- 1␣-SAv-FITC) seen indicates that the majority of cells bound bMIP-1␣. Plot C shows the competable binding of bMIP-1␣ (solid line) on TF-1 cells by a 50-fold excess of BB-10010 or MIP-1␣ (two dashed lines), as compared with the control level (dotted line). Plot D shows the binding of bMIP-1␣ to HEK293 cells transfected with CCR5 (solid line) as compared with the control level (dotted line) or level in the presence of blocking Ab (dashed line).

confirmed by its ability to bind specifically to the HEK293 cell line which freshly isolated bone marrow or peripheral blood MNC transfected with CCR5 (Fig. 1D). could be labeled with anti-CD34 Ab conjugated to APC and anti- Having validated the detection of bMIP-1␣ binding using flow Thy-1 Ab conjugated to PE, and also, importantly, stained with PI cytometry, bMIP-1␣ binding sites on normal and CML progenitor for dead cell exclusion. Cells were assessed for the presence of cells were assessed. bMIP-1␣ binding sites by labeling with bMIP-1␣, which was de-

ϩ tected by SAv-FITC. Fig. 2 presents data from seven NBM MNC ␣ ϩ MIP-1 binding sites on normal CD34 cells preparations (mean Ϯ SEM) showing the percentage of CD34 To study bMIP-1␣ binding on viable primitive cells, which are cells that express bMIP-1␣ binding sites as compared with the extremely rare in bone marrow and peripheral blood samples, and biotinylated negative control. It was found that 15.7 Ϯ 6.2% of the which can be lost during complex purification procedures, we seven normal CD34ϩ cells expressed MIP-1␣ binding sites adopted a strategy of using MNC preparations with no further (mean Ϯ SEM). The flow-cytometric profile of a typical sample is purification. We established a four-color staining procedure in shown in Fig. 3A. The level of binding site expression over and 6194 MIP-1␣ RECEPTORS ON CD34ϩ CELLS

NBM expresses the Thy-1 Ag with a view to determining whether MIP-1␣ expression is dependent upon the stage of development of hemopoietic cells. We can detect Thy-1 expression on CD34ϩ cells from NBM (as has been described previously in Ref. 35); however, when Thy-1 and bMIP-1␣ staining was assessed, it was found that the majority of the bMIP-1␣-staining population of cells did not coincide with the Thyϩ population, thereby demonstrating that MIP-1␣-Rϩ cells are generally ThyϪ (see Fig. 4). MIP-1␣ binding sites on fixed cells The work described above was performed on freshly isolated MNC from bone marrow or peripheral blood. These data are, however, somewhat different from a previous study in which the expression of MIP-1␣ receptors on fixed MNC was reported (28) and Ͼ90% of CML CD34ϩ cells were shown to express MIP-1␣ binding sites. We therefore studied the binding of bMIP-1␣ on fixed MNC compared with unfixed, viable cells. Fig. 5 shows that in a typical experiment in which an aliquot of CML MNC was fixed ϩ before Ab labeling, 87.37% of CD34 cells in the fixed sample Downloaded from bound bMIP-1␣, whereas only 16.9% of the nonfixed CD34ϩ cells expressed MIP-1␣ binding sites. It is possible that MIP-1␣ binding sites at the cell surface may be revealed after fixation because they are already occupied in vivo by MIP-1␣, and that after fixation these sites are vacated, allowing bMIP-1␣ to bind. We have as-

