HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Carlos, T. M. Search for Related Content PubMed PubMed Citation Articles by Carlos, T. M.

(Journal of Leukocyte Biology. 2001;70:171-184.)
© 2001 by Society for Leukocyte Biology

Leukocyte recruitment at sites of tumor: dissonant orchestration

Division of Hematology/Oncology, Department of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania

Correspondence: T. M. Carlos, MD, Division of Hematology/Oncology, University of Pittsburgh, 200 Lothrop Street, Pittsburgh, PA 15213-2582. E-mail: Timothy.Carlos{at}med.va.gov

	ABSTRACT

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 
Biopsies of tumors responding to interleukin 2 (IL-2) based immunotherapy have been reported to show a leukocytic infiltration. Clinical responses to IL-2-based immunotherapy, however, are limited, suggesting a failure of leukocyte localization at tumor sites. Leukocyte infiltration at inflammatory sites requires local activation of leukocytes and endothelial cells in a coordinated and defined temporal sequence. There is evidence supporting the theory that infiltration of leukocytes at tumor sites is suboptimal due to a failure of coordination of these localizing events. In this review, factors involved in leukocyte recruitment at sites of inflammation and the coordination of these factors in a successful model of inflammation, i.e., wound healing, are discussed. This example is contrasted with events at tumor sites where alterations in expression of cell adhesion molecules or in the production of activating agents may be present. Additionally, the systemic administration of an activating cytokine such as IL-2 may fail to duplicate events that normally occur within a local environment. These observations may facilitate the design of future immunotherapy trials.

Key Words: adhesion • immunotherapy • cytokine • chemokine

	INTRODUCTION

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 
A minority of cancer patients benefit from systemic interleukin (IL)-2-based therapy [^{1 2 3 4} ]. Several reports have shown that lesions of metastatic melanoma that regress after IL-2-based immunotherapy become infiltrated by mononuclear leukocytes [^{5 6 7} ]. In each of these studies, lesions biopsied before immunotherapy lacked an infiltration of mononuclear leukocytes, or these inflammatory cells were confined to the periphery or within fibrovascular septa of tumors. In responding lesions biopsied after IL-2 therapy, however, a marked increase in the number of mononuclear leukocytes was noted. In contrast, nonresponding lesions remained devoid of an inflammatory infiltrate.

For immunotherapy to be an effective form of treatment, endogenously activated or adoptively transferred leukocyte effector cells [i.e., cytotoxic T lymphocytes (CTLs) or natural killer (NK) cells] must localize at sites of tumor. Localization of inflammatory cells at these sites is dependent on a series of regional events collectively referred to as an "adhesion cascade" [⁸ , ⁹ ]. This cascade requires that agonists generated within the tumor microenvironment induce the expression of endothelial adhesion molecules and activate counter-receptors for these structures on leukocytes. These events, if successful, lead to firm adherence of effector leukocytes to microvascular endothelial cells within the tumor and diapedesis across this barrier to reach malignant cells.

This review addresses factors that may limit leukocyte recruitment at sites of tumor and, therefore, the success of immunotherapy. The first of these is the cell adhesion molecules involved in recruitment of leukocytes at sites of inflammation. Their roles during an example of successful leukocyte recruitment, i.e., wound healing, are discussed. Finally, evidence is provided suggesting that tumor-derived products adversely affect the adhesion cascade and subsequent recruitment of leukocytes at sites of tumors.

Several recent reviews have discussed the possible involvement of adhesion molecules expressed on tumor cells or endothelium during the process of metastasis [¹⁰ , ¹¹ ]. This topic is not discussed in this review. Similarly, although this review deals with the mechanisms of leukocyte recruitment at tumor sites, there is evidence suggesting that tumor-derived products may adversely affect the function of these cells after recruitment. The reader is referred to recent reviews on this topic for further information [¹² , ¹³ ].

The use of immunotherapy as a form of treatment for malignancies has met with limited success. It is our hope that this review contributes an explanation as to why immunotherapy sometimes fails and that it also stimulates strategies to improve therapy.

	CELL ADHESION MOLECULES INVOLVED DURING LEUKOCYTE RECRUITMENT

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 
The recruitment of leukocytes at inflammatory sites requires two main elements: cell adhesion molecules expressed on leukocytes or endothelial cells and agonists (e.g., cytokines, chemokines, or other substances) produced at the local site. These agonists induce the expression or affect the function of the cell adhesion molecules. After these stimuli, leukocytes circulating in the blood begin to roll along and then become firmly adherent to the vascular endothelium. Subsequently, leukocytes cross the endothelial barrier and migrate towards inflammatory signals produced at these sites [⁸ , ⁹ ]. The coordination of these adhesion molecules and the agonists involved in their regulation constitutes the orchestration of inflammation and contains several layers of complexity.

The first layer of complexity involves the expression of adhesion molecules on all or only subsets of leukocytes (Table 1 ). Although one member of the ß₂ family of leukocyte integrins, i.e., lymphocyte function associated antigen-1 (LFA-1, ß₂ or CD11a/CD18), is expressed on all leukocytes, other members, i.e., Mac-1 (_mß₂ or CD11b/CD18), p150/95 (_xß₂ or CD11c/CD18) and _dß₂ (CD11d/CD18), are primarily expressed on myeloid cells and subsets of mononuclear leukocytes [¹⁴ , ¹⁵ ]. Very-late-activation antigen 4 (VLA-4, ₄ß₁ or CD49d/CD29), a fifth integrin involved in leukocyte adhesion, is expressed on the majority of leukocytes [¹⁶ ]; nonetheless, it is not expressed on human neutrophils [¹⁷ ]. A sixth leukocyte integrin that functions in cell adhesion, ₄ß₇, is expressed on resting B lymphocytes and subsets of memory T cells but not on other leukocytes [¹⁸ , ¹⁹ ].

Finally, other leukocyte cell adhesion molecules that play a role in recruitment at sites of inflammation include platelet-endothelial adhesion molecule-1 (PECAM-1) and CD73. PECAM-1 (CD31), a member of the immunoglobulin gene superfamily, is expressed on neutrophils, monocytes, and subsets of naive T lymphocytes (e.g., CD45RA⁺, CD4⁺, and CD8⁺ cells) [^{29 30 31} ]. Lymphocyte-vascular adhesion protein (VAP)-2 (CD73) is an ectonucleotidase expressed predominantly on B lymphocytes and CD8⁺ T cells [³² , ³³ ]. Thus, among leukocyte subsets there is a moderate degree of diversity in the expression of adhesion molecules or combinations of adhesion molecules that play a role in the localization of leukocyte subsets during an inflammatory or immune response.

