Use of Immunodeficient Mice for the Evaluation of CXC Chemokines in the Regulation of Tumor-associated Angiogenesis

All authors of this article are from the Department of Internal Medicine, Division of Pulmonary and Critical Medicine, The University of Michigan Medical School, Ann Arbor, Michigan. Robert M. Strieter, M.D., is a Professor; Christina L. Addison, Ph.D., is a Post-doctoral Research Fellow; Jan E. Ehlert, Ph.D., is a Post-doctoral Research Fellow; Michael P. Keane, M.B., is an Assistant Professor; John A. Belperio, M.D., is a Research Fellow; Marie D. Burdick, B.S., is Chief Technician in Dr. Strieter’s Laboratory; and Douglas A. Arenberg, M.D., is an Assistant Professor. All authors of this article are from the Department of Internal Medicine, Division of Pulmonary and Critical Medicine, The University of Michigan Medical School, Ann Arbor, Michigan. Robert M. Strieter, M.D., is a Professor; Christina L. Addison, Ph.D., is a Post-doctoral Research Fellow; Jan E. Ehlert, Ph.D., is a Post-doctoral Research Fellow; Michael P. Keane, M.B., is an Assistant Professor; John A. Belperio, M.D., is a Research Fellow; Marie D. Burdick, B.S., is Chief Technician in Dr. Strieter’s Laboratory; and Douglas A. Arenberg, M.D., is an Assistant Professor. All authors of this article are from the Department of Internal Medicine, Division of Pulmonary and Critical Medicine, The University of Michigan Medical School, Ann Arbor, Michigan. Robert M. Strieter, M.D., is a Professor; Christina L. Addison, Ph.D., is a Post-doctoral Research Fellow; Jan E. Ehlert, Ph.D., is a Post-doctoral Research Fellow; Michael P. Keane, M.B., is an Assistant Professor; John A. Belperio, M.D., is a Research Fellow; Marie D. Burdick, B.S., is Chief Technician in Dr. Strieter’s Laboratory; and Douglas A. Arenberg, M.D., is an Assistant Professor. All authors of this article are from the Department of Internal Medicine, Division of Pulmonary and Critical Medicine, The University of Michigan Medical School, Ann Arbor, Michigan. Robert M. Strieter, M.D., is a Professor; Christina L. Addison, Ph.D., is a Post-doctoral Research Fellow; Jan E. Ehlert, Ph.D., is a Post-doctoral Research Fellow; Michael P. Keane, M.B., is an Assistant Professor; John A. Belperio, M.D., is a Research Fellow; Marie D. Burdick, B.S., is Chief Technician in Dr. Strieter’s Laboratory; and Douglas A. Arenberg, M.D., is an Assistant Professor. All authors of this article are from the Department of Internal Medicine, Division of Pulmonary and Critical Medicine, The University of Michigan Medical School, Ann Arbor, Michigan. Robert M. Strieter, M.D., is a Professor; Christina L. Addison, Ph.D., is a Post-doctoral Research Fellow; Jan E. Ehlert, Ph.D., is a Post-doctoral Research Fellow; Michael P. Keane, M.B., is an Assistant Professor; John A. Belperio, M.D., is a Research Fellow; Marie D. Burdick, B.S., is Chief Technician in Dr. Strieter’s Laboratory; and Douglas A. Arenberg, M.D., is an Assistant Professor. All authors of this article are from the Department of Internal Medicine, Division of Pulmonary and Critical Medicine, The University of Michigan Medical School, Ann Arbor, Michigan. Robert M. Strieter, M.D., is a Professor; Christina L. Addison, Ph.D., is a Post-doctoral Research Fellow; Jan E. Ehlert, Ph.D., is a Post-doctoral Research Fellow; Michael P. Keane, M.B., is an Assistant Professor; John A. Belperio, M.D., is a Research Fellow; Marie D. Burdick, B.S., is Chief Technician in Dr. Strieter’s Laboratory; and Douglas A. Arenberg, M.D., is an Assistant Professor. All authors of this article are from the Department of Internal Medicine, Division of Pulmonary and Critical Medicine, The University of Michigan Medical School, Ann Arbor, Michigan. Robert M. Strieter, M.D., is a Professor; Christina L. Addison, Ph.D., is a Post-doctoral Research Fellow; Jan E. Ehlert, Ph.D., is a Post-doctoral Research Fellow; Michael P. Keane, M.B., is an Assistant Professor; John A. Belperio, M.D., is a Research Fellow; Marie D. Burdick, B.S., is Chief Technician in Dr. Strieter’s Laboratory; and Douglas A. Arenberg, M.D., is an Assistant Professor. The use of human tumor xenografts in immunodeficient mice has provided significant insight into the biology of tumor growth and metastasis and has enabled scientists to study the complex biology of human tumor growth more effectively. Work from our laboratory and others’ supports the notion that net tumor-derived angiogenesis during