| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Tumor Targeting Group, Section of Oncology & Pathology, Division of Genomic Medicine, and
Microcirculation Unit, Surgical & Anaesthetic Sciences, Division of Clinical Sciences, University of Sheffield Medical School, Sheffield S10 2RX, and
Medisys PLC, Cell Pathology Unit, University of Aberdeen, Aberdeen AB24 5UA, United Kingdom
Correspondence: Professor Claire Lewis, Tumor Targeting Group, Section of Oncology & Pathology, Division of Genomic Medicine, University of Sheffield Medical School, Beech Hill Road, Sheffield S10, United Kingdom. E-mail: claire.lewis{at}sheffield.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
Key Words: wound healing • VEGF • hypoxia
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
In wound healing, angiogenesis serves to restore vascular perfusion to damaged tissues [9 ] and to facilitate the ingress of inflammatory leukocytes such as neutrophils and macrophages, which are integral to the healing process. In pathological states, angiogenesis is often central to disease progression. Indeed, diabetic retinopathy, rheumatoid arthritis, and psoriasis have been described as "angiogenesis-dependent" diseases [14 ]. The generation of neovessels in diabetic retinopathy represents an undesirable but predictable physiological reaction to the presence of retinal ischemia. However, angiogenesis in psoriatic skin or the inflamed synovium of rheumatoid arthritis is more difficult to explain. Certainly, hypoxia is present in both diseases [15 ], partly due to the metabolic activity of the increased inflammatory cell exudate in such affected areas. This causes the levels of oxygen, glucose, and other cellular nutrients to decrease, inducing compensatory angiogenesis. The pathobiology of these two diseases, including the role of multiple factors in the promotion of angiogenesis, is reviewed in detail elsewhere [11 , 16 , 17 ].
Perhaps the most widely studied angiogenesis-dependent disease is the growth of malignant tumors. A developing tumor reaches a threshold size of around 1–2mm3, beyond which exchange of oxygen, nutrients, and waste products with surrounding tissues by diffusion is insufficient to maintain cellular viability [18 ]. This results in the development of a central area which is hypoxic and hypoglycemic, and where cell death is extensive. To grow further, the tumor needs to stimulate the formation of its own vasculature to overcome the detrimental effects of these cellular stresses [19 ].
This is controlled mainly by factors released from local cells in
the tissue microenvironment, which activate the various endothelial
cell functions required for angiogenesis. For example, vascular
endothelial growth factor (VEGF) was first characterized as a potent
endothelial cell-specific mitogen [20
]. However, it is
now known to be one of a family of related molecules, all of which
promote neovessel formation during both vasculogenesis and angiogenesis
[21
] and act as survival factors for endothelium
[22
]. Other important angiogenic agents include basic
fibroblast growth factor (bFGF), tumor necrosis factor (TNF) 
,
thymidine phosphorylase, and interleukin (IL) 4. It is important to
recognize, however, that angiogenesis is more than a simple response to
growth factors. Rather, it results from an intricate and complex local
balance of pro- and antiangiogenic agents [23
], whose
expression might vary between neighboring microenvironments. Important
examples of naturally occurring angiogenesis inhibitors include
angiostatin [24
], thrombospondin (TSP)-1, endostatin
[25
], and fibrinogen E fragment [26
].
Other factors involved in the regulation of angiogenesis include components of the ECM, which acts as a mechanical regulator of cell phenotype. For example, endothelial cells might adopt a proliferative or differentiating phenotype according to the specific components present in neighboring ECM [27 ]. Agents that modify the ECM, such as proteases, might influence angiogenesis by altering these cell-matrix interactions. The release of matrix metalloproteases (MMPs) and plasmin by local macrophages might have profound effects on the formation of new vasculature as the ECM and hence cell-matrix interactions are modified. Moreover, both proangiogenic and antiangiogenic factors have been shown to be sequestered within the intact ECM, and proteolytic modification of this matrix might release such factors into the pericellular space. bFGF, VEGF, and transforming growth factor (TGF) ß are examples of cytokines stored in this way [28 ], as they bind to heparan sulfate proteoglycans. Modification of the ECM also yields fragments of constituent proteins, which have potent effects on the angiogenic process. The most clearly defined example of this phenomenon is the proteolytic release of the angiogenesis inhibitor endostatin, a proteolytic fragment of collagen XVIII [25 ]. Other factors that indirectly influence angiogenesis include granulocyte-macrophage colony-stimulating factor, a product of many tumors and a potent chemoattractant for macrophages [29 , 30 ]. Similarly, IL-8 is released by macrophages under hypoxic conditions and attracts monocytes and other leukocytes to sites of inflammation [31 ].
