Mobilization of Bone MarroweCDerived Cells Enhances the Angiogenic Response to Hypoxia Without Transdifferentiation Into Endothelial Cells
http://www.100md.com
循环研究杂志 2005年第11期
the Department of Molecular Physiology and Biological Physics (T.J.O., B.R.W., G.K.O.) and Department of Biomedical Engineering (T.J.O., T.C.S.), University of Virginia, Charlottesville.
Abstract
Bone marroweCderived cells (BMCs) have been implicated as a modifiers of vascular growth either directly by transdifferentiation into endothelial cells (ECs) or indirectly through growth factor release. To examine these possibilities under physiological conditions, we developed a model of hypoxia-mediated angiogenesis in the mouse spinotrapezius muscle. This allows whole-mount analysis; therefore, the morphology and location of BMCs within the vascular network may be observed along with differentiation markers. We exposed bone marrow transplant chimeric mice to hypoxia and treated a subset with granulocyte macrophage colonyeCstimulating factor. Exposure to hypoxia caused an 13% increase in capillary density relative to control. Hypoxia did not increase the overall number of muscle-resident BMCs, but did increase the number of rounded BMCs by 25%. There was no discernable BMC contribution to the endothelium, although some BMCs assumed a pericyte morphology around capillaries. Granulocyte macrophage colonyeCstimulating factor treatment further increased the number of round BMCs within the muscle and caused a 23% increase in angiogenesis. The results of this study suggest a potentially beneficial action of BMCs during hypoxia through paracrine release of growth factors but not transdifferentiation into ECs.
Key Words: adult stem cells angiogenesis bone marrow hypoxia
Introduction
There has been intense recent interest in potential therapeutic use of bone marroweCderived cells (BMCs) for augmentation of angiogenic and regenerative responses during treatment of myocardial and peripheral ischemic disease.1 Despite evidence from human clinical trials that these therapies may have some efficacy, there is no consensus concerning how much of this effect may be attributable to BMCs giving rise to vascular cells, particularly endothelium. The reports that have shown positive investment of bone marrow cells into neovasculature have essentially all used experimental models featuring hypoxia, generally coincident with ischemia, including tumor angiogenesis, arterial ligation, and wound healing.2,3
A major component of ischemia is hypoxia, reduced tissue oxygen partial pressure. Three tissue zones typically arise during ischemia: a zone of normal blood flow; a vulnerable border zone with adequate but reduced blood flow; and a zone of insufficient flow at risk for necrosis. Growth of the capillary network, angiogenesis, has been shown to occur during ischemia, particularly in the border zone,4 attributed partially to growth factors, such as vascular endothelial growth factor (VEGF), secreted within that hypoxic region of tissue. Global hypoxia profoundly remodels pulmonary and cerebral vasculature.5 Whereas there are a number of conflicting reports regarding the effect of systemic hypoxia on skeletal muscle capillarity,6,7 previous studies by our laboratory provided clear evidence of an increase in arteriolar arcade loops after hypoxia exposure in rats.8 However, it remains to be determined whether this represents a general response in multiple species, and, as yet, no studies have investigated the role of BMCs in this response.
Although systemic hypoxia may occur naturally as a physiologic condition (eg, at high altitude) or under pathologic conditions (eg, pulmonary disease), it may also be used to understand other, more local, disease states. In particular, hypoxia may be of interest as a representation of a component or precursor of ischemia. Data from both in vitro and in vivo systems have implicated hypoxia as a critical factor in the recruitment of BMCs and differentiation into endothelial cells (ECs).9 During hypoxia an area of tissue with insufficient oxygen supply exists, but vessels remain patent and vascular modifications do not need to be immediate to salvage the tissue. Systemic hypoxia models the border zone throughout the muscle. In contrast, an infarct model, such as femoral artery ligation, creates an inflammatory infiltrate and a heterogenous tissue with regions of differing blood flow, making any beneficial action of BMCs harder to discern. It also blocks the vessels that BMCs would most likely use to reach the tissue. Thus, we hypothesized that systemic hypoxia would provide an ideal model to study the role of BMCs during angiogenesis.
A consensus exists that BMCs accumulate in areas of ischemia and inflammatory angiogenesis, but several distinct modes of action have been proposed. Asahara et al and others originally suggested transdifferentiation of BMCs into new ECs that integrated into the forming neovasculature.10,11 More recent studies, such as that by Ziegelhoeffer et al, have refuted that hypothesis, speculating that indirect angiogenic growth factor delivery by BMCs to growing vessels is responsible for beneficial effects.12 Although part of this disagreement can be attributed to use of differing angiogenic stimuli, some conflicting studies have used the same model. For example, bone marroweCderived ECs within tumors or ischemic tissue have been described as frequent, rare, or nonexistent.3,12,13 In summary, there exists no consensus on the endothelial potential of BMCs or whether such contribution occurs outside of extreme situations. We sought to address the issue of whether a fundamental stimulus of physiological magnitude was capable of recruiting or mobilizing bone marroweCderived vascular cells.
Whole-mount tissue analysis has been useful in studying vascular changes,14 but such techniques have not generally been used in combination with adult transgenic mice. The major advantages of whole-mount preparations are as follows: retention of the entire microvascular network architecture in 3 dimensions; characterization of individual cells by shape and spatial orientation; and superior colocalization with multiple fluorescent antibody labels. We describe here a novel model system that permits the first whole-mount analysis of BMCs within a complete, native vascular bed using the spinotrapezius muscle, a back stabilizer. We found clear evidence for increased capillary length density after 3 weeks of hypoxia (10% oxygen) and examined the changes occurring during hypoxia-mediated angiogenesis. In hypoxia-mediated angiogenesis, BMCs do not contribute structurally to newly forming capillaries, although many were aligned with and in close proximity to capillary vessels. Granulocyte macrophage colonyeCstimulating factor (GM-CSF) mobilization of BMCs was associated with increased angiogenesis. These results suggest that hypoxia is not a sufficient stimulus to induce BMC differentiation into ECs and that any beneficial effect of these cells is indirect rather than via direct structural incorporation into the capillary network.
Materials and Methods
An expanded Materials and Methods section is available in the online data supplement at http://circres.ahajournals.org.
Animal Model of Hypoxic Angiogenesis
The following animal procedures were approved by the Animal Care and Use Committee of the University of Virginia. Hypoxia exposure was performed as described previously.8 Briefly, mice with access to food and water ad libitum were kept in standard caging in an environment of 10% O2 and 90% N2, whereas CO2 and humidity were maintained at normal levels (n=12).
Generation of Bone Marrow Chimeric Mice
Wild-type (C57BL/6J) mice 6 to 8 weeks of age were lethally irradiated (12 Gy) (n=12). On the same day, these mice were injected with 2x106 bone marrow cells harvested from the femurs and tibias of mice expressing enhanced green fluorescent protein (EGFP) ubiquitously (C57BL/6-Tg[ACTB-EGFP]1Osb; Jackson Laboratory, Bar Harbor, Me), with the exception of erythrocytes and hair. A minimum of 8 weeks was allowed for bone marrow reconstitution, which was confirmed by flow cytometry analysis of the blood and spleen. Mice expressing LacZ ubiquitously (B6.129S7-Gt[ROSA]26Sor/J, Jackson Laboratory) were also used as bone marrow donors in preliminary experiments. For Tie-2 GFP bone marrow chimeras, Tie-2 GFP mice (FVB/N-TgN[TIE2GFP]287Sato; Jackson Laboratory) were used as donors and wild-type FVB/NJ mice were used as recipients, as this transgene was not available on a C57BL/6J background strain (n=8).
Whole-Mount Histological Examination of Spinotrapezius Muscle
Muscles were analyzed in a modified version of a procedure used previously by our laboratory.8,14 Perfused muscles were removed, then fixed and stained. Length density of capillaries (total capillary length/area of muscle) was calculated by comparing 4 consecutive 1-mm2 fields of view from the same region of each muscle.