tumorigenesis of human tumors is determined, in part, by an imbalance in favor of the overexpression of angiogenic compared with angiostatic CXC chemokines ( Figure 1 ). This paradigm predicts an environment that favors angiogenesis and tumorigenesis and supports the potential for spontaneous metastases. The purpose of this article is to describe the use of immunodeficient mice as an animal model system to characterize the qualitative and quantitative presence of these angiogenic and angiostatic CXC chemokines during tumorigenesis and to determine their net contribution to human tumorigenesis and metastasis in vivo. Various cancer cell lines have been used and xenografted into immunodeficient mice (such as severe combined immunodeficiency [SCID 1 ] or nude mice) to create a human tumor/mouse chimeras. The results of these studies have provided evidence that an imbalance in the biology of angiogenic versus angiostatic CXC chemokines supports a significant portion of human tumorderived angiogenesis leading to augmented tumorigenesis and spontaneous metastases. Moreover, using the immunodeficient mouse model system has made possible the identification of potential novel strategies to manipulate therapeutically the imbalance of angiogenic compared with angiostatic CXC chemokines, which may be directly translational to human disease. Net tumor-derived angiogenesis during tumorigenesis is determined, in part, by an imbalance in favor of the overexpression of angiogenic (ELR + ), compared with angiostatic (ELR~) CXC chemokines. This paradigm predicts an environment that favors angiogenesis, and tumorigenesis and supports the potential for spontaneous metastases. In contrast, this paradigm also supports the contention that inhibition of angiogenic or augmentation of angiostatic CXC chemokine biology will lead to reduced tumor growth and metastases. Net tumor-derived angiogenesis during tumorigenesis is determined, in part, by an imbalance in favor of the overexpression of angiogenic (ELR + ), compared with angiostatic (ELR~) CXC chemokines. This paradigm predicts an environment that favors angiogenesis, and tumorigenesis and supports the potential for spontaneous metastases. In contrast, this paradigm also supports the contention that inhibition of angiogenic or augmentation of angiostatic CXC chemokine biology will lead to reduced tumor growth and metastases. Angiogenesis is the growth of new blood vessels from preexisting vessels and capillaries. This biological process is critical to a variety of physiological processes like embryogenesis and wound repair. However, it is also a process that plays a devastating role in a number of pathological processes such as tumorigenesis ( Auerbach 1981 ; Auerbach and others 1976 ; Folkman 1995 , 1993 , 1985 ; Folkman and Brem 1992 ; Folkman and Cotran 1976 ; Folkman and Klagsbrun 1987 ; Leibovich and Weisman 1988 ; Polverini 1996 ; Polverini and others 1977 ). The regulation of angiogenesis depends on a dual, yet opposing, balance of local factors that promote or inhibit neovascularization. For example, the rate of normal capillary endothelial cell turnover in adults is typically measured in months or years ( Engerman and others 1967 ; Tannock and Hayashi 1972 ), suggesting a balance between angiogenic and angiostatic factors under homeostatic conditions. By contrast, the development of granulation tissue of wounds shifts the balance in favor of a predominance of angiogenic factors leading to new functioning capillaries within a matter of days ( Leibovich and Weisman 1988 ). Angiogenesis of wound granulation tissue is locally controlled and transient. These qualities support the notion that the following process occurs in wound granulation tissue: a marked reduction in the elaboration of angiogenic factors and/or a simultaneous increase in the level of factors that inhibit neovascularization ( Bouck 1992 ). In contrast to the precise regulation in wound repair, dysregulation of angiogenesis can lead to an imbalance in the relationship of angiogenic and angiostatic factors, which favors persistent net angiogenesis that contributes to the pathogenesis of tumor growth and metastases. The complement of angiogenic and angiostatic factors may vary among different physiological and pathological settings. However, the recognition of this dual mechanism of control is critical to gain insight into this complex process and understand the regulation of angiogenesis in association with tumorigenesis. A