ϩ sessed the possible effect of any residual MIP-1␣ bound to cell http://www.jimmunol.org/ FIGURE 2. MIP-1␣ binding site expression on normal and CML CD34 cells. Seven samples of NBM MNC and ten samples of CML PBMC were surface binding sites in clinical samples by acid washing of fresh ␣ labeled for CD34 expression and bMIP-1␣ binding. Cells were labeled with CML and normal MNC before the bMIP-1 -binding assay. It was anti-CD34 APC and Thy PE, followed by staining with bMIP-1␣ detected found that this treatment had no effect on the binding of bMIP-1␣ with SAv-FITC. Results are presented as the percentage of CD34ϩ cells that to CML and normal MNC, in terms of either the percentage of cells displayed positive binding (mean Ϯ SEM) when incubated with control Ab, binding bMIP-1␣ or the level of ﬂuorescence intensity (data not bMIP-1␣, or bMIP-1␣, plus blocking Ab. See Fig. 4 for staining patterns of shown). These data strongly suggest that bMIP-1␣ receptor expres- ϩ Thy-1 expression. The percentage of CD34 cells that displayed positive bind- sion at the cell surface is observed in only a fraction of cells, although ing, as compared with the negative control, is shown to be signiﬁcant on both ␣ ϩ almost all express bMIP-1 binding sites within the cells, as revealed

Ͻ Ͻ by guest on September 26, 2021 normal and CML CD34 cells: p 0.001 and p 0.00001, respectively; and by permeabilization through fixation. the percentage of CD34ϩ cells expressing MIP-1␣ binding sites is shown to be significantly greater in NBM than in CML (p Ͻ 0.025). Data were analyzed Subcellular localization of bMIP-1␣ binding sites in fixed cells using a Student’s t test. The subcellular location of the bMIP-1␣ binding sites observed after fixation of cells was investigated further through the use of deconvolution microscopy. Immunochemical staining of THP-1 above the negative control was highly significant ( p Ͻ 0.001), and cells showed evidence of cell surface binding and also some weak binding was specific, as demonstrated by the anti-MIP-1␣ Ab cytoplasmic binding, suggesting that in addition to the cell surface (Figs. 2 and 3A). binding sites detected on fresh cells by flow cytometry, there is an ϩ intracellular pool of bMIP-1␣ binding sites (Fig. 6, a and b). The MIP-1␣ binding sites on viable CML CD34 cells sorted CD34ϩ cells from NBM (data not shown) and CML pe- Samples of 10 CML peripheral blood MNC were stained with ripheral blood (Fig. 6, c and d) also gave positive staining by anti-CD34 PE, followed by bMIP-1␣ labeling, as in the above immunocytochemistry. The staining seen on CD34ϩ cells was experiments on NBM samples. As can be seen in Figs. 2 and 3B, more patchy than that seen with THP-1 cells, suggesting clustering ϩ there is bMIP-1␣ binding to CML CD34 cells. The percentage of of the binding moiety. CD34ϩ cells have a high nuclear:cytoplas- ϩ CML CD34 cells that bind bMIP-1␣,9Ϯ 4% (mean Ϯ SEM), is, mic ratio that, combined with the patchy staining, made discrim- however, significantly lower ( p Ͻ 0.025) than that on normal ination between cell surface and cytoplasmic staining less clear. ϩ CD34 cells. As with NBM, the level of binding was significantly However, as can be seen in Fig. 6, c and d, there was evidence of above the control value ( p Ͻ 0.00001), and specificity of MIP-1␣ cell surface staining with binding concentrated in a tight ring at the binding was demonstrated using a blocking Ab (Figs. 2 and 3B). outer edge of the cell.

MIP-1␣ binding site expression on CD34ϩ subpopulations Identification of specific MIP-1␣ receptors We next characterized further the subpopulation of CD34ϩ cells Ab specific for the MIP-1␣ receptor CCR5 was used to stain MNC that express MIP-1␣ binding sites. Unlike the binding of bMIP-1␣ isolated from NBM and CML peripheral blood. As shown in Table by the homogeneous cell lines TF-1 and THP-1, bMIP-1␣ ap- I, 19 Ϯ 9.8% of CD34Ϫ cells in NBM expressed detectable levels peared to bind to a subpopulation of CD34ϩ cells from NBM or of CCR5. While the majority of CD34ϩ cells were found not to CML peripheral blood. Previous studies on the phenotypic char- express CCR5, staining was seen on a small subpopulation (4.9 Ϯ acterization of stem cells suggested that CD34ϩ Thyϩ cells are 2.6%). As was seen in the previous three-color staining experi- more primitive than CD34ϩ ThyϪ cells. We therefore used a three- ments with bMIP-1␣, these receptor-expressing cells were con- color multiparameter analysis of CD34, Thy-1, and bMIP-1␣ to fined to the CD34ϩ ThyϪ progenitor population of cells, and no phenotypically assess whether the CD34ϩbMIP-1␣ϩ population in staining was seen on CD34ϩThyϩ cells (Fig. 7). Abs to CCR3 The Journal of Immunology 6195 Downloaded from http://www.jimmunol.org/

