Pivotal Role of Lnk Adaptor Protein in Endothelial Progenitor Cell Biology for Vascular Regeneration
Despite the fact that endothelial progenitor cells (EPCs) are important for postnatal neovascularization, their origins, differentiation, and modulators are not clear. Here, we demonstrate that Lnk, a negative regulator of hematopoietic stem cell proliferation, controls endothelial commitment of c-kit+/Sca-1+/Lineage− (KSL) subpopulations of bone marrow cells. The results of EPC colony–forming assays reveal that small (primitive) EPC colony formation by CD34− KSLs and large (definitive) EPC colony formation by CD34(dim) KSLs are more robust in lnk−/− mice. In hindlimb ischemia, perfusion recovery is augmented in lnk−/− mice through enhanced proliferation and mobilization of EPCs via c-Kit/stem cell factor. We found that Lnk-deficient EPCs are more potent actors than resident cells in hindlimb perfusion recovery and ischemic neovascularization, mainly via the activity of bone marrow-EPCs. Similarly, lnk−/− mice show augmented retinal neovascularization and astrocyte network maturation without an increase in indicators of pathogenic angiogenesis in an in vivo model of retinopathy. Taken together, our results provide strong evidence that Lnk regulates bone marrow-EPC kinetics in vascular regeneration. Selective targeting of Lnk may be a safe and effective strategy to augment therapeutic neovascularization by EPC transplantation.
Stem cell–related, postnatal neovascularization requires several activities of putative stem cells and their progeny, endothelial progenitor cells (EPCs), including the ability to self-renew in bone marrow (BM), commitment and differentiation into mature endothelial cells (ECs), mobilization from BM into the circulatory system, and recruitment to sites of neovascularization.1,2 Many cytokines augment mobilization and/or recruitment of BM-derived EPCs,3,4 including granulocyte colony-stimulating factor and granulocyte/macrophage colony-stimulating factor; angiogenic growth factors such as vascular endothelial growth factor (VEGF) and stromal cell–derived factor (SDF)-1; estrogen; and pharmaceutical drugs such as statins. However, these factors act not only on immature stem/progenitor cells but also on hematopoietic cells and mature ECs. Thus, the identification of a novel molecule that specifically regulates immature populations involved in EPC kinetics in BM is warranted.
Differentiation of progenitor cells into hematopoietic and endothelial lineage cells has been intensively investigated. During development, hemangioblastic aggregates originate from the mesodermal yolk sac, migrate to the fetal liver, and finally establish themselves in the BM. The results of a number of gene-targeting studies contribute to our understanding of functional molecules such as Scl/Tal,5 c-kit, CD34, Runx-1,6 and Flk-1,7 which regulate the developmental kinetics of hemangioblasts and are also expressed in the common precursors of hematopoietic cells and ECs. Postnatal hematopoietic stem cells (HSCs) and EPCs also share common markers; however, the precise characteristics of hemangioblasts, mechanisms regulating cell growth in adults, and endothelial commitment of putative stem cells and/or common precursors for hematopoietic cells and ECs for postnatal vasculogenesis have not previously been reported.
The Lnk protein shares a pleckstrin homology domain, a Src homology 2 domain, and potential tyrosine phosphorylation sites with APS and SH-2B. It belongs to a family of adaptor proteins implicated in integration and regulation of multiple signaling events.8 Lnk has been studied in the immune system, where Lnk regulates B cell production via negative regulation of pro–B-cell expansion.8 Recently, Lnk was reported to play a critical role in maintaining the ability of HSCs to self-renew, in a study that based on BM c-Kit+/Sca-1+/Lineage (Lin)− (KSL) CD34− cells, which are a more immature HSC subpopulation than KSL CD34+ cells.9 Importantly, earlier studies report that expression of lnk is strong in immature cells, ie, c-kit+/Lin− cells, as compared with relatively mature cells, ie, c-Kit−/Lin− cells.10 Accordingly, and because mouse BM-KSLs are capable of differentiating into both hematopoietic and endothelial lineage cells and contribute to postnatal vasculogenesis,11–13 Lnk may regulate the functional kinetics of EPCs. Lnk has also been suggested to act as a negative regulator of the stem cell factor (SCF)–c-Kit signaling pathway.10 SCF reportedly stimulates proliferation and differentiation of HSCs and mobilizes stem cell populations from BM into peripheral blood (PB) by binding with its receptor, c-Kit. The SCF–c-Kit signaling pathway also supports stem cell survival and motility.14 Moreover, EPCs are recruited via interaction with membrane-bound c-Kit, which is highly expressed on ECs in ischemic tissue.15 The c-Kit–positive cells recruited to ischemic tissue reconstitute the injured heart and vasculature, via to their ability to regulate the myocardial balance of angiogenic cytokines.16
Here, we sought to test the hypothesis that a lack of Lnk signaling may enhance postnatal neovascularization via specific control of the SCF–c-Kit–mediated regenerative potential of EPCs. We provide in vitro and in vivo evidence that Lnk plays a pivotal role in specific modulation of EPCs in terms of cell growth, commitment into endothelial lineage cell types, mobilization from BM into PB, and recruitment to ischemic sites for neovascularization.