The vast number of factors now known to be capable of regulating angiogenesis illustrates the complexity of the cytokine and enzyme milieu involved in maintaining a quiescent vasculature in normal tissues. Manipulation of the angiogenic process shows promise in the development of novel strategies for the treatment of disease [32 ]. For example, promotion of angiogenesis might be used to clinical advantage to restore tissue oxygenation in diseased tissues such as the ischemic limbs of diabetic patients and the ischemic myocardium [33 ].
|
RELEASE OF ANGIOGENIC STIMULI BY MACROPHAGES |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
, and interferon (IFN) 
[36
37
38
39
]. After extravasation, monocytes undergo
differentiation into macrophages and remain resident in tissues for
30–90 days before dying or emigrating to local lymph nodes
[40
].
|
-dependent and -independent
activation of macrophages [42
], illustrating the
potential for a diversity of macrophage phenotypes. It is hardly
surprising, therefore, that macrophages are capable of releasing a wide
range of molecules, including cytokines, complement components,
reactive oxygen species, proteases, inhibitors, and growth factors,
depending on their activation status. When stimulated, macrophages
secrete an array of angiogenic cytokines and growth factors
[43
], along with an array of proteolytic enzymes
[44
], illustrated in a recent paper by Moldovan et al.
[45
] which suggested that macrophages might play a role
in myocardial revascularization by "digging" channels that become
invested with endothelial cells to form capillaries. When the net
release of proangiogenic factors outweighs the local production of
angiogenesis inhibitors, the angiogenic pathway is stimulated. Activated macrophages have been shown to accumulate during angiogenesis in the corpus luteum of the postovulatory ovary and the endometrial lining of the womb [8 , 46 ]. They are also recruited to wounds (and other sites of tissue damage), primarily to phagocytose/destroy infectious agents and fragments of ECM, and to revascularize the area. Macrophages are also found in large numbers in malignant human and murine tumors, where they are often termed tumor-associated macrophages (TAMs). Recent studies have shown TAMs to be an important source of angiogenic factors [47 ] (Table 1) . This could explain, in part, why elevated levels of TAMs have been found to correlate with increased tumor angiogenesis in breast cancer [48 ]. Further evidence of a significant role for TAMs in tumor angiogenesis has been provided by studies on endometrial carcinoma [48 ], pulmonary adenocarcinoma [49 ], and malignant melanoma [50 ].
In this review we compare the proangiogenic functions performed by macrophages in two well-characterized, angiogenesis-dependent tissues: healing wounds and malignant tumors. We also review recent evidence for the role of hypoxia and other microenvironmental "stress" factors present in poorly vascularized areas of tissues, such as low pH, low glucose levels, and high lactate levels, in the regulation of macrophage function.
|
MACROPHAGES IN WOUND ANGIOGENESIS |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
The initial, inflammatory phase of murine cutaneous wounds involves
initiation of clotting and the complement cascade, and degranulation of
platelets, which mediates the influx of leukocytes and the formation of
a provisional fibrin matrix. Neutrophil influx peaks at 1–2 days
postwounding [51
], followed by macrophage infiltration,
which peaks between 2 and 4 days. During this time, macrophages are
attracted to wounds by gradients of cellular and ECM debris emanating
from the wound and by neutrophil-derived chemotactic signals
[52
]. Working together, neutrophils and macrophages
remove debris by phagocytosis and secrete chemotactic factors to
attract fibroblasts, lymphocytes, and endothelial cells into the wound.