FIGURE 3. Flow-cytometric demonstration of MIP-1␣ binding sites on CD34ϩ cells. Flow-cytometry dot plots show typical labeling of A, NBM CD34ϩ cells, and B, CML peripheral blood CD34ϩ cells with negative control Ab (plots a and d), bMIP-1␣ (plots b and e), and the bMIP-1␣-blocking Ab (plots c and f). Plots b and e show that only a subpopulation of CD34ϩ binds bMIP-1␣, unlike the binding of bMIP-1␣ to the homogeneous cell lines in Fig. 1. by guest on September 26, 2021 were found to stain a small but distinct subpopulation (2.4 Ϯ HEK293 cells transfected with CCR5. However, CD34ϩ cells 1.4%) of CD34Ϫ cells in NBM, whereas only 1.5 Ϯ 0.8% of the from NBM or CML peripheral blood did not stain positively with ϩ CD34 cells stained positively for CCR3. Analysis of cells from this Ab after fixation, suggesting that this is not the receptor that Ϫ CML patients was also performed, and 8.9 Ϯ 3% of CD34 cells binds MIP-1␣ after fixation (Fig. 8). expressed CCR5, although less than 1% of CD34ϩ cells were positive for this receptor. CCR3 was also detected on 4.1 Ϯ 2.3% of Ϫ ϩ CD34 cells and on 2.3 Ϯ 1.1% of CD34 cells. As with NBM, Discussion three-color analysis showed that any CCR3 or CCR5 staining was confined to the CD34ϩ ThyϪ population, with no detectable stain- In the present study, we have examined MIP-1␣ binding sites on ϩ ing of CD34ϩ Thyϩ cells (Fig. 7). normal resting CD34 bone marrow progenitor cells by using a bMIP-1␣ probe. We first used the cell lines TF-1 and THP-1 as well CCR5 expression on fixed cells as cells transfected with CCR5 to validate the use of the MIP-1␣- The Ab specific for CCR5 was shown to recognize a fixation- binding kit. The assay was shown to be specific through the use of an resistant epitope, as determined by the staining of fresh and fixed anti-MIP-1␣ Ab. In addition, we were able to demonstrate specificity

FIGURE 4. MIP-1␣ binding site expression on CD34ϩ Thyϩ and CD34ϩ ThyϪ cells from NBM and CML peripheral blood. Simulta- neous analysis of CD34, MIP-1␣, and Thy expression on NBM cells (A) and CML peripheral blood (B). Plots show bMIP-1␣ binding site and Thy-1 expression on CD34ϩ-gated cells. 6196 MIP-1␣ RECEPTORS ON CD34ϩ CELLS

Table I. Expression of CCR3 and CCR5 on NBM and CML peripheral blood cellsa

CD34Ϫ CD34ϩ

NBM (n ϭ 3) CCR5 19.0 Ϯ 9.8 4.9 Ϯ 2.6 CCR3 2.4 Ϯ 1.4 1.5 Ϯ 0.8

CML (n ϭ 5) CCR5 8.9 Ϯ 3.0 0.5 Ϯ 0.2 CCR3 4.1 Ϯ 2.3 2.3 Ϯ 1.1

aMononuclear cells from normal bone marrow (n ϭ 3) and CML patients (n ϭ 5) were stained with Abs to CD34 combined with CCR3 or CCR5. Data are presented as the percentage of cells showing positive staining in the CD34Ϫ and CD34ϩ com- partments ϮSEM. All data were subjected to a student’s t test. CCR5 expression on CD34ϩ cells in the CML patients was found to be signiﬁcantly different from CCR5 expression on CD34Ϫ cells ( p Ͻ 0.05).