GM-CSF Treatment
Mice were treated with IP injections of recombinant mouse GM-CSF (0.5 e/d, Pharmingen), a dose shown to be effective in vivo,15 for 10 days starting 3 days before hypoxia (n=7). Control mice received vehicle alone (NaCl, 0.9%) (n=8).
Statistical Analysis
Significance was determined using Student t tests.
Results
Systemic Hypoxia Induced an Angiogenic Response Within the Mouse Spinotrapezius Muscle In Vivo
Age-matched C57BL/6J mice were maintained in an hypoxic environment for 21 days, a duration shown to be more than sufficient to induce vascular changes in rats,8,16 at which time both spinotrapezius muscles were removed and analyzed. The capillary bed was labeled using Bandeiraea simplicifolia (BSI) lectin and imaged using laser confocal microscopy. Angiogenesis was quantified by determining the capillary length density within whole-mounted muscles. Because of the essentially 2D nature of the very thin spinotrapezius, the fluorescently labeled capillary network was visible throughout the entire thickness of the muscle (Figure 1A). Capillary length density was measured within sequential 1-mm2 fields of view and found to be increased 13% after 21 days of hypoxia exposure (P<0.01; Figure 1). A comparison of Figure 1A and 1B shows a qualitative increase in capillary length density in hypoxic compared with normoxic spinotrapezius beds. Given the recent interest in the potential role of bone marrow cells during vessel growth, we applied this model to examine BMCs during moderate angiogenesis.
BMC Density Within Muscle Tissue Was Unchanged After Hypoxia-Mediated Angiogenesis
To identify BMCs within the muscle, bone marrow transplant chimeric mice were created by injecting bone marrow from mice ubiquitously expressing a marker gene into irradiated wild type mice. Transplanting bone marrow from a mouse transgenic for GFP (C57BL/6-Tg[ACTB-EGFP]1Osb) demonstrated a much higher sensitivity for BMC detection than -galactosidase (B6.129S7-Gt[ROSA]26Sor/J) (Figure 2A and 2B). The differences in transgenic constructs suggest that the difference may be attributable to unequal expression levels of the marker gene or sensitivity of the indicator assay. To quantitate the degree of reconstitution within solid organs other than muscle, spleens from representative mice were homogenized and analyzed by flow cytometry (Figure 2C and 2D). Both CD11b+ and CD45+ cells demonstrated 80% to 90% expression of the GFP gene, indicating good reconstitution both within subsets of leukocytes and the leukocyte compartment as a whole. The entire vasculature was perfused to remove any cells within the capillary lumen, then fixed, labeled, and imaged. A significant number of GFP+ cells were detectable throughout the muscle tissue, without any stimulus (Figure 2B). Chimeric mice were then exposed to hypoxia and the muscles of these mice were imaged and analyzed.
Hypoxia Induced an Increase in BMC Density in Proximity to Capillaries
The magnitude of angiogenesis was unchanged between wild-type and transplanted mice, similar to the results seen in other studies.12 Nearly all muscle resident BMCs fell clearly into 1 of 2 morphological categories: rounded or elongated (Figures 3 and 4 ). Elongated BMCs were defined as having a length:width ratio of greater than 2:1. Round BMCs predominantly expressed CD45 and CD11b, whereas these markers were less prevalent on elongated BMCs (Figure 4E and 4F). Elongated BMCs were half as prevalent within the capillary bed as round BMCs (Figure 3C), but 3 times more likely to be perivascular under normal conditions (35% compared with 11%, Figure 5), defined as being within 5 e of and aligned parallel to a capillary.
Given the previous reports of BMC accumulation in areas of growing tumor vasculature17 or arteriolar enlargement,12 we hypothesized that a similar accumulation might be evident after hypoxia-mediated angiogenesis. Sequential fields of view were analyzed and >8000 BMCs (n=12) were analyzed based on shape and location. Overall, BMC density and the density of elongated BMCs within the muscle tissue were unchanged, whereas the density of round BMCs was increased by 25% (Figure 3C). When the position of BMCs relative to capillaries was considered, round BMCs were much more likely to be found in a perivascular position in hypoxia-exposed animals as compared with muscles of control mice (from 11% to 18%, P<0.01; Figure 5). Conversely, the proximity of elongated BMCs to capillaries was unchanged after hypoxia-mediated angiogenesis.
Hypoxia Did Not Induce BMC Transdifferentiation Into Capillary Endothelial Cells or Smooth Muscle Cells
Previously published reports have suggested that BMCs may differentiate into ECs if angiogenesis is induced in bone marrow transplant chimeric mice.13,18 Many of these experimental models, such as wound healing and ischemia, include a hypoxic component.3,19 Hypoxia has also been shown to induce BMC differentiation toward an endothelial phenotype in vitro20 and has been speculated to be the critical factor in the formation of bone marroweCderived endothelium.21 Given the large number of BMCs residing within the spinotrapezius muscle tissue, we speculated that our model might also demonstrate the presence of bone marroweCderived ECs. More than 1000 fields of view were scrutinized by confocal microscopy for BMCs that had integrated into the capillary network or expressed endothelial markers. Many cells had assumed a potentially endothelial morphology, including some contiguous to the vasculature. In general, these BMCs were nonoverlapping, suggestive of pericytic rather than endothelial differentiation (Figure 4). In rare instances, BMCs were distinctly wrapped around a capillary but clearly distinct from the endothelial layer (Figure 4D). More than 10 000 capillary segments were examined, and only 3 were identified where BMCs appeared to have potentially integrated into the vascular wall. Optical slicing confocal microscopy was then used to determine whether the BMCs were trapped within vessels. In each of these cases, the GFP signal was located within the capillary lumen and distinct from the endothelium (Figure 6A and 6B).
To confirm our findings, further confocal microscopic studies were conducted labeling capillaries with antibodies against additional endothelial markers CD31 and Tie-2 (Figure 6C and 6D). As in the previous studies, there was a total absence of GFP+ cells that coexpressed endothelial surface markers. Studies with chimeric mice that had been allowed to recover for more than 6 months after transplant showed a similar lack of bone marroweCderived ECs, indicating that reconstitution was not a significant factor. Tie-2 LacZ mice have been used as bone marrow donors to distinguish bone marroweCderived ECs from infiltrating leukocytes.22 Because of our improved visualization of GFP compared with LacZ as a BMC marker, we used Tie-2 GFP as a bone marrow donor. Whereas the entire capillary network was GFP+ in donor mice (Figure 6E), no GFP+ cells could be detected in wild-type mice receiving Tie-2 GFP bone marrow (Figure 6F), independent of hypoxia exposure. We, therefore, conclude that none of the ECs formed during hypoxia-mediated angiogenesis derive from the bone marrow, despite numerous neighboring BMCs.
We additionally colabeled muscles from GFP bone marrow chimeras with an antibody for -smooth muscle actin (-SMA). Whereas BMCs could be seen in apposition to arterioles and venules as well as capillaries (Figure 6G and 6H), they did not express -SMA and did not assume a morphology consistent with smooth muscle cells. Despite a perivascular location, elongated BMCs do not differentiate into smooth muscle cells or -SMA+ pericytes in our model.
BMC Mobilization Increased Muscle Resident BMCs and Hypoxia-Mediated Angiogenesis
GM-CSF has been shown to increase numbers of circulating15 or engrafting23 BMCs with endothelial potential. To rule out a lack of progenitor cell mobilization as a responsible factor for the absence of bone marroweCderived ECs, GFP bone marrow transplant chimeric mice were treated with GM-CSF for 10 days, concurrent with the initiation of hypoxia exposure. GM-CSF administration was associated with a 27% increase in the number of muscle-resident BMCs within the vascular bed (Figure 7C), as well as nearly doubling the increase in capillary length density (23% compared with 13%; Figure 7E). However, no bone marroweCderived ECs were detected. We conclude from these results that the lack of BMC differentiation into endothelium during hypoxia is not a result of insufficient mobilization. Angiogenesis within skeletal muscle during moderate hypoxia does not appear to be sufficient for generation of ECs derived from bone marrow, although it is possible these cells play an important paracrine role in the angiogenic response.