These cell types are the major cellular constituents of mature wounds,
peaking at 
7 days, and participate in the synthesis of new matrix
and development of blood vessels during the proliferation phase of
epithelial cell expansion, reinstating tissue architecture and
restoring perfusion. Finally, the remodelling phase involves regression
of neovessels and proteolytic modification of collagen matrix to
improve mechanical strength and adhesion.
By day 2, macrophages are the predominant leukocyte in the wound space, and they persist through to the remodelling phase. They appear to perform a number of important functions in wounds. Leibovich and Ross [53 ] have demonstrated that depletion of tissue macrophages and circulating monocytes in a guinea pig model of injury results in reductions in both wound debridement and the ability of the wound to recruit fibroblasts and promote fibrosis. Wound angiogenesis was not addressed in this study, although it has recently been demonstrated that macrophages in aged mice produce significantly less VEGF than those in young mice, which results in reduced angiogenesis and slower wound repair, thereby indicating an important role for macrophages in mediating angiogenesis in wound healing. In contrast, depletion of neutrophils has no discernible effect on the process of tissue repair, although the lack of a rapid phagocytic response to contaminating bacteria might result in persistence of infection at the wound site [54 ].
Some of the earliest studies demonstrating the angiogenic activity of wound macrophages examined cells isolated from wound fluid [55 , 56 ]. Injection of these macrophages into rat corneas induces neovessel formation. Furthermore, culture supernatants from thioglycollate-induced guinea pig peritoneal macrophages have been found to be proangiogenic in a similar corneal model of angiogenesis [57 ]. Table 1 summarizes the evidence for expression of VEGF and other angiogenesis-modulating factors by wound macrophages.
Various studies have shown that wound macrophages, but not unstimulated monocytes or macrophages, are capable of inducing angiogenesis [58 ]. For example, corneal angiogenesis is accelerated only when autologous wound macrophages are used [59 ]. Similarly, adding macrophages isolated from a 3-week-old wound to a rabbit ear chamber model of wounding produces significant acceleration of angiogenesis [55 ]. These studies suggest that activation of macrophages by wound-specific microenvironmental factors is essential for their proangiogenic activity in such tissues.
The extremely low oxygen tensions (hypoxia) in areas of wounds appear to be one such wound signal (Fig. 1 ). Hypoxia occurs in early wounds due to lack of perfusion caused by damage to the vasculature and the intense metabolic activity of cells infiltrating the wound. However, it is apparent that hypoxia is not the initial driving force behind macrophage infiltration of wound tissue, since wounds have been shown to be normoxic immediately after injury [60 ]. Measurement of oxygen tensions in wound tissue has revealed both the temporal pattern of hypoxia and the extent of oxygen deprivation. Oxygen levels decrease over the first and second days postinjury, and oxygen tensions as low as 20–30 mm Hg on the first day, which steadily decreased to reach a minimum of 5–7 mm Hg by days 5–7 postinjury, have been reported in a rodent wound chamber model [61 ]. In extended studies using subcutaneous sponge-induced granulation tissue in rats, tissue oxygen tension has been shown to decrease from 28 to 22 mm Hg between days 7 and 14 postinjury [62 ]. These studies suggest a temporal pattern of hypoxia in wounds, with infiltration of the damaged tissue by macrophages and development of granulation tissue and neovessels occurring as oxygen tensions progressively fall.