Materials and Methods
An expanded Materials and Methods section is available in the online data supplement at http://circres.ahajournals.org.
The lnk−/− mice were generated as previously reported.8 All animal care and experiments were conducted in accordance with the institutional guidelines of Tokai University School of Medicine, Isehara, Japan.
EPC colony–forming assay (EPC-CFA), single cell–based EPC-CFA, mobilization of EPCs, and in vitro 5′-bromodeoxyuridine (BrdUrd) proliferation assay were performed.
In Vivo Study
Hindlimb Ischemia Model and Cell Transplantation
The mouse model of hindlimb ischemia was generated by ligating the proximal femoral artery of 8- to 10-week-old C57BL6/J or Balb/C nude mice.
BM Transplantation Model
C57BL/6J mice were exposed to a lethal dose of total body irradiation (10 Gy) and inoculated intravenously with 1×106 donor BM mononuclear cells (BM-MNCs).
Murine Model of Oxygen-Induced Retinopathy
Oxygen-induced retinopathy (OIR) was induced in C57/BL6 wild-type (WT) and lnk−/− mice.
Deficiency of lnk Augments Endothelial Differentiation and Upregulates Cell Growth-Relating Signals in BM-KSL Subpopulations
Although a previous report has clearly shown that self-renewal of BM-CD34− KSLs for hematopoiesis is accelerated in lnk−/− mice,9 the role of Lnk in ischemic vasculogenesis is unknown. We first examined lnk mRNA levels in various populations of BM cells and several organs of WT mice in the presence or absence of limb ischemia. Expression of lnk mRNA is strong in BM-CD34− KSLs regardless of tissue ischemia. Expression of lnk is moderate in BM-CD34+ KSLs, a relatively differentiated population as compared with CD34− KSLs. In contrast, lnk expression was faint in samples from BM-MNCs, BM-Lin− cells, skeletal muscle, and spleen independently of ischemia. These results suggest that lnk is highly expressed in BM hematopoietic and endothelial progenitors but not in mature BM cells or other organs. The lnk expression levels were especially high in the immature fraction of BM-HSC/EPCs as compared to committing fractions (Figure 1a).
The pattern of expression of lnk suggests a role in differentiation of various subpopulations among BM-KSLs. To test this, we next compared the number of BM-KSLs and derivative subpopulations in lnk−/− and WT mice. The number of KSLs, CD34− KSLs, and CD34(dim) KSLs, but not CD34(high) KSLs, was significantly greater in lnk−/− mice than in WT. These data suggest that deletion of lnk results in an increase in immature subpopulations of KSLs (Figure 1b). To compare the vasculogenic commitment of BM-KSLs in lnk−/− mice versus WT, fluorescence-activated cell-sorting analysis for endothelial markers was performed. KSLs coexpressing Flk-1 or CXCR4 were more frequent in lnk−/− mice than in WT (Figure 1c). Thus, loss of lnk appears to promote vasculogenic commitment, resulting in an increase in the EPC pool in BM. Similarly, the number of EPCs increased in PB of Lnk-deficient mice (Figures I and II in the online data supplement).
To further confirm the role of Lnk in differentiation of KSL subpopulations into endothelial lineage cells, we performed an EPC-CFA established recently in our laboratory. KSLs and their subpopulations can form 2 types of EPC colony clusters, small (primitive) and large (definitive) EPC colony clusters. Both cluster types are positive for uptake of acetyl LDL (Ac-LDL) and for expression of an EC-specific marker, isolectin B4, as revealed by chemical staining. Additionally, both are positive for Flk-1 (VEGF receptor 2) and CD31 (platelet endothelial cell adhesion molecule-1), as revealed by immunocytochemistry (Online Figure III, a through d). Moreover, colony-derived cells express the endothelial markers Flk-1 and CD31 at high levels, as detected by flow cytometric analysis. Cells from large EPC clusters, which comprise more committed EPCs with spindle-like morphology, more frequently show Ac-LDL uptake and higher expression of Flk-1 and CD31 than cells from small EPC clusters (Online Figure III, c and d).