variety of factors have been described that promote angiogenesis ( Auerbach 1981 ; Folkman 1985 , 1993 , 1996 , 1997 , 1998 ; Folkman and Brem 1992 ; Folkman and Klagsbrun 1987 ; Gastl and others 1997 ; Hotfilder and others 1997 ; Hui and Ignoffo 1998 ; Kumar and Fidler 1998 ; Lund and others 1998 ; Pluda 1997 ; Risau 1997 ; Zetter 1998 ; Ziche and others 1996 ). Perhaps the best studied of the angiogenic factors are vascular endothelial growth factor (VEGF1) and basic fibroblast growth factor (bFGF 1 ) ( Gastl and others 1997 ; Hotfilder and others 1997 ; Hui and Ignoffo 1998 ; Kumar and Fidler 1998 ; Lund and others 1998 ; Pluda 1997 ; Risau 1997 ; Zetter 1998 ; Ziche and others 1996 ). In contrast, both angiostatin and endostatin have been found to display marked inhibition of angiogenesis ( O’Reilly and others 1994 , 1997 ). Although these factors are important in the regulation of angiogenesis in both physiological and pathophysiological processes, these molecules do not fully account for all of the modulation of the neovascular response in pathological conditions, such as tumorigenesis. CXC chemokines are characteristically heparin binding proteins. On a structural level, they have four highly conserved cysteine amino acid residues, with the first two cysteines separated by one nonconserved amino acid residue, hence the name CXC ( Adams and Lloyd 1997 ; Baggiolini 1998 ; Baggiolini and others 1994 ; Baggiolini and others 1997 ; Balkwill 1998 ; Luster 1998 ; Rollins 1997 ; Strieter and Kunkel 1997 ; Taub and Oppenheim 1994 ; Walz and others 1996 ). Although the CXC motif distinguishes this family from other chemokine families, a second structural domain within this family dictates their angiogenic potential. The NH 2 -terminus of the majority of the CXC chemokines containing three amino acid residues (Glu-Leu-Arg: the “ELR” motif) precedes the first cysteine amino acid residue of the primary structure of these cytokines ( Adams and Lloyd 1997 ; Baggiolini 1998 ; Baggiolini and others 1994 ,1997; Balkwill 1998 ; Luster 1998 ; Rollins 1997 ; Strieter and Kunkel 1997 ; Taub and Oppenheim 1994 ; Walz and others 1996 ). The family members that contain the ELR motif (ELR + ) are potent promoters of angiogenesis in physiological concentrations of 1 to 10 nM ( Strieter and others 1995 ) ( Table 1 ). In contrast, members that lack the ELR motif (ELR − ) are potent inhibitors of angiogenesis in physiological concentrations of 500 pM to 1 nM ( Strieter and others 1995 ). This difference suggests on a structural/functional level that members of the CXC chemokine family are unique cytokines in their ability to behave in a disparate manner in the regulation of angiogenesis. The angiogenic members include interleukin (IL 1 )-8, epithelial neutrophil activating protein (ENA 1 )-78, growthrelated oncogenes (GRO 1 -α, β, and γ), granulocyte chemotactic protein-2, and NH 2 -terminal truncated forms of platelet basic protein (PBP 1 ), which include connective tissue activating protein-III, beta-thromboglobulin, and neutrophil activating protein-2 ( Hu and others 1993 ; Koch and others 1992 ; Strieter and others 1992 , 1995 ). ELR + and ELR − CXC chemokines, which are angiogenic and angiostatic factors, respectively lnterleukin-8 (IL-8) Epithelial neutrophil activating protein-78 (ENA-78) Growth-related oncogene alpha (GRO-α) Growth-related oncogene beta (GRO-β) Growth-related oncogene gamma (GRO-γ) Granulocyte chemotactic protein-2 (GCP-2) Platelet basic protein (PBP) Connective tissue activating protein-III (CTAP-III) Beta-thromboglobulin (β-TG) Neutrophil activating protein-2 (NAP-2) Platelet factor-4 (PF4) Interferon-γ-inducible protein (IP-10) Monokine induced by interferon-γ (MIG) lnterleukin-8 (IL-8) Epithelial neutrophil activating protein-78 (ENA-78) Growth-related oncogene alpha (GRO-α) Growth-related oncogene beta (GRO-β) Growth-related oncogene gamma (GRO-γ) Granulocyte chemotactic protein-2 (GCP-2) Platelet basic protein (PBP) Connective tissue activating protein-III (CTAP-III) Beta-thromboglobulin (β-TG) Neutrophil activating protein-2 (NAP-2) Platelet factor-4 (PF4) Interferon-γ-inducible protein (IP-10) Monokine induced by interferon-γ (MIG) ELR + and ELR − CXC chemokines, which are angiogenic and angiostatic factors, respectively lnterleukin-8 (IL-8) Epithelial neutrophil activating protein-78 (ENA-78) Growth-related oncogene alpha (GRO-α) Growth-related oncogene beta (GRO-β) Growth-related oncogene gamma (GRO-γ) Granulocyte chemotactic protein-2 (GCP-2) Platelet basic protein (PBP) Connective tissue activating protein-III (CTAP-III) Beta-thromboglobulin (β-TG) Neutrophil activating protein-2 (NAP-2) Platelet factor-4 (PF4) Interferon-γ-inducible protein (IP-10) Monokine induced by interferon-γ (MIG) lnterleukin-8 (IL-8) Epithelial neutrophil activating protein-78 (ENA-78) Growth-related oncogene alpha (GRO-α) Growth-related oncogene beta (GRO-β) Growth-related oncogene gamma (GRO-γ) Granulocyte chemotactic protein-2 (GCP-2) Platelet basic protein (PBP) Connective tissue activating protein-III (CTAP-III) Beta-thromboglobulin (β-TG) Neutrophil activating protein-2 (NAP-2) Platelet factor-4 (PF4) Interferon-γ-inducible protein (IP-10) Monokine induced by interferon-γ (MIG) The angiostatic (ELR − ) members of the CXC chemokine family include platelet factor-4 (PF4) (Thomas and others 1970), monokine induced by interferon-γ (MIG 1 ) (Farber 1992, 1993, 1990, 1997), and interferon (IFN 1 )-γ-inducible protein (IP 1 )-10 (Farber 1990, 1992, 1993, 1997; Luster and Ravetch 1987; Luster and others 1985) ( Table 1 ). Although interferon-inducible T-cell alpha chemoattractant and stromal cell-derived factor (SDF 1 )-1 are additional ELR- CXC chemokines, it remains unclear whether they inhibit angiogenesis. SDF-1 has been found to induce in vitro migration of human umbilical vein endothelial cells (Gupta and others 1998). In contrast, SDF-1 has also been found to attenuate the in vivo angiogenic activity of either ELR + CXC chemokines, bFGF, or VEGF using the rat cornea micropocket (CMP 1 ) assay of neovascularization ( Arenberg and others 1997 ). IP-10 can be induced by all three interferons (IFN-α, β and γ) (Farber 1990, 1992, 1993, 1997; Luster 1998 ; Luster and Ravetch 1987; Luster and others 1985). MIG is unique in that it is only induced by IFN-γ (Farber 1992, 1993, 1990, 1997; Luster 1998 ; Luster and Ravetch 1987; Luster and others 1985). Therefore, interferons and other cytokines that can induce the expression of interferons (IL-12 and IL-18) may have a profound effect on the production of IP-10 and MIG ( Table 2 ). Although interferons induce the production of the angiostatic CXC chemokines IP-10 and MIG, they attenuate the expression of the angiogenic CXC chemokines IL-8, GRO-γ, and ENA-78 (Gusella and others 1993; Schnyder-Candrian and others 1995) ( Table 2 ). This differential regulation of angiostatic versus angiogenic CXC chemokines by interferons may, in part, account for their previously documented inhibitory effect on angiogenesis (Angiolillo and others 1996; Coughlin and others 1998; Folkman 1997; Majewski and others 1996; Sgadari and others 1996; Vizier and others 1998; Zetter 1998 ). Stimulus specificity for the expression and production of CXC chemokines LPS, lipopolysaccharide; TNF, tumor necrosis factor; IL, interleukin; IFN, interferon; ENA, epithelial neutrophil activating protein; GRO, growth-related oncogene; IP, inducible protein, MIG, monokine induced by interferon-γ. Stimulus specificity for the expression and production of CXC chemokines LPS, lipopolysaccharide; TNF, tumor necrosis factor; IL, interleukin; IFN, interferon; ENA, epithelial neutrophil activating protein; GRO, growth-related oncogene; IP, inducible protein, MIG, monokine induced by interferon-γ. PF4 (ELR − ) was the first CXC chemokine reported to regulate angiogenesis. PF4 was found to inhibit bFGF-induced angiogenesis and attenuate growth of and in a model of tumorigenesis and others and others 1990, and others In contrast, (ELR + ) was the first CXC chemokine found to induce angiogenesis. was to both in vitro endothelial cell chemotactic and activity as as in vivo angiogenesis in the of using of angiogenesis ( Hu and others 1993 ; Koch and others 1992 ; Strieter and others 1992 ). These have been by other which have that IL-8, to or endothelial cell in vitro and angiogenesis in vivo ( 1996 ; and others 1997 ). These that has a effect on the endothelial cell and that this angiogenic activity is from ability to induce These that members of the CXC chemokine family in a disparate manner and can behave as either potent angiogenic or angiostatic factors in net neovascularization. on this that the highly conserved ELR motif of members of the CXC chemokine family is a structural/functional domain that dictates their angiogenic this endothelial cell was in the presence or of concentrations of ELR + or ELR − CXC chemokines to whether CXC