|Home | About | Journals | Submit | Contact Us | Français|
Our understanding of the mechanisms by which intravenously transplanted hematopoietic stem/progenitor cells (HSPCs) home to and engraft the bone marrow (BM) remains incomplete, but participation of adhesion molecules has been documented. We here demonstrate that blockade of the α6-integrin enhanced BM homing of human and nonhuman primate BM-derived HSPCs by >60% in the xenogeneic transplant model and led to significantly improved engraftment. The effect was limited to BM-derived HSPCs, as granulocyte-colony-stimulating factor mobilized peripheral blood or cord blood HSPCs express little or no α6 integrin. By contrast, despite high α6 integrin expression, no effect of α6 blockade on murine BM-HSPCs homing/engraftment was observed.
Homing of intravenously (i.v.) transplanted hematopoietic stem/progenitor cells (HSPCs) to bone marrow (BM) involves a series of well-coordinated interactions between HSPC adhesion receptors and stromal cell/extracellular matrix ligands [1, 2]. Expression of a number of adhesion receptors on HSPCs has been reported , and functional roles for several α-partners of the β1-integrins have been documented. Thus blockade of α4β1 integrin negatively influences homing of human and murine HSPCs [4, 5]. Divergent effects between human and murine HSPCs emanate from studies into the role of α5β1 integrin in HSPC homing/engraftment [5–8], and no effect on homing was associated with inhibition of α2β1 integrin in the mouse, presumably because α2-integrin is not expressed on the most primitive HSPCs .
Little is known about the functional role of the α6-integrin on HSPCs, even though its expression by most, if not all, BM–HSPCs has been unequivocally reported. α6-Integrin can heterodimerize with β1 and β4. Both dimers are laminin receptors. α6β4 binds laminin-5 and is thus required for epithelial integrity. Since hematopoietic cells do not appear to express β4-integrin , α6β1 (VLA6) is likely the sole hematopoietic α6-integrin heterodimer. On mature hematopoietic cells, α6 can function as an adhesion molecule and as a receptor for laminin-induced outside-in signals mediating haptotaxis. Outside hematopoiesis, complex roles for α6-integrin have been described, including roles in reproduction, epithelial integrity, and organogenesis, all of which also appear to hinge on interactions with various laminin isoforms [11–13]. Mediated through α6β1, HSPCs bind to several laminin isoforms, including, but likely not limited to, laminins 1, 8, and 10/11 [14, 15]. However, observations about the functional role of α6β1-integrin in primitive hematopoiesis are thus far incomplete [16, 17], and were therefore pursued here.
In this study, differential expression levels of α6-integrin on human BM- and granulocyte-colony-stimulating factor mobilized peripheral blood- (MPB-) HSPCs were documented, which are likely responsible for the differential functional consequences of α6-integrin blockade for homing and engraftment. In contrast to human cells, similar studies with murine HSPCs did not demonstrate α6-integrin dependent roles for homing/engraftment.
B6x129, nonobese diabetic/severe combined immunodeficiency (NOD/SCID)β2-microglobulin−/− or NOD/SCIDγc−/− (Jackson Laboratories, Bar Harbor, ME) were housed at the specific pathogen-free facilities at University of Washington Medical Center or Fred Hutchinson Cancer Research Center. All procedures were approved by the institutional animal care and use committee.
Cryopreserved human CD34+ BM- or MPB cells were gifts from Shelley Heimfeld (Fred Hutchinson Cancer Research Center, Seattle, WA) and JoAnna Reems (Puget Sound Blood Center, Seattle, WA). CD34+ cord blood (CB) cells were a gift from Chris Miller (University of Washington, Seattle, WA). Baboon CD34+ BM cells were isolated as previously described . Murine BM cells were flushed from long bones; where indicated, c-kit+ cells were purified with immunomagnetic beads according to manufacturer's instructions (Miltenyi Biotech, Auburn, CA). Cells were incubated with the α6-function-blocking antibody (Ab) GoH3 (BD-Pharmingen, San Diego, CA) or isotype-matched control Ab (BD) (20 μg/mL, 4 C, 30 min) .
For homing experiments, hosts were irradiated with a single dose of 1,250 cGy total-body irradiation (TBI), followed by i.v. injection of HSPC suspensions as indicated. Progenitor cell (colony-forming cell culture, CFU-C) homing to BM as well as, where indicated, CFU-C content in blood, spleen, and lung were assessed by plating aliquots of cell suspensions from these organs, as described , 3 or 20 h after transplantation, using commercially available semi-solid complete colony assay media (Miltenyi Biotech for human CFU-C, StemCell Technologies, Vancouver, BC, Canada for murine CFU-C) . For engraftment studies, recipients of human cells received 350 cGy TBI, recipients of mouse cells 1,250 cGy TBI. Engraftment was analyzed by quantification of donor-derived CD45+ cells in blood  or donor CFU-C contents in femurs . For human engraftment, a threshold of 1% was considered indicative of engraftment. Transplanted cell doses per recipient were 0.4–0.6×106 for human or baboon CD34+ cells, 3–5×106 for murine c-kit+ cells, or 20–30×106 for total murine BM cells.
