Sustained alterations in biodistribution of stem/progenitor cells in Tie2Cre+α4^f/f mice are hematopoietic cell autonomous

The use of Cre-lox strategy to mediate conditional deletion of embryonic lethal genes in a tissue-restricted manner has enabled the investigation of their functional roles in adult life. A commonly used approach to delete chosen floxed genes in hematopoietic cells is the interbreeding with MxCre transgenic mice harboring an interferon-inducible transgene.^1-5  Although efficient deletion in hematopoietic cells has been documented using this strategy, the extent of deletion in nonhematopoietic cells has not been rigorously assessed previously. The α4 integrin is expressed widely in hematopoietic and nonhematopoietic cells, such as endothelial cells⁶  or pericytes⁷  within the bone marrow (BM). In fact, its presence in endothelial cells is critical in tumor progression, in neoangiogenesis, and in developing blood vessels.^6,7  Conditional deletion of α4 integrin in adults has led to sustained changes in biodistribution of stem/progenitor cells at steady-state hematopoiesis.⁵  A similar, albeit transient, picture has been described with the use of antifunctional α4-integrin antibodies in vivo.⁸  Using either strategy, it has not been clear to what extent the phenotype is solely attributed to deletion of α4 in hematopoietic cells or contribution by similarly affected cells in BM microenvironment, as has been recently advocated.9-11 To address this issue the following approaches were used. First, we generated mice in which the α4 integrin is ablated both in hematopoietic and in endothelial cells, by breeding α4 floxed mice (f/f) with Tie2Cre⁺ transgenic mice with the expected α4 ablation in hematopoietic and endothelial cells.¹²  Second, we carried out transplantation experiments using either donor cells from these mice or using them as recipients of normal cells. The results from these experiments provide definitive conclusions about the contribution of α4-integrin ablation only in hematopoietic cells to the phenotype seen at homeostasis.

(C57Bl/6 × 129) mice homozygous for the floxed α4 allele (α4^f/f) were crossed to C57Bl/6 mice expressing Cre under the control of the Tie2 promoter to generate Tie2Cre⁺α4^f/f mice. Genotyping was performed by polymerase chain reaction (PCR) using the following primers: F-1: 5′-CGGGATCAGAAAGAATCCA-3′; F-2: 5′-CCACCTGGTGTATGAAAGC-3′; Rev: 5′-GATCACATACAGTTGCAAGC-3′ for discrimination of wild-type (WT), f, and Δ alleles, respectively. Tie2Cre+/VCAM-1^f/f mice were previously described.¹³ 

Directly conjugated antibodies from BD Biosciences (San Diego, CA) were B220, CD3, CD4, CD8, CD45, TER119, and GR1. Directly conjugated CD49d antibody was from Southern Biotechnology (Birmingham, AL). BM or blood cells stained with these antibodies were analyzed on a FACSCalibur (BD Biosciences) using CellQuest software (BD Biosciences).

Colony-forming unit in culture (CFU-C) assays were performed from BM, spleen, or peripheral blood (PB) samples using a methylcellulose mixture (Methocult GF; StemCell Technologies, Vancouver, BC, Canada), as described previously.⁵ 

BM cells (0.5, 2.5, or 5.0 × 10⁶/recipient) from α4-deficient or from normal mice were infused into lethally irradiated normal or α4-deficient recipients (1150 cGy whole body irradiation with a ¹³⁷Cs source). Recipients were analyzed at multiple points after transplantation. Donor reconstitution was verified by fluorescence-activated cell sorting (FACS) or PCR or both.

Endothelial cells were procured from cleansed, flushed, and minced long and pelvic bones and cultured as described.¹³  For generating BM-derived fibroblasts the cells were cultured in DMEM with 20% FCS for 4 to 7 weeks. Contaminating macrophages were removed by depletion of CD45⁺ cells, whereas endothelial cells were identified by several endothelial-specific markers.¹³  Reverse transcription-PCR (RT-PCR) was performed as described.¹³  Primers for α4: forward, 5′-TGTGGAAGGCTGGATTCTTT-3′, reverse, 5′-CGGGTCTTCTGAACAGGATT-3′; VWF: forward, 5′-AGAGGAGAGTACATCTGGGAG-3′, reverse, 5′-AGAACACCTGCACTGCATGGC-3′.