|
Exposure of cells in wounds to hypoxia is essential for promoting angiogenesis in wound healing. Support for this hypothesis has been provided by studies in which wound angiogenesis is markedly reduced when the oxygen tension gradient from the center to the periphery of the wound is abolished [65 ]. Although there is now clear evidence for hypoxic regulation of the proangiogenic activity of macrophages in vitro (Table 2 ), there have been no studies investigating whether hypoxia is required for macrophage-mediated angiogenesis in wounds. A recent histological study investigating the expression of the transcription factor hypoxia-inducible factor (HIF) 1 in mouse dermal wound tissue has revealed that HIF-1 is expressed in reepithelializing keratinocytes but not in wound macrophages, even though the latter appear to produce abundant VEGF [66 ]. This is consistent either with macrophages producing such angiogenic factors in the absence of hypoxia or with VEGF being secreted in response to hypoxia via a HIF-1-independent pathway of gene induction. However, studies in our laboratories have demonstrated expression of HIF-1 by macrophages in human dermal scar tissue (Fig. 2 ), suggesting a role for hypoxia-regulated control of gene expression in wound healing. The roles of HIF-1 and other transcription factors in mediating the effects of hypoxia on the expression of various genes in macrophages are discussed below.
|
|
Other microenvironmental stress factors are also present in wounds. These include high lactate concentrations and low pH, which could affect the phenotype and secretory profile of macrophages and other wound cells (Table 3 ). These conditions arise through reduced perfusion, resulting in a switch to anaerobic metabolism in the ischemic tissue and the subsequent accumulation of lactate, with associated tissue acidosis. Lactate has been shown to increase the release of proangiogenic factors by the murine macrophage cell line RAW264.7 [70 ]. The stimulatory activity of medium conditioned by these cells was blocked by the addition of a neutralizing anti-VEGF antibody, suggesting that it was due to lactate-induced VEGF released by these cells. Indeed, lactate has recently been shown to increase VEGF production and expression in human primary macrophages [71 ].
|
in hypoxic human cultured macrophages, whereas hypoxia alone increases
expression [29
]. Interestingly, Hunt et al. have
speculated that the microenvironmental physiology of a wound might be
important in the control of collagen production by wound fibroblasts
[74
]. In summary, it is apparent that while stress factors in general might influence the release of angiogenic factors from wound macrophages, only hypoxia has been directly linked to the control of VEGF production. No evidence is available for the role of lactate, hypoglycemia, or pH in this process. However, it should be noted that hypoglycemia has been shown to influence VEGF expression in other cell types. For example, Stein et al. have demonstrated the stabilizing effect of hypoglycemia on the VEGF transcript in glioma cells [75 ] and have also shown coregulation of VEGF with other ischemia-induced genes such as the glucose transporter, GLUT-1. Furthermore, D’Arcangelo et al. have shown that acidosis increases expression of both VEGF and bFGF in endothelial cells and inhibits their apoptotic death [76 ]. Interestingly, a study on a range of breast carcinoma cell lines has demonstrated that VEGF expression responds differently to hypoxia and hypoglycemia, suggesting cell- or tissue-specific regulation of VEGF [77 ], and could in part explain why such regulation is not apparent in macrophages. It should again be stressed that much of the work performed on wound macrophages has used animal (i.e., nonhuman) models and that further research is needed to unambiguously define the effects of hypoxia and other stress factors on macrophage involvement in human wound angiogenesis. Recently, an in vitro model of angiogenesis has been described which accurately reflects many of the steps in the angiogenic pathway in vivo and could be used in this type of investigation [78 ].
|
MACROPHAGES AND TUMOR ANGIOGENESIS |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
TAMs are now known to be an important source of proangiogenic factors
such as VEGF [86
], TNF- [87
], IL-8
[88
], and bFGF [89
]. VEGF and bFGF
stimulate endothelial cell proliferation and migration
[90
], in addition to differentiation in vitro
[91
]. Angiogenesis is also facilitated by TAM-derived
proteases released in tumors. These include MMPs 1, 2, 3, 9, and 12,
and urokinase plasminogen activator, its receptor [47
,
92
], and plasmin. These enzymes release various
proangiogenic molecules bound to heparan sulfates in
proteoglycans, as well as fragments of fibrin and collagen
[93
]. Their combined action results in the degradation
of the basement membrane and other ECM components, destabilizing the
local vasculature and permitting the migration and proliferation of
endothelial cells. This proteolytic activity also contributes to the
migration and extravasation of tumor cells during the metastatic
process [94
].