by cold competition using nonbiotinylated wild-type MIP-1␣ and BB- Downloaded from 10010, a stable disaggregated mutant of MIP-1␣ (33). We have found that, as previously reported, the majority of CD34ϩ cells are able to bind bMIP-1␣ after permeabilization by ﬁxation in methanol, implying that all CD34ϩ cells have the potential to respond to MIP-1␣. We have now extended this study to

show that unfixed cells have very little capacity to bind bMIP-1␣, http://www.jimmunol.org/ and that only a small subpopulation of CD34ϩ cells demonstrated significant binding. The flow-cytometric approach used in these ␣ FIGURE 5. MIP-1 receptor staining using fixed CD34 cells from experiments has an important limitation in that we are unable to CML peripheral blood. MIP-1␣ binding was compared on samples before directly quantitate the number of receptors per cell. However, THP-1 and after fixation. CML PBMC were fixed in cold methanol, as described ␣ in Materials and Methods. Histogram A shows the level of bMIP-1␣ iden- cells, which we have clearly shown to express MIP-1 receptors by tified on freshly isolated CML MNC (dotted line), together with the neg- flow cytometry, have previously been shown in radiobinding assays to ative control (solid line) and the blocking Ab (dashed line). B shows the express a single class of high affinity receptor for monomeric MIP-1␣, equivalent bMIP-1␣ binding to fixed MNC. with about 1300 receptors per cell detected (34). Therefore, we may by guest on September 26, 2021

FIGURE 6. Immunochemical staining of sorted CD34ϩ cells. THP and CD34ϩ cells were sorted onto slides and ﬁxed in methanol before staining with the bMIP-1␣ kit. Staining of THP-1 cells in the presence (a) and absence (b) of blocking Ab is shown. Fig. 6 (c) shows the staining of CD34ϩ cells isolated from a CML patient. Magniﬁcation is ϫ40. A more detailed image of c is shown in d. The Journal of Immunology 6197

FIGURE 7. CCR5 and CCR3 expression on CD34ϩ cells from NBM and CML peripheral blood. Plots show CCR3 or CCR5 expression on CD34ϩ- gated cells in relation to Thy-1ϩ cells. Receptor-spe- ciﬁc Abs were used to detect CCR5 staining on NBM (A) and CML (B), and CCR3 on NBM (C) and CML cells (D). Downloaded from http://www.jimmunol.org/

postulate that our detection system is sufficiently sensitive to readily subpopulation of candidate stem cells (35–37). To investigate detect less than 1000 receptors per cell. whether the low percentage of unfixed cells that showed MIP-1␣- Abs to CD34 are known to mark a heterogeneous population of binding capacity was associated with the Thyϩ stem cell compart- by guest on September 26, 2021 hemopoietic progenitor cells encompassing lineage-committed ment, we established a three-color stain for the simultaneous anal- ϩ progenitors, more primitive progenitors, and also the small Thy ysis of CD34, Thy-1, and MIP-1␣ binding site expression. We found that binding ability did not coincide with Thy-1ϩ cells, but rather was confined to a subpopulation of the Thy-1Ϫ population of cells. Thus, binding sites are found on cells with the phenotype of progenitors, but not on stem cells themselves. Presumably the MIP-1␣-binding CD34ϩ subpopulation has substantial colony-forming potential, since high levels of colony inhibition of CD34ϩ cells have been reported with MIP-1␣ treatment (5, 28). These data are consistent with that of others (see Ref. 38 for example) who have suggested that MIP-1␣ inhibits the proliferation of primitive cells that have developed beyond the hemopoietic stem cell compartment, and that the action of MIP-1␣ is direct rather than via accessory cells. The differences in staining profiles between fixed and fresh samples of CD34 cells may have important biological implications. One possibility is that there are receptor sites available within the cytosol that are made accessible by the fixation procedure. Immunochemical staining of TF-1 cells revealed evidence of cell membrane staining and some intracellular staining. Immunochemical analysis of CD34ϩ cells confirmed that these cells bind bMIP-1␣ upon fixation, although since these cells have very little cytoplasm, cell surface expression could not be clearly distinguished from cytoplasmic staining. Cytosolic receptor binding implies that an as yet unidentified stimulatory event will induce cell surface expression. Little is known about the regulation of MIP-1␣ receptor expression, although induc- tion on T lymphocytes has been shown to be affected by IL-2 (39). It FIGURE 8. CCR5 expression on fixed cells. HEK293 cells transfected with CCR5 were stained with anti-CCR5 Ab before and after fixation (A). is possible that combinations of cytokines may prove capable of in- CCR5 staining on unfixed (B) and methanol-fixed (C) CD34ϩ cells from ducing cell surface expression on stem cells. We addressed this ques- NBM is also shown. HEK293 CCR5 transfectants labeled with an irrele- tion by investigating whether stimulation by stem cell factor and GM- vant isotype-matched Ab were used as controls. CSF (growth factors used in one study by Chasty et al. (28), in which 6198 MIP-1␣ RECEPTORS ON CD34ϩ CELLS