EPC-CFA was performed for each KSL subpopulation obtained from lnk−/− or WT mice. The number of small EPC colonies derived from CD34− KSLs was significantly greater in lnk−/− mice than in WT, whereas the number derived from CD34(dim) or CD34(high) KSLs was similar in lnk−/− and WT. In contrast, the number of large EPC colonies from CD34− KSLs was similar in both groups, whereas the number from CD34(dim) or CD34(high) KSLs was significantly higher in lnk−/− mice than WT (Figure 1d). These data suggest that Lnk deficiency increases the capacity of immature stem cells to form primitive EPCs and in the capacity of relatively mature progenitor cells to differentiate into definitive EPCs.
To compare cell growth of CD34−/CD34(dim) KSLs from lnk−/− versus WT mice, we next analyzed SCF-dependent glycogen synthase kinase (GSK)3β phosphorylation, which is part of a signaling cascade indispensable for cell growth.17 The level of phosphorylation of GSK3β in CD34−/CD34(dim) KSLs was enhanced and prolonged in the lnk−/− background relative to WT. This points to an important role for Lnk in the ability of immature HSC/EPCs to cell growth, as apparently controlled by upregulation of the SCF-dependent GSK3β signaling pathway (Figure 1e).
Lnk Deficiency Upregulates Proliferation and Endothelial Commitment of EPCs Derived From KSL Populations in Culture
To explore the function of Lnk in EPC biology in terms of cell proliferation and commitment, we isolated and cultured Lin− cells, KSLs, and KSL subpopulations from WT and Lnk-deficient mice in a defined EPC culture system. In both lnk−/− and WT genetic backgrounds, KSLs in general, and CD34− KSLs and CD34(dim) KSLs in particular, proliferated efficiently in culture for 1 week, whereas BM-Lin− cells and CD34(high) KSL subpopulation cells exhibited a smaller increase in proliferation. The fold increase in cell number for KSLs, CD34− KSLs, and CD34(dim) KSLs was significantly greater in cells from Lnk-null mice than from WT. In contrast, the fold increase of Lin− cells and CD34(high) KSL subpopulation was similar in cells from lnk−/− or WT mice.
We next looked at cultured KSL subpopulations in lnk−/− and WT genetic backgrounds. The results of flow cytometric analysis reveal that cultured cells derived from CD34− KSL or CD34(dim) KSL subpopulations in lnk−/− mice were more frequently positive for the endothelial lineage markers Flk-1/Sca-1 and CXCR4/Sca-1 than those from WT. However, the number of cells positive for the endothelial markers among cells cultured from the CD34(high) KSL subpopulation was similar for lnk−/− and WT (Online Figure IV, b and c).
To determine whether EPC development from KSL subpopulations occurs at the single-cell level, we sorted single cells from each subpopulation, cultured the cells ex vivo for 1 week, and then assayed the cells using EPC-CFA and flow cytometry. EPC-CFA revealed that the number of large EPC colonies derived from a single CD34− KSL or CD34(dim) KSL, but not a single CD34(high) KSL, was significantly greater when cells were derived from lnk−/− mice. In contrast, the number of small EPC colonies derived from single cells of all subpopulations was similar in the 2 groups (Online Figure IV, d). Flow cytometry also revealed that the frequency of Sca-1+/Flk-1+ cells, an EPC-enriched population, among cultured cells derived from single CD34− KSLs or CD34(dim) KSLs, but not single CD34(high) KSLs, was significantly higher in lnk−/− than WT (Online Figure IV, e).
Lnk Deficiency Promotes Neovascularization In Vivo
The in vitro data above suggest that negative modulation of lnk gene expression may promote neovascularization in ischemic tissue. To test the in vivo effect of lnk deficiency, we generated a hindlimb ischemia model in WT and lnk−/− mice. Laser Doppler perfusion imaging revealed serial recovery of blood flow in the ischemic region of both groups; however, blood flow 14 and 28 days after ischemia was significantly greater in lnk−/− mice than in WT (Figure 2a and 2b).