A second element of complexity in the orchestration of leukocyte migration involves the regulation of adhesion molecules on leukocytes after local activation at sites of inflammation. Locally generated agonists (e.g., chemokines) that bind to receptors on leukocytes induce a change in avidity of leukocyte integrins for their endothelial counter-receptors [³⁴ ]. Leukocyte activation by local agonists results in "inside-out" signaling and an increase in binding of LFA-1 to its endothelial counter-receptors intercellular adhesion molecule (ICAM)-1 (CD54) and ICAM-2 (CD102) [³⁵ ]. Since LFA-1 is expressed on all leukocytes, this results in nonselective leukocyte recruitment across endothelium. Inside out signaling also induces binding of Mac-1 and VLA-4 to their respective endothelial ligands [ICAM-1 and vascular cell adhesion molecule (VCAM)-1 (CD106)] [³⁶ , ³⁷ ]. In this situation, however, subsets of leukocytes would be recruited since these leukocyte adhesion molecules are not expressed on all white blood cells. A second method of regulation of leukocyte adhesion molecules involves the loss of L-selectin by proteolytic cleavage after activation [^{38 39 40} ]. Local loss of L-selectin may be important for subsequent transendothelial migration. Finally, redistribution of adhesion molecules occurs after activation that may facilitate the function of adhesion molecules (e.g., PSGL-1 moves to microvilli for optimal rolling and tethering [⁴¹ ]). Alterations in expression or function of leukocyte adhesion molecules, therefore, represent a second level of complexity during recruitment of leukocytes at sites of inflammation.

A third element in the complex orchestration of leukocyte recruitment involves the expression of chemokine receptors on subsets of leukocytes (Table 1) [^{42 43 44} ]. Local production of ELR⁺ CXC chemokines (e.g., IL-8) would be expected to activate counter-receptors expressed on neutrophils (CXCR1 and CXCR2), resulting in the recruitment of neutrophils. Conversely, regional production of ELR^- CXC chemokines [e.g., monokines induced by interferon (IFN)- (Migs) and IFN--inducible protein-10 (IP-10)] and binding to the leukocyte receptor for these chemokines (CXCR3) would result in recruitment of lymphocytes. In a similar manner, regional production of CC chemokines [e.g., monocyte chemoattractant protein (MCP)-1 and macrophage inflammatory protein (MIP)-1] would lead to the activation of mononuclear leukocyte binding after interaction of these chemokines with their respective chemokines receptors [CCR2 (MCP-1) and CCR1 or CCR5 (MIP-1)]. Thus, in addition to the determination of selective recruitment of leukocyte subsets by expression of different combinations of adhesion molecules, the expression of varied chemokine receptors on subsets of leukocytes and the subsequent activation of leukocyte integrins, loss of L-selectin, and redistribution of PSGL-1 also contribute to the selective recruitment of leukocyte subsets.

A fourth component in the orchestration of leukocyte recruitment involves the local expression of endothelial adhesion molecules at sites of inflammation. Although several adhesion molecules are constitutively expressed on vascular endothelium (e.g., ICAM-1 [⁴⁵ ], ICAM-2 [⁴⁶ , ⁴⁷ ], and PECAM-1 [⁴⁸ ]), other adhesion molecules are expressed only after endothelial activation (Table 2 ). P-selectin (CD62P) is synthesized and stored within Weibel-Palade bodies [⁴⁹ ]. After activation, these granules fuse with the cell membrane resulting in surface expression of P-selectin. In contrast to P-selectin, both endothelial (E)-selectin (CD62E) and VCAM-1 require de novo protein synthesis after endothelial activation [⁵⁰ , ⁵¹ ]. Vascular adhesion protein-1 (VAP-1), an endothelial adhesion molecule involved in lymphocyte re-circulation, is not constitutively expressed on endothelial cells in vitro and is not induced by proinflammatory cytokines [⁵² ]. However, VAP-1 is expressed at sites of inflammation in vivo and selectively binds CD8⁺ T cells and NK cells in vitro [⁵³ , ⁵⁴ ].

Successful orchestration of all these events results in recruitment of leukocyte subsets across an endothelial barrier. In this cascade, circulating leukocytes roll along vascular endothelium and become tethered at this local site by adhesive interactions involving selectin ligands (e.g., L-selectin and PSGL-1) and VLA-4 [⁷³ , ⁷⁴ ]. Local production of activators of specific leukocytes (e.g., CXC or CC chemokines) causes changes in avidity of leukocyte integrins (LFA-1, Mac-1, and VLA-4) for their respective ligands (ICAM-1 for both LFA-1 and Mac-1, ICAM-2 also for LFA-1, and VCAM-1 for VLA-4) constitutively expressed or induced on endothelium. This process results in firm adhesion to the endothelium at the local site followed by leukocyte diapedesis.

The importance of adhesion molecules in successful recruitment of leukocytes during an inflammatory response can be demonstrated by observing the clinical phenotype of inherited defects of these molecules in humans or by examining the effect on leukocyte recruitment in mice lacking expression of these proteins. Defects in leukocyte recruitment have been clearly demonstrated in patients lacking the ß₂ family of leukocyte integrins [patients with Leukocyte Adhesion Deficiency (LAD)-I] [⁷⁵ ] and in patients without the ability to generate selectin ligands (patients with LAD II [⁷⁶ ]) due to abnormalities in fucosyl transferase activity. In both groups of patients, leukocyte migration at sites of inflammation is affected. In both situations, however, defects in neutrophil migration predominate since the interaction of VLA-4 with VCAM-1 is a redundant mechanism for recruitment of mononuclear leukocytes, eosinophils, and basophils. Mice genetically engineered to lack the expression of the entire family of ß₂ integrins (CD18) [⁷⁷ ] or the subunits of these integrins [⁷⁸ , ⁷⁹ ] show defective leukocyte migration at sites of inflammation. Similarly, mice lacking L-selectin [⁸⁰ ] or the fucosylated ligands for endothelial selectins [²⁵ ] also show abnormal recruitment in models of acute inflammation.

Defects in leukocyte recruitment due to abnormalities in expression of endothelial adhesion molecules also have been reported. A single patient with a genetic abnormality in endothelial adhesion molecule expression has been described [⁸¹ ]. This patient generates a soluble form of E-selectin but lacks a membrane-bound protein. This deficiency results in a clinical syndrome of recurrent infections with a lack of neutrophil recruitment at sites of inflammation. Mice genetically engineered to lack expression of single [⁸² , ⁸³ ] or dual [⁸⁴ ] endothelial selectins, ICAMs-1 and -2, and PECAM-1 have been reported [^{85 86 87} ]. Mice deficient in P-selectin show clear evidence of defective leukocyte recruitment during inflammation [⁸³ ]. Alterations in leukocyte recruitment during inflammation in mice lacking E-selectin, however, are not detected unless P-selectin is blocked by administration of a monoclonal antibody [⁸² ] or by concomitant deletion of P-selectin [⁸⁴ ]. Mice deficient in expression of ICAM-1 [⁸⁵ ], ICAM-2 [⁸⁶ ], or PECAM-1 [⁸⁷ ] have impaired leukocyte recruitment in models of inflammation. The role of VCAM-1 during recruitment of leukocytes in models of inflammation can not be examined in mice lacking VCAM-1 because induced deletions of this molecule result in impaired embryogenesis [⁸⁸ ].