chemokines display disparate angiogenic All of the ELR + CXC chemokines significant endothelial cell chemotactic activity the endothelial cell chemotactic activity of ELR − CXC chemokines was to ( Strieter and others 1995 ). Moreover, ELR − CXC chemokines endothelial cell chemotactic response to ELR + CXC chemokines, bFGF, or This was in vivo using the assay of neovascularization ( Strieter and others 1995 ) ( Table ). that the ELR motif is the critical structural/functional domain that dictates angiogenic activity for members of the CXC chemokine that amino acid residue for the ELR motif of ( Strieter and others 1995 ). In a of MIG was containing the ELR motif to the first cysteine amino acid residue of the primary structure of MIG ( Strieter and others 1995 ). cell and the assay used to the biological activity of these to induce endothelial cell both the endothelial chemotactic activity of and ( Strieter and others 1995 ). In both the angiogenic response of either or or in the assay ( Strieter and others 1995 ). In contrast, the ELR + of MIG induced a significant angiogenic and MIG the angiogenic response of this in the assay ( Strieter and others 1995 ). These a structural/functional role of the ELR motif in the angiogenic or angiostatic potential of CXC chemokines, the that the net biological balance between angiogenic and angiostatic CXC chemokines may an important role in angiogenesis. The ELR + and ELR − CXC chemokines, and their in the micropocket model of angiogenesis or in In the effect of the ELR- CXC chemokines on basic fibroblast growth factor and vascular endothelial growth factor angiogenesis IL, interleukin; ENA, epithelial neutrophil activating protein; GRO, growth-related oncogene; granulocyte chemotactic protein; platelet IP, inducible protein; MIG, monokine induced by interferon-γ. The ELR + and ELR − CXC chemokines, and their in the micropocket model of angiogenesis or in In the effect of the ELR- CXC chemokines on basic fibroblast growth factor and vascular endothelial growth factor angiogenesis IL, interleukin; ENA, epithelial neutrophil activating protein; GRO, growth-related oncogene; granulocyte chemotactic protein; platelet IP, inducible protein; MIG, monokine induced by interferon-γ. The CXC chemokines are important of tumorigenesis to their angiogenic Although has been reported to inhibit angiogenesis ( and others 1995 ), the used in this study was to 10 that found for angiogenic activity to 10 ( Arenberg and others 1997 ; Strieter and others 1995 ). This suggests that concentrations of can the angiogenic Moreover, studies in support that all a significant role in tumorigenesis to both their and angiogenic For example, β, and are all found to be highly in human ( and others 1997 ). determine the biological of the presence of these ELR + CXC chemokines in human β, and into ( and others 1997 ; and others 1997 ). The expression of of these in the cells their to in in vitro and to tumors in vivo in both nude and mice ( and others 1997 ; and others 1997 ). The tumors that highly vascular and to the of ( and others 1997 ; and others 1997 ). of with in a marked reduction of tumor-derived angiogenesis and inhibition of tumor growth ( and others 1997 ; and others 1997 ). These support the notion that the ELR + CXC chemokines, such as β, and have the ability to as potent angiogenic factors to promote tumorigenesis in The and growth of are also on angiogenesis, and has been to a significant role in human angiogenesis and tumorigenesis in nude mice ( and others 1998 ). and the expression of IL-8, bFGF, and VEGF in different human cell All cell lines of in these cells either or of or tumors in a rate that was to marked neovascularization ( and others 1998 ). The expression of was directly with neovascularization and with VEGF expression was only with production of ( and others 1998 ). was found for the expression of with either tumor neovascularization or ( and others 1998 ). These studies have also been in and in which the expression of with and of these tumors ( and others 1998 ; and others 1994 ). has been found in of cell cancer 1 ) and others and has been to be a significant angiogenic factor to tumor-derived angiogenic activity ( Arenberg and others ; and others these studies to an in vivo model system of human tumorigenesis human mouse ( Arenberg and others ), tumor-derived was found to be directly with tumorigenesis ( Arenberg and others ). with to a reduction in tumor which was by a reduction