ϩ MIP-1␣ inhibited normal CD34 cell proliferation) and also by MIP-1␣) also inhibits clonogenic epidermal keratinocyte proliferation. J. Invest. MIP-1␣ itself induced cell surface expression on CD34ϩ cells. We Dermatol. 101:113. 5. Broxmeyer, H. E., B. Sherry, S. Cooper, L. Li, R. Maze, P. M. Beckmann, were unable to see any changes in cell surface expression. A. Cerami, and P. Ralph. 1993. Comparative analysis of the human macrophage The lack of binding sites detected on the stem cells is however inflammatory protein family of cytokines (chemokines) on proliferation of human puzzling, particularly with respect to the reported expression of the myeloid progenitor cells: interacting effects involving suppression, synergistic ϩ suppression, and blocking of suppression. J. Immunol. 150:3448. MIP receptor CCR1 on the surface of CD34 cells (40, 41), as 6. Eaves, C., C. Udomsakdi, J. Cashman, M. Barnett, and A. Eaves. 1993. The biology detected by specific Ab. It is possible that there is a further level of normal and neoplastic stem cells in CML. Leuk. Lymphoma 11(Suppl. 1):245. 7. Verfaillie, C. M., P. M. Catanzarro, and W. N. Li 1994. Macrophage inflamma- of complexity within the chemokine receptor family, whereby a tory protein 1␣, interleukin 3 and diffusible marrow stromal factors maintain change in conformation of the receptor may be required for effec- human hematopoietic stem cells for at least eight weeks in vitro. J. Exp. Med. tive binding, such as within the integrin family of adhesion mol- 179:643. 8. Verfaillie, C., and J. S. Miller. 1994. CD34ϩ/CD33Ϫ cells reselected from mac- ecules. Indeed, the IL-8R CXCR1 has been reported to undergo a rophage infalammatory protein 1 ϩ interleukin-3-supplemented “stroma-noncon- conformational change after ligand binding that results in activa- tact” cultures are enriched for long-term bone marrow culture. Blood 84:1442. tion of signaling events (42). 9. Verfaillie, C. M., and J. S. Miller. 1995. A novel single-cell prolferation assay ϩ shows that long-term culture-initiating cell (LTC-IC) maintenance over time re- Taken together, these observations suggest that CD34 cells sults from the extensive proliferation of a small fraction of LTC-IC. Blood 86: have the potential to bind MIP-1␣, as evidenced by the staining of 2137. ␣ 10. Oppenheim, J. J., C. O. C. Zacharial, N. Mukaida, and R. Matsushima. 1991. fixed cells. However, the ability of MIP-1 to bind unfixed stem Properties of the novel proinflammatory supergene “intercrine” cytokine family. cells may be subject to regulation of receptor expression or con- Annu. Rev. Immunol. 9:617. formation, such that in the resting state there is very limited ca- 11. Wolpe, S. D., and A. Cerami. 1989. Macrophage inflammatory proteins 1 and 2: members of a novel superfamily of cytokines. FASEB J. 3:2565. ␣ Downloaded from pacity for these cells to bind MIP-1 at the surface. In contrast to 12. Gao, J. L., D. B. Kuhns, H. L. Tiffany, D. McDermott, X. Li, U. Francke, and the binding of MIP-1␣ to human cells, primitive mouse cells ap- P. M. Murphy. 1993. Structure and functional expression of the human macro- pear to bind well with almost all cells staining positively (A. phage inflammatory protein 1␣/RANTES receptor. J. Exp. Med. 177:1421. 13. Neote, K., D. DiGregorio, J. Y. Mak, R. Horuk, and T. J. Schall. 1993. Molecular Buckle, unpublished data). cloning, functional expression, and signaling characteristics of a C-C chemokine This study has raised the possibility that there may be a more receptor. Cell 72:415. complex mechanism underlying the effects of MIP-1␣ on primitive 14. Gong, W. H., O. M. Z. Howard, J. A. Turpin, M. C. Grimm, H. Ueda, P. W. Gray, C. J. Raport, J. J. Oppenheim, and J. M. Wang. 1998. Monocyte chemotactic progenitor cells than previously considered. Control of MIP-1␣ protein-2 activates CCR5 and blocks CD4/CCR5-mediated HIV-1 entry/replica- http://www.jimmunol.org/ activity on these cells may be mediated through the inducibility of tion. J. Biol. Chem. 273:4289. 15. Murphy, P. M. 1994. The molecular biology of leukocyte chemoattractant recep- active form of receptor expression. It will therefore be important to tors. Annu. Rev. Immunol. 12:593. understand more about the regulation of this receptor family in 16. Aramori, I., J. Zhang, S. S. G. Ferguson, P. D. Bieniasz, B. R. Cullen, and normal and Phϩ primitive cells to exploit fully the clinical poten- M. G. Caron. 1997. Molecular mechanism of desensitization of the chemokine ␣ receptor CCR-5: receptor signalling and internalization are dissociable from its tial of MIP-1 as a growth inhibitor. role as an HIV-1 co receptor. EMBO J. 16:4606. Understanding the relationship between receptor expression and 17. Nibbs, R. J. B., S. M. Wylie, I. B. Pragnell, and G. J. Graham. 1997. Cloning and binding capacity of MIP-1␣ is hampered by the lack of reagents to characterization of a novel murine ␤ chemokine receptor D6. J. Biol. Chem. 272:12495. 18. Nibbs, R. J. B., S. M. Wylie, J. Y. Yang, N. R. Landau, and G. J. Graham. 1997. detect receptor expression. PCR analysis of CCR5 in sorted Cloning and characterization of a novel promiscuous human ␤-chemokine recep-