Defective expression or regulation of adhesion molecules, therefore, results in impaired recruitment of leukocytes at sites of inflammation. These defects may be induced by failure to generate the agonists that affect the expression or function of these molecules at these sites. Conversely, defective expression or function of adhesion molecules required for optimal localization of leukocytes may be suppressed by overproduction of antagonists that impair their function.

	KINETICS OF LEUKOCYTE RECRUITMENT DURING INFLAMMATION

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 
After successful orchestration of these events, recruitment of leukocytes at sites of inflammation follows a defined temporal sequence. Early studies in rats examining leukocytic infiltration after intradermal injections of fibrinogen or fibrin demonstrated that neutrophils were the initial subset of leukocytes recruited, followed by a wave of mononuclear cells [⁸⁹ ]. In a rabbit model of acute inflammation, peak migration of radiolabeled neutrophils after intradermal injections of lipopolysaccharide (LPS), zymosan-activated serum, or N-formyl methionyl leucyl phenylalanine occurred between 2 and 3 h after injection with little additional recruitment after 6 h [⁹⁰ ]. In a study of inflammation in a rat model of adjuvant arthritis, neutrophils infiltrated affected joints 4–6 days prior to recruitment of lymphocytes [⁹¹ ]. Finally, in a study using nonhuman primates to contrast the sequence of leukocyte infiltration after intradermal injections of either LPS or mammalian tuberculin, peak neutrophil influx occurred within the first 24 h. In contrast, lymphocyte infiltration in the delayed-hypersensitivity model was maximal 4 days after injection [⁹² ]. Thus, neutrophil infiltration at sites of inflammation peaks within 1 day and is followed by a delayed accumulation of lymphocytes.

In models examining the recruitment of monocytes during acute inflammation, the initial entry of this subset of leukocytes was reported to be concurrent with neutrophil influx after intradermal injection of LPS. However, recruitment of monocytes persisted for several additional hours; hence, by 12 h postinjection, monocytes were the predominant leukocytes present [⁹³ ]. Sustained monocyte recruitment during acute inflammation beyond that of neutrophils was also reported in rats after intradermal injection of inflammatory cytokines such as TNF- [⁹⁴ ]. The kinetics of monocyte infiltration into other sites of acute inflammation (e.g., lungs) has been examined [⁹⁵ ]. After the injection of complement into a localized lung segment, neutrophil recruitment begins within 1 h and becomes maximal within 4 h after injection. Monocyte recruitment is delayed by 3–4 h but persists for 48 h. As observed in the models of intradermal injections of inflammatory agents, delayed and sustained recruitment of monocytes within lungs is noted. It is interesting that monocyte recruitment is impaired in rabbits made neutropenic by administration of mechlorethamine. This alteration is reversed by the reinfusion of neutrophils, suggesting a coordination of leukocyte subsets in the kinetics of recruitment.

A possible explanation for the observed kinetics and coordination of leukocyte recruitment in vivo is the induction of cytokines or chemokines within the local milieu by one type of leukocyte that facilitates the subsequent recruitment of other subsets of leukocytes. Initial influx of neutrophils within the lungs after hemorrhage or endotoxemia has been shown to be a ready source of IL-1 that would induce activation-dependent endothelial adhesion molecules [⁹⁶ ]. Evidence for this possibility also has been reported in vitro. Monocyte adhesion to endothelial cells has been shown to induce the production of TNF- [^{97 98} ]. Local generation of TNF- by the initial wave on monocytes induces the production of adhesion molecules on local endothelial cells that participate in recruitment of mononuclear leukocytes but not neutrophils (e.g., VCAM-1). Similarly, the binding of monocytes to endothelium in vitro induces the synthesis and release of chemokines (e.g., IL-8 [⁹⁹ ] and MCP-1 [¹⁰⁰ ]).

Successful recruitment of leukocytes at sites of inflammation, therefore, is a carefully orchestrated sequence of regional events. Because subsets of leukocytes are recruited in a temporal sequence, variability in the combinations of adhesion molecules present on the leukocytes, the endothelial adhesion molecules expressed, and factors produced within the local microenvironment that activate or regulate these molecules affect this process.

	SUCCESSFUL RECRUITMENT OF LEUKOCYTES DURING WOUND HEALING

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 
Events observed during normal wound healing demonstrate the successful orchestration of cell adhesion molecule expression, production of activating agonists within the local environment, and the kinetics of leukocyte recruitment. Induction of inflammation is the first phase of wound repair [¹⁰¹ ]. During this phase of wound healing, neutrophils and monocytes are the initial leukocyte subsets recruited. Murine models of wound healing have demonstrated that IL-1ß and TNF- are induced locally within 1 day of injury but become undetectable by day 13 [¹⁰² ]. The induction of E-selectin and P-selectin by these cytokines has been shown to be critically important in the initial phase of normal wound healing [¹⁰³ ]. Mice genetically deficient in E-selectin do not have a demonstrable defect in neutrophil recruitment within the initial 4 h after injury. Mice lacking P-selectin, in contrast, have an initial impairment in neutrophil recruitment by 1 h, but not by 4 h, after wounding. Induction of E-selectin within 4 h of wounding compensates for the absence of P-selectin. Mice genetically engineered to lack both endothelial selectins, however, have a marked impairment in neutrophil infiltration 4 h after injury. The latter group of mice also have a threefold reduction in the number of macrophages infiltrating the healing wound as compared with wild-type mice. Thus, both selectins play a key role in the initial recruitment of inflammatory cells during normal wound healing.

Murine models of wound healing have also delineated some of the chemokines that are important during the initial inflammatory phase of wound healing. MIP-2 (the murine homologue of human IL-8), MIP-1, and MCP-1 have been shown to be induced rapidly after injury [¹⁰² , ¹⁰⁴ , ¹⁰⁵ ]. The administration of neutralizing antibodies directed against these agonists diminishes monocyte infiltration. Expression of these agonists persists for several days after injury but returns to baseline within 7 days. During impaired wound healing in diabetic mice, however, expression of MIP-2 and MCP-1 as well as persistent induction of TNF- and IL-1ß has been reported [¹⁰² ]. Thus, normal wound healing is characterized by a transient induction of proinflammatory cytokines and leukocyte-specific chemokines followed by resolution of these events.