To assess the mechanism of enhanced blood flow recovery in lnk −/− mice, we first compared the mitotic capacity of EPC-enriched populations in the presence or absence of hindlimb ischemia in lnk −/− and WT genetic backgrounds. The percentage of BM Sca-1+/BrdUrd+ cells in Lin− cells without ischemia tended to be greater in lnk −/− mice than in WT, but the difference was not statistically significant. In contrast, the percentage of cycling EPCs 7 days after ischemia was significantly greater in lnk−/− mice than in WT. These data suggest that the proliferative activity of EPCs in response to ischemia is upregulated in the absence of Lnk activity (Figure 2c). Next, we compared the enhancement of ischemia-induced phosphorylation of Akt and of endothelial nitric oxide synthase (eNOS) levels in BM-Sca-1+/Lin− cells in lnk−/− mice versus WT. The results indicate that Lnk-deficient EPCs are more potent for activation of the Akt/eNOS signaling cascade, an important pathway for EPC survival and differentiation (Figure 2d and 2f).18 To clarify the potential of EPCs in lnk−/− mice for ischemic neovascularization, we used RT-PCR to compare mRNA expression of angiogenic factors and their receptors in KSL subpopulations in the presence or absence of hindlimb ischemia in lnk−/− and WT. In lnk−/− mice, genes that encode angiogenic factors or their receptors, such as vegf, ang-1, tie-1, and tie-2, were highly expressed independently of ischemic condition, whereas ang-2, an antagonist of TIE-2 signaling, was constitutively downregulated. In contrast, most angiogenic genes, which are weakly expressed at baseline, were upregulated postischemia in WT (Online Figure V). These data suggest that Lnk regulates the production of angiogenic factors, which in turn enhances EPC proliferation, differentiation, migration, and mobilization.
As for the kinetics of PB-EPCs, the number of Sca-1+ MNCs, an EPC-enriched fraction, on days 3 and 7 after hindlimb ischemia was significantly increased in lnk−/− mice as compared to WT (Figure 2e). Furthermore, the number of Sca-1+/CD31+ and Sca-1+/Flk-1+ cells in PB was greater in lnk−/− mice as compared to WT (data not shown). These outcomes suggest that the mobilization of EPCs into circulation in response to ischemia is augmented in lnk−/− mice compared with WT.
A caveat to the above is that enhanced neovascularization in lnk−/− mice could be attributable to upregulation of angiogenic effects of resident cells as well as augmentation of BM-derived EPC kinetics. To clarify the proportional contribution of these mechanisms, we performed BM transplantation (BMT) with cells from lnk−/− or WT, with donor cells marked with green fluorescent protein (GFP) transplanted into unmarked recipients. Perfusion in limb tissue at day 14 postischemia dramatically improved in WT mice that received Lnk-deficient BM cells (BMCs) as compared with WT mice receiving WT BMCs. Moreover, perfusion recovery in the hindlimb was significantly inhibited in lnk−/− mice receiving WT BMCs as compared with lnk−/− mice receiving Lnk-deficient BMCs. Importantly, hindlimb perfusion was similar in WT and lnk−/− mice receiving WT BMCs. Similarly, perfusion recovery was not significantly different between WT and lnk−/− mice receiving Lnk-deficient BMCs (Figure 3a).
Next, we detected BM-derived endothelial lineage cells incorporating into the ischemic region via immunohistochemical detection of GFP and CD31 or isolectin B4 (Figure 3b). The number of GFP+/CD31+ ECs in ischemic tissue was significantly greater in WT mice receiving Lnk-null BMCs than in those receiving WT BMCs. In addition, BM-derived ECs were more frequently observed in lnk−/− mice receiving Lnk-deficient BMCs than in those receiving WT BMCs. Similar to what was observed in the hindlimb perfusion analysis, the number of BM-derived ECs was equivalent in WT mice receiving Lnk-deficient BMCs and lnk−/− mice receiving Lnk-deficient BMCs, as well as in WT mice receiving WT BMCs and lnk−/− mice receiving WT BMCs (Figure 3c). These results suggest that BM-derived EPCs are indispensable for enhanced neovascularization in lnk−/− mice, whereas lnk deficiency in resident cells does not significantly contribute to ischemic neovascularization.