ϩ by guest on September 26, 2021 CD34 cells was inconclusive, with occasional positive staining tor D6. J. Biol. Chem. 272:32078. being observed (25). In this study, we have investigated, using 19. Cocchi, F., A. L. DeVico, A. Garzino-Demo, S. K. Ayra, R. C. Gallo, and P. Lusso. 1995. Identification of RANTES, MIP-1␣ and MIP-1␤ as the major receptor-specific Ab, the expression of the high affinity MIP-1␣ HIV-suppressive factors produced by CD8ϩ T cells. Science 270:1811. receptor CCR5, which is a coreceptor for macrophage-trophic 20. Alkhatib, G., C. Combadiere, C. C. Broder, Y. Feng, P. E. Kennedy, P. M. Murphy, and P. A. Berger. 1996. CC CKR5: a RANTES, MIP-1␣, MIP-1␤ strains of HIV-1 (20–22). In addition, CCR5 has been implicated receptor as a fusion cofactor for macrophage trophic HIV-1. Science 272:1955. in the control of stem cell proliferation (17). Although staining of 21. Deng, H., R. Liu, W. Ellmeier, S. Choe, D. Unutmz, M. Burhkart, P. Di Marzio, MNC could be seen with Ab to CCR5 and also with Ab to a further S. Marmon, R. E. Sutton, C. M. Hill, et al. 1996. Identification of a major coreceptor for primary isolates of HIV-1. Nature 381:661. coreceptor for HIV, CCR3, very little expression of either receptor ϩ 22. Doranz, B. J., J. Rucker, Y. Yi, R. J. Smyth, M. Samson, S. C. Peiper, was observed in the CD34 compartment, and in particular there M. Parmentier, R. G. Collman, and R. W. Doms. 1996. A dual trophic primary was no evidence of staining on CD34ϩ Thyϩ stem cells. These HIV-1 isolate that uses fusin and the ␤-chemokine receptors CKR-5, CKR-3 and ϩ ϩ CKR-2b as fusion cofactors. Cell 85:1149. profiles suggest that resting CD34 Thy cells are not potential 23. Feng, Y., C. C. Broder, P. E. Kennedy, and A. E. Berger. 1996. HIV-1 entry targets for the macrophage-tropic strains of HIV-1 and that the cofactor: functional cDNA cloning of a seven transmembrane, G protein-coupled growth-inhibitory effects of MIP-1␣ are not mediated through receptor. Science 272:872. 24. Zauli, G., G. Furlini, M. Vitale, M. C. Re, D. Gibellini, L. Zamai, G. Visani, CCR5 on resting primitive stem cells. P. Borgatti, S. Capitani, and M. Laplaca. 