In human wound healing, a similar pattern of chemokine expression has been reported [¹⁰⁶ ]. Infiltration of neutrophils peaks within 24 h of injury in normal volunteers. This influx occurs concurrently with the induction of IL-8 and GRO, CXC chemokines that bind to CXCR1 and CXCR2 expressed on neutrophils. Monocyte recruitment peaks within 48 h and occurs with induction of MCP-1 but not MIP-1 or RANTES (regulated) on activation, normal T cell expressed and secreted. Influx of lymphocytes after injury is relatively constant and has been observed during induction of IP-10 and Mig [¹⁰⁶ ]. Recruitment of leukocyte subsets during the sequence of normal wound healing, therefore, follows the pattern of chemokine expression.

In contrast to proinflammatory cytokines, local production of IL-10 has an inhibitory role in leukocyte recruitment and chemokine generation during wound healing. IL-10 has been shown to decrease production of early cytokines (e.g., TNF- and IL-1) involved in normal wound healing and to diminish generation of chemokines (MCP-1 and MIP-1) [¹⁰⁷ ]. Administration of neutralizing antibodies against IL-10 leads to increased production of these agonists and augmented recruitment of neutrophils and monocytes.

Successful recruitment of leukocytes during wound healing reflects a balance between agonists (e.g., proinflammatory cytokines and leukocyte-specific chemokines) and antagonists (e.g., anti-inflammatory cytokines) within the local environment. In addition, the kinetics of leukocyte recruitment and of agonist expression suggests that this process is finely regulated. An imbalance in these factors may contribute to the characterization of tumors as "wounds that don’t heal" [¹⁰⁸ ].

	FAILURE OF ORCHESTRATION DURING IMMUNOTHERAPY OF TUMORS

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 
Leukocytes commonly fail to infiltrate tumors but remain localized within the periphery or fibrous septa of the tumor [¹⁰⁸ ]. Defective leukocyte recruitment as normally observed in wound repair might be one explanation for this abnormality. The lack of orchestrated recruitment of leukocytes might also contribute to a poor response to immunotherapy. Although the concept of the adhesion cascade was developed from observations made in models of acute inflammation, there is evidence suggesting that the cascade could play a role in recruitment of leukocytes at sites of tumor, which might be germane not only for localization of endogenous leukocytes but also during the recruitment of immune effector cells at tumor sites after adoptive transfer.

In vivo microscopy studies
In animal models examining the rolling and adhesion of normal leukocytes on vascular endothelium within tumors in vivo, alterations in the adhesion cascade have been observed. In a murine model of mammary adenocarcinoma, rolling and adherent leukocytes were observed in blood vessels within normal tissue but not within tumors [¹⁰⁹ ]. Addition of a local chemoattractant (i.e., formyl methionyl leucyl phenylalanine) increased the number of rolling leukocytes on endothelial cells within normal tissue only. Pretreatment of the animals with LPS or TNF- to induced expression of activation-dependent endothelial adhesion molecules leading to increased leukocyte adhesion in both normal and tumor tissue, but firm adhesion, a phenomenon mediated by the interaction of leukocyte integrins and their endothelial counter-receptors, remained impaired within the latter. Since normal leukocytes were used in this study, a primary alteration in the endothelial component of the adhesion cascade was proposed [¹⁰⁹ ]. Defective leukocyte rolling on tumor vascular endothelium in vivo has also been reported using both the murine mammary adenocarcinoma model and a human glioblastoma xenograft model [¹¹⁰ ]. Borgstrom et al. evaluated the role of endothelial selectins and leukocyte ß₂ integrins during leukocyte-tumor endothelial-cell interactions [¹¹¹ ]. Defective rolling was again noted on unactivated tumor vascular endothelium in vivo but was increased on tumor endothelium after the addition of TNF-ß/lymphotoxin. Leukocyte rolling was blocked by pretreatment with monoclonal antibodies that inhibited the adhesive function of E-selectin or P-selectin. Local addition of a chemoattractant (i.e., LTB4) was followed by increasingly firm adhesion and transendothelial migration into tumor stroma. Adhesion and migration of leukocytes could be ablated by pretreatment with a blocking monoclonal antibody against CD18 [¹¹¹ ]. These reports demonstrate a primary defect in the initial steps involved in the adhesion cascade, steps mediated by the interaction of leukocyte selectin ligands (e.g., L-selectin, PSGL-1, CLA, SLeX, or other fucosylated molecules) with endothelial selectins (e.g., E-selectin or P-selectin) or with leukocyte VLA-4 interacting with endothelial VCAM-1. These studies also demonstrate a role of leukocyte integrins in firm adhesion and transendothelial-cell migration at sites of tumor.

Cell adhesion molecules involved in localization of leukocytes in preclinical tumor models
In studies examining the involvement of cell adhesion molecules during localization of leukocytes at sites of tumors, roles for several adhesion molecules have been demonstrated. In a study examining the recruitment of NK cells to subcutaneous B16 melanoma tumors after systemic administration of polyinosinic-polycytidylic and poly-L-lysine, pretreatment with a blocking antibody directed against VCAM-1 abrogated an increase in tumor-associated NK cells whereas inhibition of the adhesive function of ICAM-1 had no effect [¹¹² ]. In contrast, roles for both VLA-4 and LFA-1 in localization of T cells at sites of murine fibrosarcoma and ovarian cancers were shown by Ogawa et al. [¹¹³ ]. Although leukocytes are rarely observed within these tumors prior to IL-12 immunotherapy, massive infiltrations of mononuclear cells were seen in responding tumors. Pretreatment of mice with blocking antibodies directed against LFA-1 and VLA-4 inhibited leukocyte migration into tumors and ablated the therapeutic effect. Finally, a clear role for LFA-1 during localization and eradication of tumors has been reported in mice lacking this adhesion molecule [¹¹⁴ ]. Cytotoxic T cells generated from mice lacking LFA-1 were unable to localize within or eliminate an immunogenic fibrosarcoma tumor.

Although there is some evidence suggesting roles for VLA-4 and LFA-1 in localization of effector leukocytes at tumor sites, there is increasing evidence that expression of L-selectin is not critical for recruitment [¹¹⁵ , ¹¹⁶ ]. Plautz et al. [¹¹⁵ ] reported that sensitized T cells obtained from lymph nodes draining subcutaneous MCA205 fibrosarcoma tumors could be sorted into populations either expressing or lacking L-selectin. T cells that were localized within tumors after adoptive transfer were predominantly L-selectin^low. Whereas the L-selectin^low population represented only 11% of sensitized T cells obtained from the draining lymph node, reinfusion of these cells was highly effective in localization at and eradication of intracranial tumors [¹¹⁵ ]. In a second study from the same investigators, >95% of T cells localized at sites of experimental pulmonary or subcutaneous metastases of fibrosarcoma were L-selectin^neg and had higher levels of expression of other adhesion molecules (i.e., LFA-1 and ₄ß₇) [¹¹⁶ ]. These investigators also reported that the populations of L-selectin^neg T cells express higher levels of some cytokines (i.e., IFN-, IL-2, IL-4, and IL-10) than unfractionated populations of T cells [¹¹⁷ ]. Thus, the lack of L-selectin did not adversely affect localization of sensitized T cells at sites of intracranial, subcutaneous, or pulmonary tumors. Lack of L-selectin of sensitized T cells, however, might mark a population of effector cells with alternate adhesion mechanisms that are involved in recruitment of effector cells at these sites.