Lnk Deficiency Enhances EPC Kinetics in Response to Ischemia-Related Cytokines
To identify specific cytokines responsible for enhanced mobilization of BM-EPCs in lnk−/− mice, we investigated the effect of several potent bioactive factors on EPC mobilization in lnk−/− and WT genetic backgrounds. To do this, we administered G-CSF, SDF-1α, SCF, VEGF, or PBS to mice once daily over 5 days and determined the number of PB-MNCs on day 7. In both lnk−/− and WT mice, each factors resulted in a significant increase in the number of PB-MNCs as compared with mock treatment (PBS). The number of PB-MNCs after administration of each factor was significantly greater in lnk−/− mice than in WT. Notably, SCF and VEGF led to a more than 4-fold difference in PB-MNC number in lnk−/− versus WT (Figure 4a). The results of an EPC culture assay using PB-MNCs also revealed that the number of circulating EPCs detected after infusion of any of the factors tested significantly increased in lnk−/− mice as compared with WT. This difference between the 2 groups was particularly remarkable following infusion of SCF or VEGF (Figure 4b). To evaluate the scale of the Lnk-dependent SCF effect on EPC mobilization, we next looked at cell kinetics over time after SCF infusion in lnk−/− or WT mice. The results of serial quantification of PB-MNCs revealed a significant increase in PB-MNCs in lnk−/− mice that was detectable at day 2 and reached a peak on day 6 (Figure 4c). Furthermore, the results of serial FACS analysis revealed a significant increase in the PB-EPC–enriched cell fraction (ie, in Sca-1+/CD31+ or Sca-1+/VE-cadherin+ cells) that was detectable at day 0 and still observable at day 8 after initiation of SCF infusion in lnk−/− mice, as compared with levels in WT (Figure 4d and 4e). We next performed an in vitro proliferation assay to ask whether SCF upregulates proliferative activity of EPCs in lnk−/− mice as well as mobilization. In WT mice, SCF did not affect the mitotic activity of Sca-1+/Lin− cells. In contrast, treatment with 10 ng/mL of SCF significantly augments proliferation of EPC-enriched fraction cells in lnk−/− mice (Figure 4f). Taken together, these data suggest that ischemia-related cytokines, in particular SCF/c-kit, are critical for both proliferation and mobilization of EPCs in lnk−/− mice.
Lnk-Deficient EPCs Rescue Hindlimb Ischemia Following Therapeutic Administration
To evaluate the therapeutic potential of lnk gene–modified EPCs in ischemic neovascularization, we isolated and intravenously transplanted BM Sca-1+/Lin− cells from lnk−/− or WT mice into nude mice with hindlimb ischemia. As shown in Figure 5a and 5b, transplantation of Lnk-null EPCs resulted in robust hindlimb perfusion as compared with WT-EPCs at equal dosing. The results of immunohistochemical analysis using the EC markers isolectin B4 and CD31 surface antigen clearly show that the capillary density at ischemic tissues is higher in animals receiving Lnk-deficient EPCs than in those receiving WT-EPCs or a mock treatment (PBS) control (Figure 5c and 5d and Online Figure VI).
Lnk Deficiency Enhances Neonatal Revascularization in OIR
We next sought to test the effect of Lnk on vascular regeneration in retinal vascular disease. To do this, we generated an animal model of neonatal retinopathy, OIR, by exposing lnk−/− or WT mice to 75% oxygen from postnatal day (P)7 to P12 (Figure 6a). In WT mice with OIR, avascular regions of the retina were readily apparent at neonatal P17. In contrast, lnk−/− mice with OIR had 4-fold smaller retinal avascular areas than WT (Figure 6b and 6c). We also observed functional regeneration of the astrocyte network, accompanied by upregulation of blood vessel regeneration, in lnk−/− mice (Online Figure VII), suggesting that enhanced neovascularization may contribute to preservation of retinal interstitial structure in the Lnk-deficient microenvironment.