BMC Expression of Growth Factors and Matrix Metalloproteinases
To identify potential candidates that regulate BMC influence on hypoxia-mediated angiogenesis, we stained with antibodies for VEGF and matrix metalloproteinase 9 (MMP-9), 2 BMC-derived factors shown to be capable of influencing vascular growth by disparate mechanisms and potentially involved in a paracrine effect on angiogenesis during hypoxia.24,25 Although high autofluorescence made detailed quantification impossible, a higher number of BMCs, both elongated and round, were positive for VEGF after hypoxia exposure (Figure 8). In addition, more BMCs were found to be expressing MMP-9 after hypoxia exposure, although in this case, staining was almost exclusively restricted to round BMCs. Although this does not prove that hypoxia-mediated angiogenesis is dependent on BMC release of either factor, this further suggests a role for BMCs in angiogenesis of physiologic magnitude.
Discussion
In this article, we describe for the first time a model of hypoxia-induced angiogenesis within mouse skeletal muscle. After 21 days of hypoxia exposure, we found a 13% increase in spinotrapezius capillary length density. This result was consistent with previous hypoxia studies of Smith and Marshall in the rat spinotrapezius16 and of Deveci et al in the rat diaphragm,7 a muscle similar in composition to the spinotrapezius. However, of major significance, despite examining more than 10 000 capillary segments and 8000 BMCs that invested within the vascular bed, we found no compelling evidence that any of these cells transdifferentiated into ECs. We did observe 3 of 8000 BMCs where there may have been coincident expression of EGFP and the endothelial marker BSI lectin. Confocal microscopy, however, demonstrated that these were circulating cells trapped within capillaries rather than integrated into the capillary wall. This interpretation is supported by a complete lack of GFP+ cells detected when Tie-2 GFP mice were used as bone marrow donors rather than CX1 GFP mice. Although it is possible that the lack of bone marroweCderived ECs seen in this model could be attributable to a negative effect of hypoxia on the bone marrow, our observation of increased BMCs within the muscle suggest this is not responsible for negative findings. Taken together, our results provide compelling evidence that BMC transdifferentiation does not play a significant role in the angiogenic response to systemic hypoxia.
The present results are in marked contrast to those of Aicher et al3 and Asahara et al,13 who reported that BMCs were a source of ECs in skeletal muscle. Our results are consistent with those of a recent report by Ziegelhoeffer et al, who found no evidence of transdifferentiation of BMCs into ECs within a femoral artery ligation model of ischemia.12 There are several key aspects of our studies that imparted a higher degree of resolution than some previous studies reporting BMC transdifferentiation into endothelium. First, early studies, even those using fluorescence, did not use confocal microscopy and, as such, may not have adequately resolved single cells, as is necessary to clearly determine whether BMCs expressed endothelial markers. Second, the whole-mount analyses used in our hypoxia model offered several advantages as compared with analysis of tissue sections including: (1) a much larger sample size in terms of BMC numbers; (2) spatial information on the location of BMCs within the vascular bed; and (3) more effective perfusion to remove BMCs that are nonadherent or weakly associated with the capillary luminal surface. Third, rigorous assessment of expression of endothelial markers within BMCs by multiple antibodies and bone marrow donors was performed, because studies using transgenic Tie-2 LacZ mice as donors13 are confounded by the fact that expression of this marker has been shown not to be entirely restricted to ECs.17
The most compelling evidence for bone marroweCderived endothelium has been found in models that included revascularization of a previously avascular area7 or in association with severe necrosis.26 Human studies have shown evidence that intense vascular injury induces progenitor mobilization and BMC transdifferentiation in cases of tumor or organ transplant.27 These results have led to speculation that the severity of injury is a major determinant of the frequency of eventual transdifferentiation of BMCs into ECs.15 A critical question is whether the differences in these models versus the current study are simply a function of the magnitude of the investment of inflammatory cells or whether there are key differences in the nature of stimuli associated with necrosis/severe injury versus systemic hypoxia that are important in inducing progenitors within BMC populations to differentiate into ECs. We believe the latter is more likely, based on observations that we saw no evidence of transdifferentiation of BMCs into ECs despite the presence of very large numbers of BMCs in close proximity to capillaries in our model, including cases where BMC numbers were augmented by GM-CSF administration. These results support the hypothesis that the lack of integration into the vasculature was not attributable to lack of mobilization but to lack of a signal for transdifferentiation. Although femoral artery ligation may induce a hypoxia sufficient to mobilize BMCs into the circulation,15 an ischemia severe enough to cause necrosis may be necessary to induce endothelial transdifferentiation.
An unexpected, but very intriguing, observation in the present studies was that we detected a large number of cells within the muscle tissue before stimulus. BMCs are generally described as sparse within unperturbed muscle of GFP bone marrow chimeric mice, regardless of whether the studies report BMC transdifferentiation.12,28 The differences in BMC frequency seen between previous studies and ours is likely a function of the greater sensitivity and ease in quantifying BMCs in our whole-mount preparation. In any case, these results are of potential major importance in that they suggest that, even in normal tissues, there is a large resident population of BMC derived cells. As such, reports in the literature claiming the existence of local populations of local mesenchymal stem cells may in fact be derived from bone marrow hematopoietic cell lineages and simply reside for prolonged periods of time within tissues, possibly retaining plasticity even in aged animals.29
On exposure of the chimeric mice to hypoxia, we detected a small increase in the subset of rounded BMCs within the spinotrapezius muscle. More compelling, we also found that this subset was more likely to be in a perivascular position after hypoxia exposure. Again, this shares similarities with a phenomenon described in arterioles: an accumulation of BMCs around collateral vessels.30 The 2 major subsets of BMCs known to reside within muscle tissue are macrophages and satellite cells. Satellite cells have a role in producing growth factors aiding muscle regeneration after injury but do not have a known function in vascular growth. Conversely, macrophage density has been correlated with an increase in angiogenesis.31 The shift in location of round BMCs during hypoxia-mediated angiogenesis is, therefore, of interest, as it suggests that these cells may play an important paracrine role in modulating the angiogenic response to systemic hypoxia. We further showed production of both VEGF and MMP-9 by increased numbers of BMCs during hypoxia, and both factors are capable of supporting angiogenesis through effects on ECs or extracellular matrix, respectively.24,25 Macrophages are known to produce a variety of EC growth factors in response to hypoxia.32 Kinnaird et al demonstrated that injected BMCs assumed a position next to collateral arterioles and released growth factors in a paracrine manner, leading to arteriolar enlargement.25 Lyden et al demonstrated a crucial role for BMCs in tumor growth, whereby wild-type bone marrow transplant rescued tumor angiogenesis in a mutant mouse.33 Although this was initially attributed to BMC contribution to the endothelium, it may have been attributable to lack of growth factor production by BMCs during initial stages of tumor vascularization. This hypothesis is supported by more recent studies showing perivascular, but not endothelial, BMCs within growing tumors.17,34 Results of the present studies showing that administration of GM-CSF augmented the angiogenic response to systemic hypoxia and increased the frequency of BMCs in close proximity to capillaries provides further support for the hypothesis that BMCs may play a key paracrine role in mediating angiogenesis during hypoxia.
In conclusion, systemic hypoxia is capable of inducing angiogenesis within mouse skeletal muscle, but despite a high density of BMCs within the muscle tissue before hypoxia, we found no evidence that BMCs are a significant source of endothelial or smooth muscle cells within the neovasculature. However, systemic hypoxia was associated with a marked increase in the frequency of BMCs within the microvascular bed, including many rounded cells in close proximity to capillaries. The results of this study suggest a potentially beneficial action of BMCs during systemic hypoxia through a paracrine release of growth factors but not transdifferentiation into endothelial cells.
Acknowledgments
This work was supported by NIH grants HL65958 and 5T32 HL007284. We thank Dr Lisa Palmer for assistance in developing the hypoxia model and Janet Gorman and Gina Wimer for technical assistance.