1994. A subset of human CD34ϩ hematopoietic progenitors express low levels of CD4, the high affinity receptors for human immunodeficiency virus type-1. Blood 84:1896. Acknowledgments 25. Deichmann, M., R. Kronenwett, and R. Haas. 1997. Expression of human immunodeficiency virus type I coreceptors CXCR-4 (fusin, LESTR) and CCR-5 in We thank all clinicians in the North West who supplied us with patient CD34ϩ haemopoietic progenitor cells. Blood 89:3522. samples, and also Dr. Jane Owen-Lynch and Andrew Pierce for helpful 26. Broxmeyer, H. E., N. L. Hague, G. W. Sledge, H. Rasmussen, and M. S. Gordon. discussions, and Furheen Rafiq for establishing the immunocytochemical 1995. Suppression of marrow and mobilization of blood myeloid progenitors in staining method for MIP-1␣. vivo by BB-1001010010, a genetically engineered variant of human macrophage inflammatory protein (MIP)-1-␣, in a phase-1 clinical trial in patients with re- lapsed/refractory breast cancer. Blood 86:37. References 27. Eaves, C. J., J. D. Cashman, S. D. Wolpe, and A. C. Eaves. 1993. Unrespon- siveness of primitive chronic myeloid leukemia cells to macrophage inflamma- 1. Schall, T. J. 1994. The chemokines. In The Cytokine Handbook, 2nd Ed. A. tory protein 1␣, an inhibitor of primitive normal hematopoietic cells. Proc. Natl. Thompson, ed. Academic Press, London, p. 419. Acad. Sci. USA 90:12015. 2. Graham, G. J., E. G. Wright, R. Hewick, S. D. Wolpe, N. M. Wilkie, 28. Chasty, R. C., G. S. Lucas, P. J. Owen Lynch, A. Pierce, and A. D. Whetton. 1995. D. Donaldson, S. Lorimore, and I. B. Pragnell. 1990. Identification and charac- Macrophage inflammatory protein-1␣ receptors are present on cells enriched for terization of an inhibitor of haemopoietic stem cell proliferation. Nature 344:442. CD34 expression from patients with chronic myeloid leukemia. Blood 86:4270. 3. Broxmeyer, H. E., B. L. L. Sherry, S. Cooper, O. Oh, P. Tekamp-Olson, 29. Owen-Lynch, P. J., J. A. Adams, M. L. Brereton, A. D. Whetton, and A. L. Yin. B. S. Kwon, and A. Cerami. 1990. Enhancing and suppressing effects of recom- 1996. Blast cells from patients with acute myeloid leukemia are refractory to the binant murine macrophage inflammatory proteins on colony forming formation in inhibitory effects of MIP-1␣. Br. J. Haematol. 95:77. vitro by bone marrow myeloid progenitor cells. Blood 76:1110. 30. Han, Z. C., M. Lu, L. Junmin, M. Defard, B. Boval, N. Schlegel, and J. P. Caen. 1997. 4. Parkinson, E. K., G. J. Graham, P. Daubersies, J. E. Burns, C. Heufler, M. Plumb, Platelet factor 4 and other CXC chemokines support the survival of normal haemo- G. Schuler, and I. B. Pragnell. 1993. Hematopoietic stem cell inhibitor (SCI poietic cells and reduce the chemosensitivity of cells to cytotoxic agents. Blood 89:2328. The Journal of Immunology 6199