Although the inhibition of leukocyte or endothelial adhesion molecule function impairs leukocyte recruitment across an endothelial barrier, it is important to consider other functions of these molecules. In a study infusing T cells obtained from lymph nodes draining subcutaneous MCA205 fibrosarcoma tumors, shortened survival in mice bearing intracranial or pulmonary tumors was observed after the administration of an antibody that inhibited the adhesive function of LFA-1 [¹¹⁸ ]. Antibodies directed against ICAM-1, VLA-4, or VCAM-1 were without effect. Surprisingly, no inhibition of T-cell migration into brain tumors was observed after the administration of LFA-1. Thus, prevention of leukocyte emigration was not the reason for failure of efficacy in this model. Rather, an inhibition of LFA-1-dependent T-cell interaction with cells expressing ligands for this adhesion molecule (e.g., ICAM-1, ICAM-2, or ICAM-3 [¹¹⁹ ] expressed on macrophages or tumors) within the region is the suggested mechanism for the observed impaired survival.

From the studies cited, roles for leukocyte LFA-1 and VLA-4 interacting with their respective endothelial ligands (i.e., ICAMs 1–3 and VCAM-1) appear to be germane for recruitment of effector leukocytes to sites of experimental tumors. Conversely, leukocyte expression of L-selectin does not appear to be important but marks a subset of effector cells with augmented potential to localize at sites of tumor.

Altered expression of endothelial adhesion molecules on human tumor endothelium
From the preclinical animal studies cited above, endothelial selectins involved in leukocyte rolling (i.e., E-selectin and P-selectin) and endothelial molecules involved in firm adhesion (i.e., ICAM-1, ICAM-2, and VCAM-1) appear to be important during leukocyte recruitment at sites of tumor. Thus, alterations in these molecules would be expected to impair localization of leukocytes.

In the past 7 years, numerous studies have reported the expression of adhesion molecules on endothelial cells within human tumors [^{120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137} ]. Some of these studies [¹²² , ¹³² ] have assessed the level of expression of these molecules by reporting the values as fractions of endothelial cells within an area that is delineated by PECAM-1, a pan-endothelial marker. Other studies have used a qualitative scale based on the intensity of staining. In addition, a minor portion of these studies addresses the level of expression of these molecules at sites of metastatic disease [¹²⁴ , ¹²⁶ , ¹²⁹ , ¹³² , ¹³⁴ ], whereas the majority address expression at sites of primary tumors. Finally, only two studies report changes in expression of these molecules in human tumors after chemoimmunotherapy [¹²⁴ , ¹³² ].

To examine the level of expression of a panel of endothelial adhesion molecules on sites of metastatic melanoma, a kind of tumor presently being treated with IL-2-based immunotherapy, the numbers of vessels within biopsied tumors from 20 patients were quantified using CD31 staining to define these structures. The numbers of vessels expressing ICAM-1, ICAM-2, VCAM-1, E-selectin, and P-selectin were then expressed as percentages of the number of CD31+ vessels (Table 3 ), [¹³⁷ ]. This analysis found that expression of PECAM-1 is common within metastatic melanoma tumors. A second constitutively expressed endothelial adhesion molecule, ICAM-2, was present on 50% of the vessels within the biopsied specimens. In contrast, the presence of adhesion molecules whose expression is modulated by cytokines was limited [ICAM-1, 10 (20%); VCAM-1, 1 (4%); E-selectin, 1 (3%); and P-selectin, 10 (15%)]. Three other studies have examined the expression of adhesion molecules on vascular endothelium within melanoma tumors. Schadendorf et al. reported a high degree of E-selectin and P-selectin expression at sites of primary and metastatic tumors, principally localized within regions of neovascularization [¹²⁶ ]. Renard et al. reported a high level of PECAM-1 and ICAM-1 within melanoma tumors in transit but low levels of expression of VCAM-1 and E-selectin [¹²⁴ ]. Finally, Nooijen et al. reported that the majority of vessels within primary tumors express P-selectin, but the level of expression at metastatic sites is diminished [¹³⁴ ]. These reports suggest that expression of activation-dependent endothelial adhesion molecules (i.e., VCAM-1, E-selectin, and P-selectin) is limited at sites of metastatic melanoma tumors with expression often seen at peripheral areas of tumors associated with neovascularization.

Two studies have reported changes in expression of VCAM-1 and E-selectin noted after isolated limb perfusion for melanoma in transit or sarcoma [¹²⁴ , ¹³² ]. At baseline, high levels of PECAM-1 and ICAM-1 were present within tumor vascular endothelium with weak or absent expression of E-selectin or VCAM-1. After infusion of TNF-, E-selectin was strongly induced on tumor vascular endothelium while no change was noted in expression of VCAM-1 [¹²⁴ , ¹³² ].

In summary, constitutive expression of PECAM-1 and ICAM-2 is often found within primary and metastatic sites of malignancies. Expression of endothelial adhesion molecules that are affected by cytokines or other agents (e.g., NO), in contrast, is frequently absent within tumors and is often limited to areas of neovascularization in the periphery of the tumor. Because the latter area is often the location of leukocytes attempting to infiltrate the tumor [¹⁰⁸ ], lack of expression of activation-dependent endothelial adhesion molecules within the tumor may be a mechanism by which tumors escape immune surveillance.

Tumor-derived products suppress expression of endothelial adhesion molecules
An increasing number of reports have demonstrated that tumor-derived products adversely affect the expression of endothelial adhesion molecules. Inhibition of VCAM-1 up-regulation on pulmonary venules near B16 melanoma tumors after systemic injection of LPS has been reported [¹³⁸ ]. Although 95% of pulmonary vessels expressed VCAM-1 after LPS injection in nontumor regions, only 18% of vessels near melanoma tumors were VCAM-1 positive. Coincubation of B16 tumor cells with a murine endothelioma cell line confirmed that up-regulation of VCAM-1 after addition of LPS, IL-1, or TNF- is prevented by the presence of the tumor cells. Although the factor(s) responsible for this inhibition was not characterized in this study, several possible explanations can be offered.