References
Lee MS, Makkar RR. Stem-cell transplantation in myocardial infarction: a status report. Ann Intern Med. 2004; 140: 729eC737.
Rafii S, Lyden D. Therapeutic stem and progenitor cell transplantation for organ vascularization and regeneration. Nat Med. 2003; 9: 702eC712.
Aicher A, Heeschen C, Mildner-Rihm C, Urbich C, Ihling C, Technau-Ihling K, Zeiher AM, Dimmeler S. Essential role of endothelial nitric oxide synthase for mobilization of stem and progenitor cells. Nat Med. 2003; 9: 1370eC1376.
Olivetti G, Ricci R, Beghi C, Guideri G, Anversa P. Response of the border zone to myocardial infarction in rats. Am J Pathol. 1986; 125: 476eC483.
Pichiule P, LaManna JC. Angiopoietin-2 and rat brain capillary remodeling during adaptation and deadaptation to prolonged mild hypoxia. J Appl Physiol. 2002; 93: 1131eC1139.
Hoppeler H. Vascular growth in hypoxic skeletal muscle. Adv Exp Med Biol. 1999; 474: 277eC286.
Deveci D, Marshall JM, Egginton S. Chronic hypoxia induces prolonged angiogenesis in skeletal muscles of rat. Exp Physiol. 2002; 87: 287eC291.
Price RJ, Skalak TC. Arteriolar remodeling in skeletal muscle of rats exposed to chronic hypoxia. J Vasc Res. 1998; 35: 238eC244.
Li TS, Hamano K, Suzuki K, Ito H, Zempo N, Matsuzaki M. Improved angiogenic potency by implantation of ex vivo hypoxia prestimulated bone marrow cells in rats. Am J Physiol Heart Circ Physiol. 2002; 283: H468eCH473.
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: 964eC966.
Grant MB, May WS, Caballero S, Brown GA, Guthrie SM, Mames RN, Byrne BJ, Vaught T, Spoerri PE, Peck AB, Scott EW. Adult hematopoietic stem cells provide functional hemangioblast activity during retinal neovascularization. Nat Med. 2002; 8: 607eC612.
Ziegelhoeffer T, Fernandez B, Kostin S, Heil M, Voswinckel R, Helisch A, Schaper W. Bone marrow-derived cells do not incorporate into the adult growing vasculature. Circ Res. 2004; 94: 230eC238.
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: 221eC228.
Price RJ, Owens GK, Skalak TC. Immunohistochemical identification of arteriolar development using markers of smooth muscle differentiation. Evidence that capillary arterialization proceeds from terminal arterioles. Circ Res. 1994; 75: 520eC527.
Takahashi T, Kalka C, Masuda H, Chen D, Silver M, Kearney M, Magner M, Isner JM, Asahara T. Ischemia- and cytokine-induced mobilization of bone marrow-derived endothelial progenitor cells for neovascularization. Nat Med. 1999; 5: 434eC438.
Smith K, Marshall JM. Physiological adjustments and arteriolar remodelling within skeletal muscle during acclimation to chronic hypoxia in the rat. J Physiol (Lond). 1999; 521: 261eC272.
De Palma M, Venneri MA, Roca C, Naldini L. Targeting exogenous genes to tumor angiogenesis by transplantation of genetically modified hematopoietic stem cells. Nat Med. 2003; 9: 789eC795.
Espinosa-Heidmann DG, Caicedo A, Hernandez EP, Csaky KG, Cousins SW. Bone marrow-derived progenitor cells contribute to experimental choroidal neovascularization. Invest Ophthalmol Vis Sci. 2003; 44: 4914eC4919.
Badiavas EV, Abedi M, Butmarc J, Falanga V, Quesenberry P. Participation of bone marrow derived cells in cutaneous wound healing. J Cell Physiol. 2003; 196: 245eC250.
Akita T, Murohara T, Ikeda H, Sasaki Ki, Shimada T, Egami K, Imaizumi T. Hypoxic preconditioning augments efficacy of human endothelial progenitor cells for therapeutic neovascularization. Lab Invest. 2003; 83: 65eC73.
Rabelink TJ, de Boer HC, de Koning EJP, van Zonneveld AJ. Endothelial progenitor cells: more than an inflammatory response Arterioscler Thromb Vasc Biol. 2004; 24: 834eC838.
Asahara T, Takahashi T, Masuda H, Kalka C, Chen D, Iwaguro H, Inai Y, Silver M, Isner JM. VEGF contributes to postnatal neovascularization by mobilizing bone marrow-derived endothelial progenitor cells. EMBO J. 1999; 18: 3964eC3972.
Cho HJ, Kim HS, Lee MM, Kim DH, Yang HJ, Hur J, Hwang KK, Oh S, Choi YJ, Chae IH, Oh BH, Choi YS, Walsh K, Park YB. Mobilized endothelial progenitor cells by granulocyte-macrophage colony-stimulating factor accelerate reendothelialization and reduce vascular inflammation after intravascular radiation. Circulation. 2003; 108: 2918eC2925.
Silvestre JS, Mallat Z, Tamarat R, Duriez M, Tedgui A, Levy BI. Regulation of matrix metalloproteinase activity in ischemic tissue by interleukin-10: role in ischemia-induced angiogenesis. Circ Res. 2001; 89: 259eC264.
Kinnaird T, Stabile E, Burnett MS, Shou M, Lee CW, Barr S, Fuchs S, Epstein SE. Local delivery of marrow-derived stromal cells augments collateral perfusion through paracrine mechanisms. Circulation. 2004; 109: 1543eC1549.
Hess DC, Hill WD, Martin-Studdard A, Carroll J, Brailer J, Carothers J. Bone marrow as a source of endothelial cells and NeuN-expressing cells after stroke. Stroke. 2002; 33: 1362eC1368.
Peters BA, Diaz LA, Polyak K, Meszler L, Romans K, Guinan EC, Antin JH, Myerson D, Hamilton SR, Vogelstein B, Kinzler KW, Lengauer C. Contribution of bone marrow-derived endothelial cells to human tumor vasculature. Nat Med. 2005; 11: 261eC262.
Wagers AJ, Sherwood RI, Christensen JL, Weissman IL. Little evidence for developmental plasticity of adult hematopoietic stem cells. Science. 2002; 297: 2256eC2259.
Conboy IM, Conboy MJ, Wagers AJ, Girma ER, Weissman IL, Rando TA. Rejuvenation of aged progenitor cells by exposure to a young systemic environment. Nature. 2005; 433: 760eC764.
Heil M, Ziegelhoeffer T, Pipp F, Kostin S, Martin S, Clauss M, Schaper W. Blood monocyte concentration is critical for enhancement of collateral artery growth. Am J Physiol Heart Circ Physiol. 2002; 283: H2411eCH2419.
Manoonkitiwongsa PS, Jackson-Friedman C, McMillan PJ, Schultz RL, Lyden PD. Angiogenesis after stroke is correlated with increased numbers of macrophages: the clean-up hypothesis. J Cereb Blood Flow Metab. 2001; 21: 1223eC1231.
Lewis JS, Lee JA, Underwood JC, Harris AL, Lewis CE. Macrophage responses to hypoxia: relevance to disease mechanisms. J Leukoc Biol. 1999; 66: 889eC900.
Lyden D, Hattori K, Dias S, Costa C, Blaikie P, Butros L, Chadburn A, Heissig B, Marks W, Witte L, Wu Y, Hicklin D, Zhu Z, Hackett NR, Crystal RG, Moore MA, Hajjar KA, Manova K, Benezra R, Rafii S. Impaired recruitment of bone-marrow-derived endothelial and hematopoietic precursor cells blocks tumor angiogenesis and growth. Nat Med. 2001; 7: 1194eC1201.