Previous results suggest that enhanced angiogenesis/vasculogenesis in the retina may result in pathogenic side effects such as excess inflammation and abnormal blood vessel formation, eventually leading to retinal bleeding.19 However, histological examination of our treated OIR model tissue revealed a smaller number of abnormally sprouting vessels in lnk−/− mice than in WT (Figure 6d through 6f). Moreover, the incidence of retinal hemorrhage at P17 was markedly lower in lnk−/− mice than in WT (Online Figures VIII and IX). These results suggest that lnk deficiency leads to an accelerated rate of retinal neovascularization without stimulating pathogenic blood vessel formation. To investigate this further, we isolated tissue from lnk deficient mice with OIR and used laser microdissection to look at the production of angiogenic growth factors in situ. Levels of VEGF, angiopoietin-1, eNOS, and leukemia inhibitory factor in vascular plexuses were significantly higher in lnk−/− mice than in WT (Figure 7b). Importantly, enhanced expression of ang-1 may inhibit pathogenic angiogenesis by inducing the maturation of newly formed blood vessels.20 The source of angiogenic cytokines in Lnk-null OIR is likely to be at least in part BM-derived EPCs that are recruited into the retina, as both EPCs cultured in vitro under hypoxic conditions (Online Figure X).
The results of previous studies12,21,22 have clearly demonstrated that BM-derived hematopoietic stem cells such as BM-KSLs serve as a reservoir of EPC origin cells in adults. In addition to having a long-term capacity for multilineage hematopoiesis, transplanted KSLs have also been shown to give rise to functional endothelial cells, even after single-cell transplantation or serial transplantation in the presence of retinal ischemic injury.12,21 Although differentiation of hematopoietic and endothelial lineages has been intensively investigated,5–8 molecular targets that regulate endothelial commitment of putative stem cells for postnatal vasculogenesis remain to be uncovered. Identification of molecules that control the commitment and differentiation of adult multipotent stem cells into specific lineages would be a big step toward improved therapeutic treatment in regenerative medicine. Toward identifying a modulator of endothelial development, Guthrie et al have shown that the NO pathway induces new blood vessel formation via EPCs derived from the transplanted KSLs.22 Recently, our group reported that Jagged-1–dependent Notch signaling affects EPC bioactivities including proliferation, endothelial commitment, and mobilization from BM.23 However, these molecules regulate multiple functions in various types of mature and immature cells. In the present study, we found a pivotal function of Lnk adaptor protein as a downstream target of the SCF–c-Kit axis that modulates vasculogenesis in BM stem cells. Lnk was most robustly expressed in BM-CD34− KSLs, which are immature putative stem cells. Lnk is also expressed at moderate levels in CD34+ KSLs, a relatively differentiated stem cell type, but is not expressed in more mature cells such as BM-Lin− cells and MNCs. The results of in vitro analysis using EPC-CFA clearly indicate that Lnk deficiency results in upregulation of commitment of stem cell subpopulations into endothelial lineage cell types. Indeed, lnk deficiency enhances commitment of CD34− KSLs into primitive EPCs (ie, small EPC colonies). Interestingly, lnk deficiency also augments the activity of CD34(dim)/(high) KSLs for the formation of definitive EPCs (large EPC colonies). These findings suggest that Lnk may regulate not only lineage commitment but also differentiation and maturation of EPCs. The specific expression of Lnk in stem cells and the capacity of Lnk to control lineage commitment/differentiation of BM stem cells suggest a pivotal role for Lnk as a regulator of EPCs in adults.
In addition to suggesting roles for Lnk in EPC commitment and differentiation, the results presented here also indicate that Lnk deficiency results in higher levels of proliferation of BM-KSLs and their subpopulations in vitro. Thus, we used an animal model of hindlimb ischemia to assess the effects of Lnk deficiency on EPC kinetics in vivo. Lnk deficiency results in enhanced recovery of hindlimb perfusion via upregulated proliferation of BM-derived EPCs, their enhanced mobilization activity into PB, and markedly increased recruitment into sites of ischemia. These data strongly suggest that both production of quiescent stem cells in the BM and the supply of stem cells from the BM pool for ischemic vasculogenesis may be controlled by Lnk. Furthermore, overexpression of angiogenic cytokines in Lnk-deficient KSL subpopulations suggests the importance of paracrine effects of KSL subpopulations for in situ angiogenesis as well as their autocrine effect for direct vasculogenesis. Interestingly, the results of a series of BMT experiments show that Lnk deficiency in BM-derived EPCs, but not resident EPCs/ECs, specifically augments neovascularization post hindlimb ischemia. These results provide the first direct evidence that the Lnk adapter protein plays a pivotal role in regulating the bioactivities of BM-derived EPCs for postnatal neovascularization.