Angiogenic factors (i.e., bFGF, VEGF, and TGF-ß) have been reported to adversely affect the expression of endothelial adhesion molecules [¹³⁹ , ¹⁴⁰ ]. Endothelial cells isolated from fresh tumor tissue have been shown to have diminished expression of ICAM-1 and ICAM-2 [¹³⁹ ]. The addition of bFGF to endothelial cells in vitro initially stimulates and then suppresses the expression of these molecules, whereas PECAM-1 is unaffected. Induction of E-selectin and VCAM-1 and up-regulation of ICAM-1 by TNF- are prevented by pretreatment with bFGF. Additional work by these investigators demonstrated that bFGF also inhibits up-regulation of ICAM-1 by IL-1 and IFN- and that both VEGF and TGF-ß but not IL-8 prevent cytokine induction of E-selectin and VCAM-1 [¹⁴⁰ ]. Thus, angiogenic factors produced by tumor might suppress the expression of endothelial adhesion molecules that are requisite for optimal leukocyte recruitment, and this may be another mechanism by which tumors escape immune surveillance.

An imbalance between proinflammatory and anti-inflammatory cytokines produced within the local tumor microenvironment could be a second mechanism contributing to tumor-induced suppression of endothelial adhesion molecules [^{141 142 143 144 145 146 147 148} ]. In studies examining the regional production of cytokines within tumors, prominent production of Th2 cytokines (i.e., IL-10 and TGF-ß) with limited expression of proinflammatory Th1 cytokines (i.e., IL-2 and IFN-) appears to follow a relatively consistent pattern. In studies involving the production of regional cytokines within renal [¹⁴⁴ , ¹⁴⁵ ], non-small-cell lung [¹⁴⁷ , ¹⁴⁸ ], breast [¹⁴¹ ], and ovarian [¹⁴¹ ] carcinomas and malignant melanoma [¹⁴³ ], limited local generation of IL-2 or IFN- was reported. Concomitantly, increased production of IL-10 [^{144 145 146 147 148} ] and TGF-ß [¹⁴³ ] was noted. In one study of interest, the cytokine pattern within seborrheic keratosis was contrasted with that found in basal cell carcinoma. Although Th1-type cytokines (i.e., IL-2, IFN-, and TNF-ß) were prom-inently expressed within the benign neoplasm, a conversion to a Th2 cytokine pattern (i.e., IL-4, IL-5, and IL-10) was noted in the carcinomas [¹⁴² ]. Thus, in addition to angiogenic factors limiting the expression of endothelial adhesion molecules, local production of IL-10 and TGF-ß, which have been shown to down-regulate expression of activation-dependent endothelial adhesion molecules [^{64 65 66} ], might contribute to the observed diminished expression of VCAM-1 or E-selectin and the escape from immune surveillance.

Administration of proinflammatory cytokines (e.g., TNF-) to large tumors generating these inhibitory agents might not reverse the tumor-induced suppression. As noted previously, in clinical studies of isolated limb perfusion with TNF-, E-selectin but not VCAM-1 has been induced on vascular endothelial cells within melanoma tumors [¹²⁴ , ¹³² ]. In a murine melanoma model, PECAM-1 has been expressed on vascular endothelial cells both within the subcutaneous region and within the tumor (Fig. 1A ). In contrast, after the injection of 1µg of rmTNF- into the subcutaneous B16 tumor, VCAM-1 is readily induced only on normal vascular endothelium (Fig. 1B) [¹⁴⁹ ]. These observations suggest that therapies directed at decreasing tumor volume are critical to optimize expression of requisite endothelial adhesion molecules and to maximize leukocyte recruitment.

View larger version (190K): [in this window] [in a new window]   Figure 1. Lack of induction of VCAM-1 of tumor vascular endothelium after local injection of rmTNF-. TNF- (1 µg) was injected into a subcutaneous tumor of B16 melanoma and expression of endothelial adhesion molecules was examined 24 h later. Whereas PECAM-1 was expressed on vascular endothelial cells both within subcutaneous tissue and tumor [A (arrows)], expression of VCAM-1 was limited to vessels within the subcutaneous tissue [B (arrows)]. Arrowheads reflect the boundary between normal tissue and tumor.

After IL-2 infusion, evidence of endothelial activation on vessels within normal skin has been shown [^{150 151 152} ]. Perivascular infiltrations of mononuclear leukocytes have been documented after the systemic administration of IL-2 [¹⁵⁰ ] and up-regulation of ICAM-1 [^{150 151 152} ], and induction of E-selectin has been demonstrated [¹⁵¹ ]. However, in a report examining the changes in expression of endothelial adhesion molecules within metastatic breast carcinoma after the administration of systemic IL-2, no significant changes were noted in the expression of ICAM-1, P-selectin, E-selectin, or VCAM-1 on vascular endothelium within the tumors. In contrast, increased expression of these molecules has been noted on normal tissues biopsied at sites remote from metastases [¹⁵³ ]. Thus, systemic delivery of IL-2 may not recapitulate events meant to be local. Nonspecific immune activation may not induce local recruitment of effector leukocytes at sites of tumor.

Several reports have demonstrated that systemic activation of leukocytes may paradoxically result in impaired recruitment at sites of inflammation. Studies in vitro have demonstrated that the addition of a CD29-activating monoclonal antibody that induces the high-avidity binding site on VLA-4 prevents subsequent transendothelial migration of eosinophils [¹⁵⁴ ]. After activation by the intravenous administration of IL-8, neutrophil recruitment in rabbit [¹⁵⁵ ] and nonhuman-primate [¹⁵⁶ ] models of acute inflammation is hindered in a similar manner. Finally, systemic administration of LPS or TNF- also prevents neutrophil emigration at sites of local inflammation [¹⁵⁷ ]. These studies suggest that leukocyte recruitment is a regional event directed by localized changes in adhesion molecules and their counter-receptors.

Similar observations after the systemic delivery of IL-2 have been made in several preclinical tumor models. In a murine model of graft versus leukemia, localization of allogeneic, sensitized T cells at sites of leukemic tumor was inhibited by systemic administration of IL-2 [¹⁵⁸ ]. Although localization of recruited leukocytes is not examined directly in this report, deleterious effects of systemic IL-2 have been reported in two other studies. Although systemic delivery of sensitized T cells is curative in mice bearing intracerebral fibrosarcoma tumors, the addition of systemic IL-2 eliminates the benefit of the adoptive immunotherapy [¹⁵⁹ ]. Furthermore, the delivery of sensitized T cells to mice bearing the same tumor in the lungs demonstrates an augmented benefit after the addition of systemic IL-2. Thus, systemic IL-2 may alter lymphocyte trafficking at some but not all sites of tumor. Finally, in studies involving the transduction of murine tumors with a costimulatory molecule (B7.1) with or without additional tumor transduction with murine IL-2, only animals injected with tumors expressing both elements were cured. No protective effect was observed if mice bearing B7.1-transduced tumors were given systemic IL-2 [¹⁶⁰ ]. These studies demonstrate that systemic IL-2 might deter rather than facilitate immunotherapy, perhaps by failing to recapitulate the regional events that occur during leukocyte recruitment at sites of inflammation.