Gothert JR, Gustin SE, van Eekelen JA, Schmidt U, Hall MA, Jane SM, Green AR, Gottgens B, Izon DJ, Begley CG. Genetically tagging endothelial cells in vivo: bone marrow-derived cells do not contribute to tumor endothelium. Blood. 2004; 104: 1769eC1777.(Thomas J. O’Neill, IV, Br)
Abstract
Bone marroweCderived cells (BMCs) have been implicated as a modifiers of vascular growth either directly by transdifferentiation into endothelial cells (ECs) or indirectly through growth factor release. To examine these possibilities under physiological conditions, we developed a model of hypoxia-mediated angiogenesis in the mouse spinotrapezius muscle. This allows whole-mount analysis; therefore, the morphology and location of BMCs within the vascular network may be observed along with differentiation markers. We exposed bone marrow transplant chimeric mice to hypoxia and treated a subset with granulocyte macrophage colonyeCstimulating factor. Exposure to hypoxia caused an 13% increase in capillary density relative to control. Hypoxia did not increase the overall number of muscle-resident BMCs, but did increase the number of rounded BMCs by 25%. There was no discernable BMC contribution to the endothelium, although some BMCs assumed a pericyte morphology around capillaries. Granulocyte macrophage colonyeCstimulating factor treatment further increased the number of round BMCs within the muscle and caused a 23% increase in angiogenesis. The results of this study suggest a potentially beneficial action of BMCs during hypoxia through paracrine release of growth factors but not transdifferentiation into ECs.
Key Words: adult stem cells angiogenesis bone marrow hypoxia
Introduction
There has been intense recent interest in potential therapeutic use of bone marroweCderived cells (BMCs) for augmentation of angiogenic and regenerative responses during treatment of myocardial and peripheral ischemic disease.1 Despite evidence from human clinical trials that these therapies may have some efficacy, there is no consensus concerning how much of this effect may be attributable to BMCs giving rise to vascular cells, particularly endothelium. The reports that have shown positive investment of bone marrow cells into neovasculature have essentially all used experimental models featuring hypoxia, generally coincident with ischemia, including tumor angiogenesis, arterial ligation, and wound healing.2,3
A major component of ischemia is hypoxia, reduced tissue oxygen partial pressure. Three tissue zones typically arise during ischemia: a zone of normal blood flow; a vulnerable border zone with adequate but reduced blood flow; and a zone of insufficient flow at risk for necrosis. Growth of the capillary network, angiogenesis, has been shown to occur during ischemia, particularly in the border zone,4 attributed partially to growth factors, such as vascular endothelial growth factor (VEGF), secreted within that hypoxic region of tissue. Global hypoxia profoundly remodels pulmonary and cerebral vasculature.5 Whereas there are a number of conflicting reports regarding the effect of systemic hypoxia on skeletal muscle capillarity,6,7 previous studies by our laboratory provided clear evidence of an increase in arteriolar arcade loops after hypoxia exposure in rats.8 However, it remains to be determined whether this represents a general response in multiple species, and, as yet, no studies have investigated the role of BMCs in this response.
Although systemic hypoxia may occur naturally as a physiologic condition (eg, at high altitude) or under pathologic conditions (eg, pulmonary disease), it may also be used to understand other, more local, disease states. In particular, hypoxia may be of interest as a representation of a component or precursor of ischemia. Data from both in vitro and in vivo systems have implicated hypoxia as a critical factor in the recruitment of BMCs and differentiation into endothelial cells (ECs).9 During hypoxia an area of tissue with insufficient oxygen supply exists, but vessels remain patent and vascular modifications do not need to be immediate to salvage the tissue. Systemic hypoxia models the border zone throughout the muscle. In contrast, an infarct model, such as femoral artery ligation, creates an inflammatory infiltrate and a heterogenous tissue with regions of differing blood flow, making any beneficial action of BMCs harder to discern. It also blocks the vessels that BMCs would most likely use to reach the tissue. Thus, we hypothesized that systemic hypoxia would provide an ideal model to study the role of BMCs during angiogenesis.
A consensus exists that BMCs accumulate in areas of ischemia and inflammatory angiogenesis, but several distinct modes of action have been proposed. Asahara et al and others originally suggested transdifferentiation of BMCs into new ECs that integrated into the forming neovasculature.10,11 More recent studies, such as that by Ziegelhoeffer et al, have refuted that hypothesis, speculating that indirect angiogenic growth factor delivery by BMCs to growing vessels is responsible for beneficial effects.12 Although part of this disagreement can be attributed to use of differing angiogenic stimuli, some conflicting studies have used the same model. For example, bone marroweCderived ECs within tumors or ischemic tissue have been described as frequent, rare, or nonexistent.3,12,13 In summary, there exists no consensus on the endothelial potential of BMCs or whether such contribution occurs outside of extreme situations. We sought to address the issue of whether a fundamental stimulus of physiological magnitude was capable of recruiting or mobilizing bone marroweCderived vascular cells.
Whole-mount tissue analysis has been useful in studying vascular changes,14 but such techniques have not generally been used in combination with adult transgenic mice. The major advantages of whole-mount preparations are as follows: retention of the entire microvascular network architecture in 3 dimensions; characterization of individual cells by shape and spatial orientation; and superior colocalization with multiple fluorescent antibody labels. We describe here a novel model system that permits the first whole-mount analysis of BMCs within a complete, native vascular bed using the spinotrapezius muscle, a back stabilizer. We found clear evidence for increased capillary length density after 3 weeks of hypoxia (10% oxygen) and examined the changes occurring during hypoxia-mediated angiogenesis. In hypoxia-mediated angiogenesis, BMCs do not contribute structurally to newly forming capillaries, although many were aligned with and in close proximity to capillary vessels. Granulocyte macrophage colonyeCstimulating factor (GM-CSF) mobilization of BMCs was associated with increased angiogenesis. These results suggest that hypoxia is not a sufficient stimulus to induce BMC differentiation into ECs and that any beneficial effect of these cells is indirect rather than via direct structural incorporation into the capillary network.
Materials and Methods
An expanded Materials and Methods section is available in the online data supplement at http://circres.ahajournals.org.
Animal Model of Hypoxic Angiogenesis
The following animal procedures were approved by the Animal Care and Use Committee of the University of Virginia. Hypoxia exposure was performed as described previously.8 Briefly, mice with access to food and water ad libitum were kept in standard caging in an environment of 10% O2 and 90% N2, whereas CO2 and humidity were maintained at normal levels (n=12).
Generation of Bone Marrow Chimeric Mice
Wild-type (C57BL/6J) mice 6 to 8 weeks of age were lethally irradiated (12 Gy) (n=12). On the same day, these mice were injected with 2x106 bone marrow cells harvested from the femurs and tibias of mice expressing enhanced green fluorescent protein (EGFP) ubiquitously (C57BL/6-Tg[ACTB-EGFP]1Osb; Jackson Laboratory, Bar Harbor, Me), with the exception of erythrocytes and hair. A minimum of 8 weeks was allowed for bone marrow reconstitution, which was confirmed by flow cytometry analysis of the blood and spleen. Mice expressing LacZ ubiquitously (B6.129S7-Gt[ROSA]26Sor/J, Jackson Laboratory) were also used as bone marrow donors in preliminary experiments. For Tie-2 GFP bone marrow chimeras, Tie-2 GFP mice (FVB/N-TgN[TIE2GFP]287Sato; Jackson Laboratory) were used as donors and wild-type FVB/NJ mice were used as recipients, as this transgene was not available on a C57BL/6J background strain (n=8).
Whole-Mount Histological Examination of Spinotrapezius Muscle
Muscles were analyzed in a modified version of a procedure used previously by our laboratory.8,14 Perfused muscles were removed, then fixed and stained. Length density of capillaries (total capillary length/area of muscle) was calculated by comparing 4 consecutive 1-mm2 fields of view from the same region of each muscle.
GM-CSF Treatment
Mice were treated with IP injections of recombinant mouse GM-CSF (0.5 e/d, Pharmingen), a dose shown to be effective in vivo,15 for 10 days starting 3 days before hypoxia (n=7). Control mice received vehicle alone (NaCl, 0.9%) (n=8).