TAMs produce IL-8, a CXC chemokine, which, like VEGF, binds to heparin
in the ECM and has been shown to stimulate the proliferation and
migration of endothelial cells in vitro [95
] and
angiogenesis in rat corneas in vivo [96
]. Interestingly
TAM-derived cytokines can act indirectly on angiogenesis by stimulating
TAM activity in an autocrine manner. For example, VEGF and TNF-
released by TAMs stimulate the release of tissue-type plasminogen
activator and urokinase plasminogen activator [97
,
98
].
Recruitment of new blood vessels is a two-stage process in malignant tumors. The first stage involves tumor cell expression of an angiogenic phenotype, the so-called "angiogenic switch" [99 ]. This is caused by activation of various oncogenes and/or mutation of tumor suppressor genes in tumor cells, which collectively triggers the activation of a range of proangiogenic genes [99 ]. The second phase involves coupling the first phase of tumor cell-regulated angiogenesis to the costimulatory effects of various microenvironmental stress factors present in tumors (Table 3) . Several have shown hypoxia to be a potent stimulus for the release of angiogenic factors by tumor cells in vitro and in vivo [100 , 101 ]. The presence of hypoxic areas is a typical feature of malignant human tumors; for example, in breast carcinomas, polarographic microelectrodes have been used to measure multiple areas within each tumor, demonstrating a mean pO2 of <25 mm Hg, compared to a mean pO2 of 65 mm Hg in surrounding normal breast tissue. Moreover, neovessels in tumors often malfunction due to structural and functional abnormalities [102 ]. Tumor vessels are often immature, exhibiting tortuosity, blind ends, and increased permeability, resulting in spatial heterogeneity of blood supply both between and within individual tumors [103 ].
TAMs are also attracted into and/or immobilized in avascular [104 ] and necrotic hypoxic [105 ] areas of vascularized human tumors both by specific factors produced by hypoxic tumor cells and by the direct immobilizing effect of hypoxia on macrophages [106 ]. Moreover, higher numbers of TAMs are seen to extravasate from the bloodstream into tumors containing high levels of necrosis than into those with low or no necrosis [105 ]. This is an important phenomenon, as evidence has recently emerged to show that TAMs are stimulated by hypoxia in such sites to co-operate with tumor cells to promote revascularization [84 ]. The level of hypoxia present in ischemic areas of tumors stimulates the release of VEGF from human macrophages in vitro [107 ]. Moreover, TAMs up-regulate VEGF in avascular and necrotic areas of breast carcinomas [84 , 108 ]. However, the increased production of VEGF might also be due, in part, to the presence of other "stress factors" generated in areas of poor vascular supply. Figure 2 shows the expression of HIF-1 by macrophages in specific areas of human breast carcinoma.
Glucose consumption, like oxygen consumption, generally overwhelms
supply due to the high metabolic activity of tumor cells and
compromised delivery [109
] (Table 3)
. The corresponding
accumulation of metabolic waste products, mainly lactic acid as
originally described by Warburg [110
], results in tissue
acidosis, particularly in larger, rapidly growing tumors
[111
]. An interesting aspect of tumor pH is that the
extracellular pH of tumors is often lower than that of normal tissue
(7.0 compared to 7.5), whereas intracellular pH is often greater
(7.4 compared to 7.0) [112
]. This reversal of the
normal pH gradient leads to greater invasiveness of tumor cells in
vitro, associated with enhanced proteolysis, for example, by increased
MMP activation [113
] and cell surface localization of
the serine protease cathepsin B [114
]. Although this
might not affect macrophages directly, studies have shown that low
extracellular pH affects the ability of macrophages to maintain their
intracellular pH and hence impairs function [115
]. Thus,
various TAM functions might be reduced rather than augmented in the
acidic microenvironment of a tumor. As described above, exposure to the
low pH levels found in wound tissues has no effect on the angiogenic
activity of macrophages [72
], suggesting that this
function of TAMs might not be altered by acidosis in tumors.