Using OIR as a model for retinal damage, we also found that signs of pathogenic angiogenesis in the retina, such as tuft formation and retinal hemorrhage, were much lower in Lnk-deficient mice than in WT. Regeneration of a mature astrocyte network, along with robust neovascularization in lnk−/− mice, further supports the idea that knockdown of Lnk can have a beneficial and nonpathogenic effect in retinal vascular disease (Figure 6a through 6h and Online Figures VIII and IX). This notion may be explained by the beneficial effects of Ang-1 stimulation of vessel maturation.22 Consistent with this, quantitative RT-PCR using microdissected retinal tissue revealed higher levels of expression of ang-1 and other angiogenic cytokines, VEGF and eNOS, in Lnk-null mice than in WT (Figure 7b).
In conclusion, we provide strong evidence that Lnk is a definitive regulator of BM-EPC kinetics, including the ability to cell growth, endothelial commitment, mobilization, and recruitment for vascular regeneration. Selective targeting of Lnk may be a safe and effective approach to augment therapeutic neovascularization by EPC transplantation.
We thank Rie Ito for technical assistance and Sachie Ota for administrative support. We also thank all our colleagues for their helpful advice and encouragement. We appreciate assistance from the Research Center for Regenerative Medicine, the Teaching and Research Support Center, and the animal facility in the Tokai University School of Medicine and RIKEN Center for Developmental Biology.
Sources of Funding
This work was supported by the Academic Frontier Promotion Program of the Ministry of Education, Culture, Sports, Science, and Technology in Japan and a grant from the RIKEN Center for Developmental Biology.
↵*Both authors contributed equally to this work.
Original received December 16, 2008; revision received February 12, 2009; accepted March 17, 2009.
Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997; 275: 964–967.
Asahara T, Masuda H, Takahashi T, Kalka C, Pastore C, Silver M, Kearne M, Magner M, Isner JM. Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization. Circ Res. 1999; 85: 221–228.
Iwakura A, Luedemann C, Shastry S, Hanley A, Kearney M, Aikawa R, Isner JM, Asahara T, Losordo DW. Estrogen-mediated, endothelial nitric oxide synthase-dependent mobilization of bone marrow-derived endothelial progenitor cells contributes to reendothelialization after arterial injury. Circulation. 2003; 108: 3115–3121.
Hirai H, Samokhvalov IM, Fujimoto T, Nishikawa S, Imanishi J, Nishikawa S. Involvement of Runx1 in the down-regulation of fetal liver kinase-1 expression during transition of endothelial cells to hematopoietic cells. Blood. 2005; 106: 1948–1955.
Choi K, Kennedy M, Kazarov A, Papadimitriou JC, Keller G. A common precursor for hematopoietic and endothelial cells. Development. 1998; 125: 725–732.
Takaki S, Morita H, Tezuka Y, Takatsu K. Enhanced hematopoiesis by hematopoietic progenitor cells lacking intracellular adaptor protein, Lnk. J Exp Med. 2002; 195: 151–160.
Bailey AS, Jiang S, Afentoulis M, Baumann CI, Schroeder DA, Olson SB, Wong MH, Fleming WH. Transplanted adult hematopoietic stems cells differentiate into functional endothelial cells. Blood. 2004; 103: 13–19.
Dentelli P, Rosso A, Balsamo A, Colmenares Benedetto S, Zeoli A, Pegoraro M, Camussi G, Pegoraro L, Brizzi MF. C-KIT, by interacting with the membrane-bound ligand, recruits endothelial progenitor cells to inflamed endothelium. Blood. 2007; 109: 4264–4271.
Hiasa K, Ishibashi M, Ohtani K, Inoue S, Zhao Q, Kitamoto S, Sata M, Ichiki T, Takeshita A, Egashira K. Gene transfer of stromal cell-derived factor-1alpha enhances ischemic vasculogenesis and angiogenesis via vascular endothelial growth factor/endothelial nitric oxide synthase-related pathway: next-generation chemokine therapy for therapeutic neovascularization. Circulation. 2004; 109: 2454–2461.
Guthrie SM, Curtis LM, Mames RN, Simon GG, Grant MB, Scott EW. The nitric oxide pathway modulates hemangioblast activity of adult hematopoietic stem cells. Blood. 2005; 105: 1916–1922.
Kwon SM, Eguchi M, Wada M, Iwami Y, Hozumi K, Iwaguro H, Masuda H, Kawamoto A, Asahara T. Specific Jagged-1 signal from bone marrow microenvironment is required for endothelial progenitor cell development for neovascularization. Circulation. 2008; 118: 157–165.