31. Wognum, A. W., P. M. Lansdorp, R. K. Humphries, and G. Krystal. 1990. De- 37. Mayani, H., M. T. Little, W. Dragowska, G. Thornbury, and P. M. Lansdorp. tection and isolation of the erythropoietin receptor using biotinylated erythropoi- 1995. Differential effects of the hematopoietic inhibitors MIP-1␣, TGF-␤, and etin. Blood 76:697. TNF-␣ on cytokine-induced proliferation of subpopulations of CD34ϩ cells pu- 32. Wognum, A. W., Y. Westerman, T. P. Visser, and G. Wagemaker. 1994. Dis- rified from cord blood and fetal liver. Exp. Hematol. 23:422. tribution of receptors for granulocyte-macrophage colony-stimulating factor on 38. Jacobsen, S. E., F. W. Ruscetti, M. Ortiz, J. M. Gooya, and J. R. Keller. 1994. The immature CD34ϩ bone marrow cells, differentiating monomyeloid progenitors growth response of Lin-Thy-1ϩ hematopoietic progenitors to cytokines is deter- and mature blood subsets. Blood 84:764. mined by the balance between synergy of multiple stimulators and negative co- 33. Hunter, M. G., L. Bawden, D. Brotherton, S. Craig, S. Cribbes, L. G. Czaplewski, operation of multiple inhibitors. Exp. Hematol. 22:985. T. M. Dexter, A. H. Drummond, A. H. Gearing, C. M. Heyworth, et al. 1995. 39. Loetscher, P., M. Seitz, M. Baggiolini, and B. Moser. 1996. Interleukin-2 regu- BB-10010: an active variant of human macrophage inflammatory protein-1␣ with lates CC chemokine receptor expression and chemotactic responsiveness in T improved pharmaceutical properties. Blood 86:4400. lymphocytes. J. Exp. Med. 184:569. 34. Avalos, B. R., K. J. Bartynski, P. J. Elder, M. S. Kotur, W. G. Burton, and 40. Su, S., N. Mukaida, J. B. Wang, H. Nomura, and K. Matsushima. 1996. Preparation N. M. Wilkie. 1994. The active monomeric form of macrophage inflammatory of specific polyclonal antibodies to a C-C chemokine receptor CCR-1 and determi- protein-1␣ interacts with high- and low-affinity classes of receptors on human nation of CCR1 expression on various types of leukocytes. J. Leukocyte Biol. 60:658. hematopoietic cells. Blood 84:1790. 41. Su, S., N. Mukaida, J. Wang, Y. Zhang, A. Takami, S. Nakao, and 35. Baum, C. M., I. L. Weissman, A. S. Tsukamoto, A.-M. Buckle, and B. Peault. K. Matsushima. 1997. Inhibition of immature progenitor cell proliferation by 1992. Isolation of a candidate human haematopoietic stem cell population. Proc. macrophage inflammatory protein-1␣ by interacting mainly with a C-C chemo- Natl. Acad. Sci. USA 89:2804. kine receptor, CCR-1. Blood 90:605. 36. Mayani, H., and P. Lansdorp. 1994. Thy-1 expression is linked to functional 42. Wu, L., N. Ruffing, X. Shi, W. Newman, D. Soler, C. Mackay, and S. Qiu. 1996. properties of primitive haematopoietic progenitor cells from human umbilical Discrete steps in binding and signalling of interleukin 8 with its receptor. J. Biol. cord blood. Blood 83:2410. Chem. 271:31202. Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021