Statistical Analysis
Significance was determined using Student t tests.
Results
Systemic Hypoxia Induced an Angiogenic Response Within the Mouse Spinotrapezius Muscle In Vivo
Age-matched C57BL/6J mice were maintained in an hypoxic environment for 21 days, a duration shown to be more than sufficient to induce vascular changes in rats,8,16 at which time both spinotrapezius muscles were removed and analyzed. The capillary bed was labeled using Bandeiraea simplicifolia (BSI) lectin and imaged using laser confocal microscopy. Angiogenesis was quantified by determining the capillary length density within whole-mounted muscles. Because of the essentially 2D nature of the very thin spinotrapezius, the fluorescently labeled capillary network was visible throughout the entire thickness of the muscle (Figure 1A). Capillary length density was measured within sequential 1-mm2 fields of view and found to be increased 13% after 21 days of hypoxia exposure (P<0.01; Figure 1). A comparison of Figure 1A and 1B shows a qualitative increase in capillary length density in hypoxic compared with normoxic spinotrapezius beds. Given the recent interest in the potential role of bone marrow cells during vessel growth, we applied this model to examine BMCs during moderate angiogenesis.
BMC Density Within Muscle Tissue Was Unchanged After Hypoxia-Mediated Angiogenesis
To identify BMCs within the muscle, bone marrow transplant chimeric mice were created by injecting bone marrow from mice ubiquitously expressing a marker gene into irradiated wild type mice. Transplanting bone marrow from a mouse transgenic for GFP (C57BL/6-Tg[ACTB-EGFP]1Osb) demonstrated a much higher sensitivity for BMC detection than -galactosidase (B6.129S7-Gt[ROSA]26Sor/J) (Figure 2A and 2B). The differences in transgenic constructs suggest that the difference may be attributable to unequal expression levels of the marker gene or sensitivity of the indicator assay. To quantitate the degree of reconstitution within solid organs other than muscle, spleens from representative mice were homogenized and analyzed by flow cytometry (Figure 2C and 2D). Both CD11b+ and CD45+ cells demonstrated 80% to 90% expression of the GFP gene, indicating good reconstitution both within subsets of leukocytes and the leukocyte compartment as a whole. The entire vasculature was perfused to remove any cells within the capillary lumen, then fixed, labeled, and imaged. A significant number of GFP+ cells were detectable throughout the muscle tissue, without any stimulus (Figure 2B). Chimeric mice were then exposed to hypoxia and the muscles of these mice were imaged and analyzed.
Hypoxia Induced an Increase in BMC Density in Proximity to Capillaries
The magnitude of angiogenesis was unchanged between wild-type and transplanted mice, similar to the results seen in other studies.12 Nearly all muscle resident BMCs fell clearly into 1 of 2 morphological categories: rounded or elongated (Figures 3 and 4 ). Elongated BMCs were defined as having a length:width ratio of greater than 2:1. Round BMCs predominantly expressed CD45 and CD11b, whereas these markers were less prevalent on elongated BMCs (Figure 4E and 4F). Elongated BMCs were half as prevalent within the capillary bed as round BMCs (Figure 3C), but 3 times more likely to be perivascular under normal conditions (35% compared with 11%, Figure 5), defined as being within 5 e of and aligned parallel to a capillary.
Given the previous reports of BMC accumulation in areas of growing tumor vasculature17 or arteriolar enlargement,12 we hypothesized that a similar accumulation might be evident after hypoxia-mediated angiogenesis. Sequential fields of view were analyzed and >8000 BMCs (n=12) were analyzed based on shape and location. Overall, BMC density and the density of elongated BMCs within the muscle tissue were unchanged, whereas the density of round BMCs was increased by 25% (Figure 3C). When the position of BMCs relative to capillaries was considered, round BMCs were much more likely to be found in a perivascular position in hypoxia-exposed animals as compared with muscles of control mice (from 11% to 18%, P<0.01; Figure 5). Conversely, the proximity of elongated BMCs to capillaries was unchanged after hypoxia-mediated angiogenesis.
Hypoxia Did Not Induce BMC Transdifferentiation Into Capillary Endothelial Cells or Smooth Muscle Cells
Previously published reports have suggested that BMCs may differentiate into ECs if angiogenesis is induced in bone marrow transplant chimeric mice.13,18 Many of these experimental models, such as wound healing and ischemia, include a hypoxic component.3,19 Hypoxia has also been shown to induce BMC differentiation toward an endothelial phenotype in vitro20 and has been speculated to be the critical factor in the formation of bone marroweCderived endothelium.21 Given the large number of BMCs residing within the spinotrapezius muscle tissue, we speculated that our model might also demonstrate the presence of bone marroweCderived ECs. More than 1000 fields of view were scrutinized by confocal microscopy for BMCs that had integrated into the capillary network or expressed endothelial markers. Many cells had assumed a potentially endothelial morphology, including some contiguous to the vasculature. In general, these BMCs were nonoverlapping, suggestive of pericytic rather than endothelial differentiation (Figure 4). In rare instances, BMCs were distinctly wrapped around a capillary but clearly distinct from the endothelial layer (Figure 4D). More than 10 000 capillary segments were examined, and only 3 were identified where BMCs appeared to have potentially integrated into the vascular wall. Optical slicing confocal microscopy was then used to determine whether the BMCs were trapped within vessels. In each of these cases, the GFP signal was located within the capillary lumen and distinct from the endothelium (Figure 6A and 6B).
To confirm our findings, further confocal microscopic studies were conducted labeling capillaries with antibodies against additional endothelial markers CD31 and Tie-2 (Figure 6C and 6D). As in the previous studies, there was a total absence of GFP+ cells that coexpressed endothelial surface markers. Studies with chimeric mice that had been allowed to recover for more than 6 months after transplant showed a similar lack of bone marroweCderived ECs, indicating that reconstitution was not a significant factor. Tie-2 LacZ mice have been used as bone marrow donors to distinguish bone marroweCderived ECs from infiltrating leukocytes.22 Because of our improved visualization of GFP compared with LacZ as a BMC marker, we used Tie-2 GFP as a bone marrow donor. Whereas the entire capillary network was GFP+ in donor mice (Figure 6E), no GFP+ cells could be detected in wild-type mice receiving Tie-2 GFP bone marrow (Figure 6F), independent of hypoxia exposure. We, therefore, conclude that none of the ECs formed during hypoxia-mediated angiogenesis derive from the bone marrow, despite numerous neighboring BMCs.
We additionally colabeled muscles from GFP bone marrow chimeras with an antibody for -smooth muscle actin (-SMA). Whereas BMCs could be seen in apposition to arterioles and venules as well as capillaries (Figure 6G and 6H), they did not express -SMA and did not assume a morphology consistent with smooth muscle cells. Despite a perivascular location, elongated BMCs do not differentiate into smooth muscle cells or -SMA+ pericytes in our model.
BMC Mobilization Increased Muscle Resident BMCs and Hypoxia-Mediated Angiogenesis
GM-CSF has been shown to increase numbers of circulating15 or engrafting23 BMCs with endothelial potential. To rule out a lack of progenitor cell mobilization as a responsible factor for the absence of bone marroweCderived ECs, GFP bone marrow transplant chimeric mice were treated with GM-CSF for 10 days, concurrent with the initiation of hypoxia exposure. GM-CSF administration was associated with a 27% increase in the number of muscle-resident BMCs within the vascular bed (Figure 7C), as well as nearly doubling the increase in capillary length density (23% compared with 13%; Figure 7E). However, no bone marroweCderived ECs were detected. We conclude from these results that the lack of BMC differentiation into endothelium during hypoxia is not a result of insufficient mobilization. Angiogenesis within skeletal muscle during moderate hypoxia does not appear to be sufficient for generation of ECs derived from bone marrow, although it is possible these cells play an important paracrine role in the angiogenic response.