Adhesion to tissue sections was assessed as described . Frozen tissue sections of liver (fresh) and lung (prefixed) from lethally irradiated mice were prepared. A 1:1 mix of control and anti-α6-Ab-treated CD34+ cells, loaded with red (SNARF) or green (CFSE) cell dyes (Molecular Probes, Eugene, OR) according to manufacturer's instructions, was allowed to adhere to the sections (1 h, room temperature). After washing by 10 times dipping into phosphate-buffered saline at room temperature, to remove nonadherent cells, slides were dried and mounted with Vectashield+DAPI. Adhering cells were counted by fluorescence microscopy in all sections (Microscope: Leica DMIRE2, Wetzlar, Germany, with triple bandpass filter (420, 520, and 610 nm); camera: Spot RT Slider, Diagnostic Instruments, Sterling Heights, MI; Acquisition software: Openlab 3.1.7, Improvision, Lexington, MA). Red, green, and blue pseudocolored images were merged after adjustment of contrast and brightness.
Flow cytometry, using directly conjugated Abs (BD-Pharmingen), was performed according to standard protocols using the FACSCalibur (BD Immunocytometry Systems, San Jose, CA).
Homing efficiency was compared using the t-test. Probability of engraftment among groups was compared using the χ2 test. Descriptive statistics, t-tests and χ2 statistics were calculated using Excel (Microsoft, Redmond, WA).
α6-Integrin is expressed on virtually all CD34+ cells from human and baboon BM (Fig. 1A and C). By comparison, expression of α6-integrin was much lower on human MPB-CD34+ (Fig. 1B), and virtually absent from human cord blood (CB) CD34+ cells (Fig. 1B). Thus α6-integrin joins the list of adhesion molecules for which reduced expression on MPB- and CB-CD34+ cells was shown [23, 24].
HSPCs were incubated in vitro with saturating concentrations of α6-integrin-blocking Ab (anti-α6-Ab) (Fig. 1C) or with similar concentrations of isotype-matched control Abs prior to transplantation. This treatment had no effect on colony formation in vitro (Fig. 1D). Homing of anti-α6-Ab-treated CFU-C from human BM, assessed 20 h after injection into lethally irradiated NOD/SCIDβ2-microglobulin−/− mice, was increased by more than two-thirds compared to control-treated HSPCs (mean±SEM from three independent experiments: control Ab 1.2±0.13% vs. anti-α6-Ab 2.2±0.17% of injected human CFU-C/2 femurs, Fig. 1E). Similar results were observed when baboon BM-HSPCs were used as transplants (Fig. 1G). At the same time, the number of HSPCs recovered from spleens was significantly reduced in recipients of anti-α6-Ab treated HSPCs (Fig. 1E and G). Recovery from lungs was highly variable, but also trended toward lower values (Fig. 1E). Since enumeration of CFU-C homed to liver and lung could not be successfully achieved with our in vivo homing assays, a surrogate in vitro adhesion assay to sections from irradiated lung and liver was performed and demonstrated decreased binding of anti-α6-Ab-treated CD34+ cells compared to controls (Supplementary Fig. 1; Supplementary materials are available online at http://www.liebertpub.com). In keeping with their superior homing, 8-week engraftment of human cells in sublethally conditioned NOD/SCIDγc−/− mice was significantly increased after incubation with anti-α6-Ab (Fig. 1H). In contrast to BM-HSPCs, BM homing of human MPB-HSPC was not significantly affected by anti-α6-Ab incubation (Fig. 1F). In agreement with these data, short-term engraftment of these cells was also no different between anti-α6-Ab and control-treated HSPCs (not shown).
Similarly to human and baboon BM-CD34+ cells, virtually all murine c-kit+ cells express α6, at a similar density as primate HSPCs (Fig. 2A). Nevertheless, homing of anti-α6-Ab-treated (saturating conditions, Fig. 2B) murine HSPCs in isogeneic (Fig. 2C and D) or in NOD/SCIDβ2-microglobulin−/− (Fig. 2E) hosts, tested 3 or 20 h after transplantation, was no different from control-treated HSPCs. This outcome was observed whether whole BM or c-kit-enriched BM-HSPCs were used, in normal as in splenectomized recipients (Fig. 2D). Likewise, short-term engraftment was not affected by anti-α6-Ab incubation (Fig. 2F).