	SUMMARY

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 
Data presented in this review suggest that adhesive events that are requisite for leukocyte recruitment at sites of inflammation are defective within tumors. The majority of these abnormalities are caused by tumor-derived products that suppress the induction of activation-dependent endothelial adhesion molecules. Therefore, strategies geared towards optimal reduction of tumor burden with effective chemoradiotherapy followed by immunotherapy in an adjuvant setting could be the best method of circumventing this suppression.

An additional problem with current immunotherapy strategies using systemic IL-2 might result from systemic versus local activation of effector leukocytes. Systemic IL-2-based immunotherapy relies on the localization of endogenous effector cells (e.g., CTL or NK cells) at sites of tumor. Strategies using systemic delivery of cytokines/chemokines that activate leukocytes may fail to recapitulate events meant to occur in the local milieu during leukocyte recruitment. Thus, although some regressing tumors may become infiltrated with mononuclear leukocytes, the limited clinical responses currently observed after systemic IL-2-based therapy suggest that the majority of tumors escape the induction of local inflammation.

Another problem is that the failure of normal kinetics of leukocyte recruitment might also play a major role in the suboptimal responses observed after immunotherapy. Immunotherapy mediated solely by T cells fails to duplicate the normal sequence of leukocyte subset recruitment observed during inflammation and wound healing. Immunotherapy strategies of focusing on neutrophil-mediated approaches [reviewed in reference 161] have also been effective. Thus, the sequential recruitment of early effector cells (e.g., leukocytes providing innate immunity such as granulocytes, monocytes, or NK cells) may be necessary for optimal immunotherapy.

A role for immunotherapy in the treatment of common malignancies remains to be defined. Recruitment of endogenous leukocyte subsets or adoptively transferred subpopulations of leukocytes at sites of tumor appears to require some of the same adhesion molecules involved during leukocyte localization at sites of inflammation or wound healing. With the observations made in the latter paradigms, localization of leukocytes at sites of tumor will require the induction of endothelial adhesion molecules and the production of local proinflammatory agonists (e.g., cytokines or chemokines) to direct this process. Strategies based on therapies that systemically activate leukocytes, therefore, may result in suboptimal induction of inflammation at sites of tumor because these events typically are meant to occur within a regionalized microenvironment. In deference to the environmentalists, this could be one situation in which it is best to think locally and not act globally.

	ACKNOWLEDGEMENTS

 
The author acknowledges the contributions of Darcy Franicola and Beatrice Tiangco, MD, to this manuscript. Helpful conversations with collaborators at the University of Pittsburgh Cancer Institute, especially Drs. Ronald Herberman, Per Basse, Bill Chambers, Simon Watkins, John Bryant, and Elaine Elder, were instrumental during the writing of this review. This work was supported by a grant from the National Institutes of Health (1PO1 CA68550-01A2).

Received April 1, 2001; revised April 3, 2001; accepted April 3, 2001.

	REFERENCES

TOP ABSTRACT INTRODUCTION CELL ADHESION MOLECULES INVOLVED... KINETICS OF LEUKOCYTE... SUCCESSFUL RECRUITMENT OF... FAILURE OF ORCHESTRATION DURING... SUMMARY REFERENCES

 