BMC Expression of Growth Factors and Matrix Metalloproteinases
To identify potential candidates that regulate BMC influence on hypoxia-mediated angiogenesis, we stained with antibodies for VEGF and matrix metalloproteinase 9 (MMP-9), 2 BMC-derived factors shown to be capable of influencing vascular growth by disparate mechanisms and potentially involved in a paracrine effect on angiogenesis during hypoxia.24,25 Although high autofluorescence made detailed quantification impossible, a higher number of BMCs, both elongated and round, were positive for VEGF after hypoxia exposure (Figure 8). In addition, more BMCs were found to be expressing MMP-9 after hypoxia exposure, although in this case, staining was almost exclusively restricted to round BMCs. Although this does not prove that hypoxia-mediated angiogenesis is dependent on BMC release of either factor, this further suggests a role for BMCs in angiogenesis of physiologic magnitude.
Discussion
In this article, we describe for the first time a model of hypoxia-induced angiogenesis within mouse skeletal muscle. After 21 days of hypoxia exposure, we found a 13% increase in spinotrapezius capillary length density. This result was consistent with previous hypoxia studies of Smith and Marshall in the rat spinotrapezius16 and of Deveci et al in the rat diaphragm,7 a muscle similar in composition to the spinotrapezius. However, of major significance, despite examining more than 10 000 capillary segments and 8000 BMCs that invested within the vascular bed, we found no compelling evidence that any of these cells transdifferentiated into ECs. We did observe 3 of 8000 BMCs where there may have been coincident expression of EGFP and the endothelial marker BSI lectin. Confocal microscopy, however, demonstrated that these were circulating cells trapped within capillaries rather than integrated into the capillary wall. This interpretation is supported by a complete lack of GFP+ cells detected when Tie-2 GFP mice were used as bone marrow donors rather than CX1 GFP mice. Although it is possible that the lack of bone marroweCderived ECs seen in this model could be attributable to a negative effect of hypoxia on the bone marrow, our observation of increased BMCs within the muscle suggest this is not responsible for negative findings. Taken together, our results provide compelling evidence that BMC transdifferentiation does not play a significant role in the angiogenic response to systemic hypoxia.
The present results are in marked contrast to those of Aicher et al3 and Asahara et al,13 who reported that BMCs were a source of ECs in skeletal muscle. Our results are consistent with those of a recent report by Ziegelhoeffer et al, who found no evidence of transdifferentiation of BMCs into ECs within a femoral artery ligation model of ischemia.12 There are several key aspects of our studies that imparted a higher degree of resolution than some previous studies reporting BMC transdifferentiation into endothelium. First, early studies, even those using fluorescence, did not use confocal microscopy and, as such, may not have adequately resolved single cells, as is necessary to clearly determine whether BMCs expressed endothelial markers. Second, the whole-mount analyses used in our hypoxia model offered several advantages as compared with analysis of tissue sections including: (1) a much larger sample size in terms of BMC numbers; (2) spatial information on the location of BMCs within the vascular bed; and (3) more effective perfusion to remove BMCs that are nonadherent or weakly associated with the capillary luminal surface. Third, rigorous assessment of expression of endothelial markers within BMCs by multiple antibodies and bone marrow donors was performed, because studies using transgenic Tie-2 LacZ mice as donors13 are confounded by the fact that expression of this marker has been shown not to be entirely restricted to ECs.17
The most compelling evidence for bone marroweCderived endothelium has been found in models that included revascularization of a previously avascular area7 or in association with severe necrosis.26 Human studies have shown evidence that intense vascular injury induces progenitor mobilization and BMC transdifferentiation in cases of tumor or organ transplant.27 These results have led to speculation that the severity of injury is a major determinant of the frequency of eventual transdifferentiation of BMCs into ECs.15 A critical question is whether the differences in these models versus the current study are simply a function of the magnitude of the investment of inflammatory cells or whether there are key differences in the nature of stimuli associated with necrosis/severe injury versus systemic hypoxia that are important in inducing progenitors within BMC populations to differentiate into ECs. We believe the latter is more likely, based on observations that we saw no evidence of transdifferentiation of BMCs into ECs despite the presence of very large numbers of BMCs in close proximity to capillaries in our model, including cases where BMC numbers were augmented by GM-CSF administration. These results support the hypothesis that the lack of integration into the vasculature was not attributable to lack of mobilization but to lack of a signal for transdifferentiation. Although femoral artery ligation may induce a hypoxia sufficient to mobilize BMCs into the circulation,15 an ischemia severe enough to cause necrosis may be necessary to induce endothelial transdifferentiation.
An unexpected, but very intriguing, observation in the present studies was that we detected a large number of cells within the muscle tissue before stimulus. BMCs are generally described as sparse within unperturbed muscle of GFP bone marrow chimeric mice, regardless of whether the studies report BMC transdifferentiation.12,28 The differences in BMC frequency seen between previous studies and ours is likely a function of the greater sensitivity and ease in quantifying BMCs in our whole-mount preparation. In any case, these results are of potential major importance in that they suggest that, even in normal tissues, there is a large resident population of BMC derived cells. As such, reports in the literature claiming the existence of local populations of local mesenchymal stem cells may in fact be derived from bone marrow hematopoietic cell lineages and simply reside for prolonged periods of time within tissues, possibly retaining plasticity even in aged animals.29
On exposure of the chimeric mice to hypoxia, we detected a small increase in the subset of rounded BMCs within the spinotrapezius muscle. More compelling, we also found that this subset was more likely to be in a perivascular position after hypoxia exposure. Again, this shares similarities with a phenomenon described in arterioles: an accumulation of BMCs around collateral vessels.30 The 2 major subsets of BMCs known to reside within muscle tissue are macrophages and satellite cells. Satellite cells have a role in producing growth factors aiding muscle regeneration after injury but do not have a known function in vascular growth. Conversely, macrophage density has been correlated with an increase in angiogenesis.31 The shift in location of round BMCs during hypoxia-mediated angiogenesis is, therefore, of interest, as it suggests that these cells may play an important paracrine role in modulating the angiogenic response to systemic hypoxia. We further showed production of both VEGF and MMP-9 by increased numbers of BMCs during hypoxia, and both factors are capable of supporting angiogenesis through effects on ECs or extracellular matrix, respectively.24,25 Macrophages are known to produce a variety of EC growth factors in response to hypoxia.32 Kinnaird et al demonstrated that injected BMCs assumed a position next to collateral arterioles and released growth factors in a paracrine manner, leading to arteriolar enlargement.25 Lyden et al demonstrated a crucial role for BMCs in tumor growth, whereby wild-type bone marrow transplant rescued tumor angiogenesis in a mutant mouse.33 Although this was initially attributed to BMC contribution to the endothelium, it may have been attributable to lack of growth factor production by BMCs during initial stages of tumor vascularization. This hypothesis is supported by more recent studies showing perivascular, but not endothelial, BMCs within growing tumors.17,34 Results of the present studies showing that administration of GM-CSF augmented the angiogenic response to systemic hypoxia and increased the frequency of BMCs in close proximity to capillaries provides further support for the hypothesis that BMCs may play a key paracrine role in mediating angiogenesis during hypoxia.
In conclusion, systemic hypoxia is capable of inducing angiogenesis within mouse skeletal muscle, but despite a high density of BMCs within the muscle tissue before hypoxia, we found no evidence that BMCs are a significant source of endothelial or smooth muscle cells within the neovasculature. However, systemic hypoxia was associated with a marked increase in the frequency of BMCs within the microvascular bed, including many rounded cells in close proximity to capillaries. The results of this study suggest a potentially beneficial action of BMCs during systemic hypoxia through a paracrine release of growth factors but not transdifferentiation into endothelial cells.
Acknowledgments
This work was supported by NIH grants HL65958 and 5T32 HL007284. We thank Dr Lisa Palmer for assistance in developing the hypoxia model and Janet Gorman and Gina Wimer for technical assistance.
References
Lee MS, Makkar RR. Stem-cell transplantation in myocardial infarction: a status report. Ann Intern Med. 2004; 140: 729eC737.