Many earlier studies have addressed the effects of antifunctional Abs on homing and engraftment of HSPCs [5,6,9, 25]. Although the interaction between HSPCs and Ab in vivo is likely transient, the interaction is sufficiently long to affect homing, because the bulk of HSPCs homing to BM do so within 3 h of transplantation, and little change occurs thereafter . Thus negative effects of antifunctional Abs on homing were regularly reflected in short-term engraftment [7,21,25,27, 28]. The observation that α6-blockade can enhance homing (and, consequently, engraftment) of human and baboon BM–derived HSPCs is notable in that it is the first modification of an integrin to have a positive effect. Although our homing data for human/baboon HSPCs were generated using xenogeneic recipients, it should be stressed that all previous studies of human HSPCs homing were also performed in this model. Our data also demonstrate functional consequences of the differential expression of α6-integrin on BM-and MPB-HSPCs, by showing that homing of MPB-HSPCs, which express little α6, is not affected by α6-blockade. Based on these data we can speculate that the previously reported superior BM homing of human MPB-HSPCs compared with BM-HSPC may among other parameters be a reflection of differential α6-integrin expression, although the differences between the cell types are complex, and additional factors are clearly involved .
Our data further describe a novel difference in the molecular mechanisms of BM homing between human and mouse HSPCs. Similar effects concern α5-integrin and CXCR4, which at least in some reports were considered essential for human, but redundant for murine HSPC homing [5–7,21, 30]. In this respect, earlier reports on negative effects of α6-blockade on murine BM-HSPCs homing without an effect on short-term engraftment  are inherently contradictory. Furthermore, they are difficult to reconcile with the absence of effects of α6-deletion on fetal liver HSPC homing  or with the data presented in this article, which were generated using very similar techniques, including use of the same Ab clone and concentration. Earlier studies have also demonstrated the absence of an effect of anti-α6-Ab on human CB-HSPC homing , an observation that is consistent with our data reporting the absence of α6 on CB-CD34+ cells reported here.
With respect to mechanisms involved in the observed enhancement of homing in the absence of α6-integrin, we can only speculate at present. Laminin 1, a major constituent of the perivascular extracellular matrix (ECM) and of basal membranes, is highly abundant and ubiquitously expressed. It would likely be the first laminin encountered by HSPCs, even more so after lethal irradiation, when endothelial surfaces have been denuded. We thus speculate that retention of HSPC on laminin 1, but also on other ECM constituents, such as fibronectin and other laminins, is a relevant “trap” for transplanted HSPCs, retention of which in nonhematopoietic organs is well documented. If such trapping is reduced (such as by integrin blockade), the number of HSPCs circulating in blood would be increased, similar to what has been described for blocking of liver uptake of HSPCs by asialoglycoprotein receptor inhibition . Thus the number of HSPCs available for homing through canonical homing receptors (α4, LFA-1, CXCL12/CXCR4, etc.) would be increased. Our finding of reduced HSPC recovery from spleens is consistent with this interpretation. Given the small total number of HSPCs retained in spleen, however, the increased marrow homing of anti-α6-treated HSPCs cannot be attributed to reduced retention in spleen alone. Instead, we propose that the modestly reduced retention of anti-α6-Ab-treated HSPCs in all extramedullary organs, also demonstrated by us, is in aggregate responsible for the observed phenotype.
The theoretical possibility that outside-in signals through α6 are elicited by the antifunctional Ab cannot be ruled out. For instance, laminin-induced haptotaxis, mediated through α6-integrin, has been reported . Such effects could contribute to the observed phenotype.
In summary, α6-integrin blockade improves homing and engraftment of human HSPCs in mice, although this effect is limited to BM-derived cells, and if verified in a large animal model, it presents a strategy to enhance engraftment. Our studies furthermore describe different expression levels of α6-integrin between human CD34+ HSPCs from BM, MPB, and CB, and identify a novel differential homing mechanism between human and murine HSPCs.
The study was supported by a CCEMH pilot grant from the Fred Hutchinson Cancer Research Center (FHCRC) (to H.B.), by National Institutes of Health (NIH) grant HL58734 (T.P.), and by NIH grant P01 HL53750 (to H.P.K.). H.B. conceived of the studies, planned and performed experiments, analyzed data, and wrote the article. G.V.P. and M.W. performed experiments. H.P.K. provided vital reagents and advised about experiments. T.P. planned experiments and wrote the article.