All statistical analyses were performed with a 2-tailed Student t test.

Tie2Cre⁺α4^f/f mice are viable and healthy with no apparent ill effects when observed for more than a year. Phenotypically all circulating PB or BM cells were virtually α4⁻ (Figure 1A-B), as were BM endothelial cells, whereas α4 deletion in bone fibroblasts/mesenchymal cells was partial at best (Figure 1B). Detailed studies on hematopoiesis disclosed that white blood cell (WBC) counts were above control levels (Figure 1C) with no deviation in lineage-specific contributions (data not shown). Circulating progenitors were several-fold higher than controls (Figure 1C) and this difference was maintained through 16 months of postnatal life (Figure S1A, available on the Blood website; see the Supplemental Figure link at the top of the online article). Overall the hematopoietic phenotype was similar to that previously described for Tie2Cre⁺VCAM-1^f/f mice.¹³  However, circulating progenitors in Tie2Cre⁺α4^f/f were higher (Figure 1C). The fact that α4 integrin has additional ligands other than VCAM-1, (ie, fibronectin, osteopontin, VWF, ADAMS28, α4-chain itself) could explain the more efficient release of α4^Δ/Δ progenitors. Alternatively, one may suggest that efficient excision of α4 integrin from endothelial cells or other microenvironmental cells in this model may have a contributing role.¹¹ 

To test these possibilities, transplantation experiments were carried out. We transplanted α4-deficient BM cells (Figure 2A) into lethally irradiated Cre(−) littermates (C57B1/6 × 129) and assessed the levels of circulating progenitors after complete hematopoietic reconstitution. The α4^+/+ cells given to α4^+/+ or to α4^Δ/Δ lethally irradiated recipients were also followed similarly (Figure 2B). Although α4^+/+ cells given to α4^+/+ or α4^Δ/Δ recipients did not show increased levels of circulating progenitors after transplantation, α4-deficient cells transplanted to normal recipients had significantly increased numbers of circulating progenitors up to 29 weeks after transplantation, compared to their concurrent controls. Thus, α4^Δ/Δ hematopoietic cells transplanted to a normal BM environment are necessary and sufficient to provide a picture similar to one seen in animals that did not receive a transplant (Figure S1B). Alternatively, α4 deletion in endothelial cells or pericytes⁷  in BM does not seem to contribute in any measurable way to the altered biodistribution of hematopoietic progenitor cells at steady state.

The data presented here and the information we recently published¹⁴  display several common features to those described¹⁵  with transplantation of CXCR4^−/− fetal cells: (1) high numbers of circulating progenitor cells starting at 2 weeks after transplantation and continuing for at least 16 weeks; (2) rescue of engraftment defects when higher numbers of donor cells are used; and (3) the presence of a significant number of long-term repopulating cells in circulation, as documented by transplantation experiments using blood.^5,15  The enhanced release of progenitors from BM to blood seen in both cases is cell autonomous and would suggest that the 2 pathways, α4/VCAM-1 or CXCR4/CXCL12 (SDF-1), have a nonredundant influence on the retention of progenitors in BM at steady-state hematopoiesis.

Although alterations in progenitor biodistribution were seen previously in other genetically deficient models, that is, with Rac2 GTPase^−/−4  or CaR^−/− cells,¹⁶  a picture recapitulating after transplantation the phenotype seen in animals that did not receive a transplant has not been documented in these cases. A dramatic increase was seen after transplantation with Rac2^−/− Rac1^Δ/Δ doubly deficient donor cells, but this was transient, lasting 4 weeks after transplantation,¹⁷  unlike our results. Surprisingly, transplantation of MxCre⁺β1^f/f cells and their ablation after transplantation, theoretically affecting several β1-integrin heterodimers, did not lead to any alterations in stem/progenitor biodistribution,²  inviting the speculation that β1 integrins expressed in nonhematopoietic cells or niche cells are more important.^9,10  Although β1/β7-deficient animals display a phenotype (in BM and PB) similar to our Tie2Cre⁺α4^f/f mice, this phenotype, unlike the one in our mice, reverts to normal with aging as a result of putative compensatory changes.¹¹  Whatever the case, our results clearly show that active engagement of α4 integrin in hematopoietic cells plays an important role in physiologic progenitor retention within BM.

This work was supported by National Institutes of Health grant HL046557.