Rafii S, Lyden D. Therapeutic stem and progenitor cell transplantation for organ vascularization and regeneration. Nat Med. 2003; 9: 702eC712.
Aicher A, Heeschen C, Mildner-Rihm C, Urbich C, Ihling C, Technau-Ihling K, Zeiher AM, Dimmeler S. Essential role of endothelial nitric oxide synthase for mobilization of stem and progenitor cells. Nat Med. 2003; 9: 1370eC1376.
Olivetti G, Ricci R, Beghi C, Guideri G, Anversa P. Response of the border zone to myocardial infarction in rats. Am J Pathol. 1986; 125: 476eC483.
Pichiule P, LaManna JC. Angiopoietin-2 and rat brain capillary remodeling during adaptation and deadaptation to prolonged mild hypoxia. J Appl Physiol. 2002; 93: 1131eC1139.
Hoppeler H. Vascular growth in hypoxic skeletal muscle. Adv Exp Med Biol. 1999; 474: 277eC286.
Deveci D, Marshall JM, Egginton S. Chronic hypoxia induces prolonged angiogenesis in skeletal muscles of rat. Exp Physiol. 2002; 87: 287eC291.
Price RJ, Skalak TC. Arteriolar remodeling in skeletal muscle of rats exposed to chronic hypoxia. J Vasc Res. 1998; 35: 238eC244.
Li TS, Hamano K, Suzuki K, Ito H, Zempo N, Matsuzaki M. Improved angiogenic potency by implantation of ex vivo hypoxia prestimulated bone marrow cells in rats. Am J Physiol Heart Circ Physiol. 2002; 283: H468eCH473.
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: 964eC966.
Grant MB, May WS, Caballero S, Brown GA, Guthrie SM, Mames RN, Byrne BJ, Vaught T, Spoerri PE, Peck AB, Scott EW. Adult hematopoietic stem cells provide functional hemangioblast activity during retinal neovascularization. Nat Med. 2002; 8: 607eC612.
Ziegelhoeffer T, Fernandez B, Kostin S, Heil M, Voswinckel R, Helisch A, Schaper W. Bone marrow-derived cells do not incorporate into the adult growing vasculature. Circ Res. 2004; 94: 230eC238.
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: 221eC228.
Price RJ, Owens GK, Skalak TC. Immunohistochemical identification of arteriolar development using markers of smooth muscle differentiation. Evidence that capillary arterialization proceeds from terminal arterioles. Circ Res. 1994; 75: 520eC527.
Takahashi T, Kalka C, Masuda H, Chen D, Silver M, Kearney M, Magner M, Isner JM, Asahara T. Ischemia- and cytokine-induced mobilization of bone marrow-derived endothelial progenitor cells for neovascularization. Nat Med. 1999; 5: 434eC438.
Smith K, Marshall JM. Physiological adjustments and arteriolar remodelling within skeletal muscle during acclimation to chronic hypoxia in the rat. J Physiol (Lond). 1999; 521: 261eC272.
De Palma M, Venneri MA, Roca C, Naldini L. Targeting exogenous genes to tumor angiogenesis by transplantation of genetically modified hematopoietic stem cells. Nat Med. 2003; 9: 789eC795.
Espinosa-Heidmann DG, Caicedo A, Hernandez EP, Csaky KG, Cousins SW. Bone marrow-derived progenitor cells contribute to experimental choroidal neovascularization. Invest Ophthalmol Vis Sci. 2003; 44: 4914eC4919.
Badiavas EV, Abedi M, Butmarc J, Falanga V, Quesenberry P. Participation of bone marrow derived cells in cutaneous wound healing. J Cell Physiol. 2003; 196: 245eC250.
Akita T, Murohara T, Ikeda H, Sasaki Ki, Shimada T, Egami K, Imaizumi T. Hypoxic preconditioning augments efficacy of human endothelial progenitor cells for therapeutic neovascularization. Lab Invest. 2003; 83: 65eC73.
Rabelink TJ, de Boer HC, de Koning EJP, van Zonneveld AJ. Endothelial progenitor cells: more than an inflammatory response Arterioscler Thromb Vasc Biol. 2004; 24: 834eC838.
Asahara T, Takahashi T, Masuda H, Kalka C, Chen D, Iwaguro H, Inai Y, Silver M, Isner JM. VEGF contributes to postnatal neovascularization by mobilizing bone marrow-derived endothelial progenitor cells. EMBO J. 1999; 18: 3964eC3972.
Cho HJ, Kim HS, Lee MM, Kim DH, Yang HJ, Hur J, Hwang KK, Oh S, Choi YJ, Chae IH, Oh BH, Choi YS, Walsh K, Park YB. Mobilized endothelial progenitor cells by granulocyte-macrophage colony-stimulating factor accelerate reendothelialization and reduce vascular inflammation after intravascular radiation. Circulation. 2003; 108: 2918eC2925.
Silvestre JS, Mallat Z, Tamarat R, Duriez M, Tedgui A, Levy BI. Regulation of matrix metalloproteinase activity in ischemic tissue by interleukin-10: role in ischemia-induced angiogenesis. Circ Res. 2001; 89: 259eC264.
Kinnaird T, Stabile E, Burnett MS, Shou M, Lee CW, Barr S, Fuchs S, Epstein SE. Local delivery of marrow-derived stromal cells augments collateral perfusion through paracrine mechanisms. Circulation. 2004; 109: 1543eC1549.
Hess DC, Hill WD, Martin-Studdard A, Carroll J, Brailer J, Carothers J. Bone marrow as a source of endothelial cells and NeuN-expressing cells after stroke. Stroke. 2002; 33: 1362eC1368.
Peters BA, Diaz LA, Polyak K, Meszler L, Romans K, Guinan EC, Antin JH, Myerson D, Hamilton SR, Vogelstein B, Kinzler KW, Lengauer C. Contribution of bone marrow-derived endothelial cells to human tumor vasculature. Nat Med. 2005; 11: 261eC262.
Wagers AJ, Sherwood RI, Christensen JL, Weissman IL. Little evidence for developmental plasticity of adult hematopoietic stem cells. Science. 2002; 297: 2256eC2259.
Conboy IM, Conboy MJ, Wagers AJ, Girma ER, Weissman IL, Rando TA. Rejuvenation of aged progenitor cells by exposure to a young systemic environment. Nature. 2005; 433: 760eC764.
Heil M, Ziegelhoeffer T, Pipp F, Kostin S, Martin S, Clauss M, Schaper W. Blood monocyte concentration is critical for enhancement of collateral artery growth. Am J Physiol Heart Circ Physiol. 2002; 283: H2411eCH2419.
Manoonkitiwongsa PS, Jackson-Friedman C, McMillan PJ, Schultz RL, Lyden PD. Angiogenesis after stroke is correlated with increased numbers of macrophages: the clean-up hypothesis. J Cereb Blood Flow Metab. 2001; 21: 1223eC1231.
Lewis JS, Lee JA, Underwood JC, Harris AL, Lewis CE. Macrophage responses to hypoxia: relevance to disease mechanisms. J Leukoc Biol. 1999; 66: 889eC900.
Lyden D, Hattori K, Dias S, Costa C, Blaikie P, Butros L, Chadburn A, Heissig B, Marks W, Witte L, Wu Y, Hicklin D, Zhu Z, Hackett NR, Crystal RG, Moore MA, Hajjar KA, Manova K, Benezra R, Rafii S. Impaired recruitment of bone-marrow-derived endothelial and hematopoietic precursor cells blocks tumor angiogenesis and growth. Nat Med. 2001; 7: 1194eC1201.
Gothert JR, Gustin SE, van Eekelen JA, Schmidt U, Hall MA, Jane SM, Green AR, Gottgens B, Izon DJ, Begley CG. Genetically tagging endothelial cells in vivo: bone marrow-derived cells do not contribute to tumor endothelium. Blood. 2004; 104: 1769eC1777.(Thomas J. O’Neill, IV, Br)