in spontaneous to the ( Arenberg and others ). The of tumor growth and was directly with reduced tumor-derived angiogenesis. These have been using other cell and that cell lines that have potential in nude and this is directly with their angiogenic Although may an important angiogenic CXC ENA-78 may be a more important angiogenic CXC chemokine in ( Arenberg and others 1998 ). of tumors found to display a and significant of ENA-78 protein with tumor neovascularization. The biological of this was in a mouse model of human tumorigenesis using human cell ENA-78 expression in tumors was directly with tumor Moreover, human with both tumor growth and spontaneous The reduction in tumor growth was by a marked in tumor and an increase in of the tumor The of these cells was not to a effect of of ENA-78 was in vitro evidence that ENA-78 effect on This is with the that angiostatic of tumors is with tumor cell ( O’Reilly and others 1996 , 1997 ). in vivo and in vitro of cells was by the presence of to Although was a significant of ENA-78 expression with tumor and of ENA-78 not inhibit tumor This the that angiogenic activity induced by tumors is to or factors in a In to ENA-78 and IL-8, are other angiogenic factors that have not been other ELR + CXC chemokines or and/or chemokine angiogenic factors (bFGF or the that CXC chemokines significant angiogenic factors in human tumors that strategies angiogenic CXC angiogenesis may be a novel to The of ELR + is the of ELR − CXC chemokines in angiogenesis. cell lines tumors in nude mice ( and others 1988 ). determine whether IP-10 or MIG the angiogenic activity of Sgadari and of these tumor cell lines in nude The expression of IP-10 and MIG was in the tumors that spontaneous compared with tumors that in nude The of the tumors was directly to evidence of reduced angiogenesis, tumor vascular and tissue determine whether this effect was to IP-10 or MIG, cell lines in nude mice to with IP-10 or in marked reduction in angiogenesis ( Sgadari and others 1997 ; and others 1997 ). Although both IP-10 and MIG can induce T-cell the with their CXC chemokine ( Baggiolini 1998 ; Baggiolini and others 1997 ; 1997 ; Luster 1998 ; Rollins 1997 ), the ability of both of these ELR- CXC chemokines to inhibit angiogenesis and induce in nude mice supports the that these chemokines their in a laboratory has the role of IP-10 in angiogenesis with ( Arenberg and others ). The of IP-10 from human tumor in normal The increase in IP-10 from human tissue was to the of IP-10 in cell 1 ) compared with Moreover, of IP-10 from in augmented angiogenic activity ( Arenberg and others ). The marked difference in the and of IP-10 with compared with is and a possible mechanism for the biological of these two of with of the is potential is and evidence of angiogenesis is for ( 1988 ; ; and others 1995 ). The studies described to a mouse model system to the effect of IP-10 on human cell tumor growth in a and mice with either or cells and others The production of IP-10 from and tumors was with tumor growth ( Arenberg and others IP-10 in the compared with The of spontaneous in mice tumors IP-10 from either the primary tumor or a In mice tumors with either to tumors with IP-10 to of IP-10 in tumors in a increase in their 10 In contrast, of IP-10 in tumors reduced both their and which was directly to a reduction in angiogenesis ( Figure 2 ). These support the notion that tumor-derived IP-10 is an important angiostatic factor in and strategies to IP-10 and/or other angiostatic CXC chemokine may lead to marked tumor of inducible protein other for compared with of human tumor growth in severe combined immunodeficiency of a mouse tumors with of a mouse tumors with of inducible protein other for compared with of human tumor growth in severe combined immunodeficiency of a mouse tumors with of a mouse tumors with Angiogenesis is by an balance of angiogenic and angiostatic The studies described using immunodeficient mice as animal of human tumorigenesis have that CXC chemokines to be important factors that regulate angiogenesis in association with tumorigenesis in a variety of These support the notion that either inhibition of angiogenic or augmentation of angiostatic CXC chemokine biology may be a novel in the of human the ability to study these molecules in immunodeficient mice has to the biology of human CXC chemokines in the regulation of angiogenesis in a and manner in the of human tumorigenesis. This was in part, by of and as as by an