The effects of lactate on TAMs have not been studied directly, but inference from the preceding section might suggest that the release of angiogenic factors might be enhanced under high lactate conditions, which could contribute to tumor angiogenesis. Lactate has recently been shown to correlate with increasing metastatic activity and worsened prognosis in human head and neck cancers and cervical carcinoma [116 ]. These findings have not been associated with alteration of TAM function or angiogenesis, but they do illustrate that elevated microenvironmental lactate levels might influence tumor behavior.
In addition to their direct effects, hypoxia and/or other stress factors might also exert an indirect effect on the proangiogenic activity of TAMs. The release of cytokines/factors might be modulated, and they, in turn, recruit/stimulate TAMs in ischemic areas. For instance, tumor cells are known to produce large amounts of VEGF [34 ], which is a potent chemoattractant for macrophages [117 ] and is controlled by oxygen tension [118 ]. bFGF, another hypoxia-regulated cytokine, is also produced by tumor cells [34 ] and macrophages in vitro [119 ] and is strongly associated with angiogenesis in neoplasms [120 ]. It is also possible that activation of TAMs during extravasation across the tumor endothelium into the tumor [121 ], and/or during exposure to cytokines/signals within the tumor milieu, might modulate TAM susceptibility to stress factors.
|
COMPARISON OF MACROPHAGE FUNCTION IN WOUNDS AND TUMORS |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
Circumstantial evidence suggests that one difference between wounds and tumors might be the relative contribution of macrophages to the regulation of angiogenesis. In tumors, macrophages are only one source of angiogenic factors, the majority of which are derived from the tumor cell population and surrounding stroma (although this could vary with the angiogenic potential of the tumor cell population). We have recently demonstrated that on the rare occasions when tumor cells do not produce VEGF in breast carcinomas, TAMs represent the sole source of this important cytokine. By contrast, macrophages recruited to wounds appear to be the principal source of angiogenic factors, as well as mediating the influx of smooth muscle cells and fibroblasts. However, direct comparisons of the contributions of macrophages to angiogenesis are made almost impossible by the extreme complexity of these pathologies.
Another difference between the activities of macrophages in wound healing and tumors is their relative persistence at the target site. In wounds, macrophages remain present until debridement of debris is complete and perfusion and normoxia are restored. The repaired tissue then becomes less cellular, and fibrosis dominates. It is unclear how macrophages are removed from the wound site. They might undergo apoptosis or necrosis, enter the local vasculature, or migrate to a new site. By contrast, the heterogenous, dynamic nature of the vascular network in tumors produces areas of ischemia in which there are high concentrations of chemoattractive agents for TAMs. This causes TAMs to persist within tumors for several months, accumulating in these ischemic areas, where their angiogenic activity is most pronounced [84 , 104 ]. It might also reflect the presence of other, unknown tumor-specific factors which prevent the death or departure of TAMs from tumors via lymphatic vessels.
Of the various stress factors present in wounds and tumors, the effects of hypoxia on macrophage gene expression are becoming increasingly well characterized and have already been reviewed by us in depth [108 ]. The most germane of these are summarized briefly below. However, as many of these effects have been demonstrated in vitro, it remains to be seen whether any or all of these changes occur in macrophages in hypoxic areas of human wounds and tumors in vivo.
|
MACROPHAGE RESPONSES TO STRESS FACTORS |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
Some genes that are hypoxically up-regulated at the transcriptional
level have been identified [124
, 125
], but
only one of these, VEGF, has been demonstrated to be controlled by
oxygen tension in macrophages [118
]. Increased
transcription of VEGF is mediated by binding of the transcription
factor HIF-1 to the hypoxia response element consensus sequence
(5'-CTGT-3') in its promoter. HIF-1 is a heterodimer, consisting of an
and a ß subunit [126
]. The mRNAs for these are
constitutively expressed in the majority of cell types
[127
]. However, HIF-1 protein levels increase rapidly in
cells exposed to hypoxia, suggesting a posttranslational mechanism of
control [128
]. Recent studies have shown that HIF-1
is rapidly degraded by a ubiquitin-proteasome pathway in normoxia
[129
]. This process is inhibited by hypoxia, which
increases production, stabilization, and DNA binding of HIF-1 at
oxygen concentrations of <2%, peaking at 0.5% [130
].
HIF-1 is up-regulated in macrophages exposed to hypoxia, along with
another form of HIF, HIF-2, which comprises HIF-1ß coupled to another
subunit, HIF-2 (EPAS-1) [131
]. HIF-2 has DNA
binding/activation properties similar to those of HIF-1
[132
]. Although HIF-2 has recently been identified as
the major form of HIF produced by macrophages in solid tumors
[133
], we have now shown this to be incorrect, as HIF-1
is also produced by primary macrophages and macrophage cell lines in
hypoxia in vitro, and in avascular areas of breast carcinomas [B.
Burke, N. Tang, K. Corke, D. Tazzyman, K. Ameri, M. Wells, and C. E. Lewis, unpublished results]. A third form of the subunit,
HIF-3, has recently been characterized [135
],
suggesting that a family of related transcription factors might exist.
Up-regulation of this third form of HIF has yet to be identified in
hypoxic macrophages.
Various HIF-independent oxygen-signaling mechanisms have also been
described in macrophages. For example, the binding of activator protein
1, a complex of the c-fos and c-jun gene
products, to its DNA motif within target gene promoters is markedly
reduced when a cysteine residue within this domain is oxidized
[136
]. Similarly, expression of the transcription factor
early growth response-1 is increased under hypoxic conditions in human
monocytes and macrophages [137
]. NF-
B is also known
to be active in macrophages under hypoxic conditions
[138
].
Lactate is known to affect in vitro expression of angiogenic activity by macrophages [139 ]. However, the role of lactate in regulating specific angiogenic gene expression in these cells has not been studied in depth. Zabel et al. have postulated that enhanced production of NAD+ and poly(adenosine diphosphate-ribose) polymerase results in increased transcription of (unspecified) angiogenic factor genes, and Constant et al. have shown that VEGF expression is increased by this mechanism [71 ].
Recently, we have designed a novel gene therapy strategy that exploits these responses of macrophages to hypoxia to deliver therapeutic genes to hypoxic areas of tumors. This is important because these sites are notoriously unresponsive to conventional anticancer therapies. In this strategy, hypoxia-responsive elements from the promoter of oxygen-regulated genes are used to control the expression of a therapeutic gene inserted into macrophages. Once transfected, macrophages migrate into hypoxic areas of a given tumor mass and up-regulate their expression of the therapeutic gene [131 ]. We are currently using this method to target expression of novel antiangiogenic peptides [26 ] to hypoxic (i.e., the most highly angiogenic) areas of tumors.
This form of macrophage-mediated gene therapy could also be used to target high-level expression of proangiogenic factors specifically to ischemic tissues in wounds. The localized, hypoxia-induced expression of genes and chemoattractants produced by transfected macrophages could enhance granulation by enhancing fibroblast and endothelial cell infiltration into the wound. The tendency of macrophages to accumulate in hypoxic areas of atherosclerotic plaques, psoriatic skin, and the arthritic synovium [reviewed by us in ref. 108 ] has prompted the suggestion that this macrophage-based gene delivery system could also be used to deliver therapeutic genes to these tissues. Moreover, genes (and their protein products) expressed by hypoxic macrophages could represent targets for the development of new drugs for such diseases.
|
CONCLUDING REMARKS |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
Received April 1, 2001; revised June 12, 2001; accepted June 14, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRELEASE OF ANGIOGENIC STIMULI...MACROPHAGES IN WOUND...MACROPHAGES AND TUMOR...COMPARISON OF MACROPHAGE...MACROPHAGE RESPONSES TO STRESS...CONCLUDING REMARKSREFERENCES |
|---|
-dependent and -independent pathways of macrophage activation J. Leukoc. Biol. 67,677-682[Abstract]
and IL-1 Int. J. Cancer 85,182-188[Medline]
and HIF-1 target genes during multi-stage epidermal carcinogenesis and wound healing Cancer Res 60,6189-6195