1. Sparano, J. A., Dutcher, J. P. (1993) Interleukin-2 for treatment of advanced melanoma Hematol. Oncol. Ann. 1,289-295
2. Keilholz, U., Conradt, C., Legha, S. S., Khayat, D., Scheibenbogen, C., Thatcher, N., Goey, S. H., Gore, M., Dorval, T., Hancock, B., Punt, C. J. A., Drummer, R., Avril, M. F., Brocker, E. B., Benhammouda, A., Eggermont, A. M. M., Pritsch, M. (1998) Results of interleukin-2-based treatment in advanced melanoma: a case record-based analysis of 631 patients J. Clin. Oncol. 16,2921-2929[Abstract/Free Full Text]
3. Law, T. M., Motzer, R. J., Mazumdar, M., Sell, K. W., Walther, P. J., O’Connell, M., Khan, A., Vlamis, V., Vogelzang, N. J., Bajorin, D. F. (1995) Phase III randomized trial of interleukin-2 with or without lymphokine-activated killer cells in the treatment of patients with advanced renal cell carcinoma Cancer 76,824-832[Medline]
4. Bukowski, R. M. (1997) Natural history and therapy of metastatic renal cell carcinoma: the role of interleukin-2 Cancer 80,1198-1220[Medline]
5. Cohen, P. J., Lotze, M. T., Roberts, J. R., Rosenberg, S. A., Jaffe, E. S. (1987) The immunopathology of sequential tumor biopsies in patients treated with interleukin-2: correlation of response with T cell infiltration and HLA-DR expression Am. J. Pathol. 129,208-216[Abstract]
6. Rubin, J. T., Elwood, L. J., Rosenberg, S. A., Lotze, M. T. (1989) Immunohistochemical correlates of response to recombinant interleukin-2-based immunotherapy in humans Cancer Res 49,7086-7092[Medline]
7. Cole, D. J., Taubenberger, J. K., Pockaj, B. A., Yannelli, J. R., Carter, C., Carrasquillo, J., Leitman, S., Steinberg, S. M., Rosenberg, S. A., Yang, Y. C. (1994) Histopathological analysis of metastatic melanoma deposits in patients receiving adoptive immunotherapy with tumor-infiltrating lymphocytes Cancer Immunol. Immunother. 38,299-303[Medline]
8. Butcher, E. C. (1991) Leukocyte-endothelial cell recognition: three (or more) steps to specificity and diversity Cell 67,1033-1036[Medline]
9. Springer, T. A. (1994) Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm Cell 76,301-314[Medline]
10. Meyer, T., Hart, I. R. (1998) Mechanisms of tumour metastasis Eur. J. Cancer 34,214-221
11. Alexandroff, A. B., McIntyre, C. A., Porter, J. C., Zeuthen, J., Vile, R. G., Taub, D. D. (1998) Sticky and smelly issues: lessons on tumour cell and leucocyte trafficking, gene and immunotherapy of cancer Br. J. Cancer 77,1806-1811[Medline]
12. Kiessling, R., Wasserman, K., Horiguchi, S., Kono, K., Sjoberg, J., Pisa, P., Petersson, M. (1999) Tumor-induced immune dysfunction Cancer Immunol. Immunother. 48,353-362[Medline]
13. Whiteside, T. L. (1999) Signaling defects in T lymphocytes of patients with malignancy Cancer Immunol. Immunother. 48,343-345[Medline]
14. Arnaout, M. A. (1990) Structure and function of the leukocyte adhesion molecules CD11/CD18 Blood 75,1037-1050[Free Full Text]
15. Van der Vieren, M., Le Trong, H., Wood, C. L., Moore, P. F., St. John, T., Staunton, D. E., Gallatin, W. M. (1995) A novel leukointegrin, alpha d beta 2, binds preferentially to ICAM-3 Immunity 3,683-690[Medline]
16. Hemler, M. E., Elices, M. J., Parker, C., Takada, Y. (1990) Structure of the integrin VLA-4 and its cell-cell and cell-matrix adhesion functions Immunol. Rev. 114,45-65[Medline]
17. Kirveskari, J., Bobo, P., Granfors, K., Leirisalo-Repo, M., Jalkanen, S., Salmi, M. (2000) Expression of 4-integrins on human neutrophils J. Leukoc. Biol. 68,243-250[Abstract/Free Full Text]
18. Schweighoffer, T., Tanaka, Y., Tidswell, M., Erle, D. J., Horgan, K. J., Ginther-Luce, G. E., Lazarovits, A. I., Buck, D., Shaw, S. (1993) Selective expression of integrin 4ß7 on a subset of human CD4+ memory T cells with hallmarks of gut-trophism J. Immunol. 151,717-729[Abstract]
19. Erle, D. J., Briskin, M. J., Butcher, E. C., Garcia-Pardo, A., Lazarovits, A. I., Tidswell, M. (1994) Expression and function of the MAdCAM-1 receptor, integrin ₄ß₇, on human leukocytes J. Immunol. 153,517-528[Abstract]
20. Laszik, Z., Jansen, P. J., Cummings, R. D., Tedder, T. F., McEver, R. P., Moore, K. L. (1996) P-selectin glycoprotein ligand-1 is broadly expressed in cells of myeloid, lymphoid, and dendritic lineage and in some nonhematopoietic cells Blood 88,3010-3021[Abstract/Free Full Text]
21. Borges, E., Tietz, W., Steegmaier, M., Moll, T., Hallmann, R., Hamann, A., Vestweber, D. (1997) P-selectin glycoprotein lignad-1 (PSGL-1) on T helper 1 but not on T helper 2 cells binds to P-selectin and supports migration into inflamed skin J. Exp. Med. 185,573-578[Abstract/Free Full Text]
22. McEver, R. P., Cummings, R. D. (1997) Role of PSGL-1 binding to selectins in leukocyte recruitment J. Clin. Invest. 100,485-492[Medline]
23. Jutila, M. A., Kishimoto, T. K., Butcher, E. C. (1990) Regulation and lectin activity of the human neutrophil peripheral lymph node homing receptor Blood 76,1781-1783
24. Varki, A. (1997) Selectin ligands: will the real ones please stand up? J. Clin. Invest. 99,158-162[Medline]
25. Maly, P., Thall, A. D., Petryniak, B., Rogers, C. E., Smith, P. L., Marks, R. M., Kelly, R. J., Gersten, K. M., Cheng, G., Saunders, T. L., Camper, S. A., Camphausen, R. T., Sullivan, F. X., Isogai, Y., Hindsgaul, O., von Andrian, U. H., Lowe, J. B. (1996) The (1, 3) fucosyltransferase Fuc-TVII controls leukocyte trafficking through an essential role in L-, E-, and P-selectin ligand synthesis Cell 86,643-653[Medline]
26. Picker, L. J., Martin, R. J., Trumble, A., Newman, L. S., Collins, P. A., Bergstresser, P. R., Leung, D. Y. M. (1994) Differential expression of lymphocyte homing receptors by human memory/effector T cells in pulmonary versus cutaneous immune effector sites Eur. J. Immunol. 24,1269-1277[Medline]
27. Blander, J. M., Visintin, I., Janeway, C. A., Jr, Medzhitov, R. (1999) (1, 3)-Fucosyltransferase VII and (2, 3)-sialyltransferase IV are up-regulated in activated CD4 T cells and maintained after their differentiation into Th1 and migration into inflammatory sites J. Immunol. 163,3746-3752[Abstract/Free Full Text]
28. Easton, E. W., Schiphorst, W. E. C. M., van Drunen, E., van der Schoot, E., van den Eijnden, D. H. (1993) Human myeloid 3-fucosyltransferase is involved in the expression of sialyl-Lewis^x determinant, a ligand for E- and P-selectin Blood 81,2978-2986[Abstract/Free Full Text]
29. Stockinger, H., Gadd, S. J., Eher, R., Majdic, O., Schreiber, W., Kasinrerk, W., Strass, B., Schnabl, E., Knapp, W. (1990) Molecular characterization and functional analysis of the leukocyte surface protein CD31 J. Immunol. 145,3889-3897[Abstract]
30. Stockinger, H., Schreiber, W., Majdic, O., Holter, W., Maurer, D., Knapp, W. (1992) Phenotype of human T cells expressing CD31, a molecule of the immunoglobulin supergene family Immunology 75,53-58[Medline]
31. Tanaka, Y., Albelda, S. M., Horgan, K. J., van Seventer, G. A., Shimizu, Y., Newman, W., Hallam, J., Newman, P. J., Buck, C. A., Shaw, S. (1992) CD31 expressed on distinctive T cell subsets is a preferential amplifier of ß₁ integrin-mediated adhesion J. Exp. Med. 176,245-253[Abstract/Free Full Text]
32. Thomson, L. F., Ruedi, J. M., Glass, A., Moldenhauer, G., Moller, P., Low, M. G., Klemens, M. R., Massaia, M., Lucas, A. H. (1990) Production and characterization of monoclonal antibodies to the glycosyl phosphatidylinositol-anchored lymphocyte differentiation antigen ecto-5'-nucleotidase (CD73) Tissue Antigens 35,9-19[Medline]
33. Airas, L., Salmi, M., Jalkanen, S. (1993) Lymphocyte-vascular adhesion protein-2 is a novel 70-kDa molecule involved in lymphocyte adhesion to vascular endothelium J. Immunol. 511,4228-4238
34. Stewart, M., Hogg, N. (1996) Regulation of leukocyte integrin function: affinity vs avidity J. Cell. Biochem. 61,554-561[Medline]
35. Dustin, M. L., Springer, T. A. (1989) T cell receptor cross-linking transiently stimulates adhesiveness through LFA-1 Nature 341,619-624[Medline]
36. Elemer, G. S., Edgington, T. S. (1994) Monoclonal antibody to an activation neoepitope on _mß₂ inhibits multiple _mß₂ functions J. Immunol. 152,5836-5844[Abstract]
37. Yednock, T. A., Cannon, C., Vandevert, C., Goldbach, E. G., Shaw, G., Ellis, D. K., Liaw, C., Fritz, L. C., Tanner, L. I. (1995) ₄ß₁ Integrin-dependent cell adhesion is regulated by a low affinity receptor pool that is conformationally responsive to ligand J. Biol. Chem. 270,28740-28750[Abstract/Free Full Text]
38. Kishimoto, T. K., Jutila, M. A., Berg, E. L., Butcher, E. C. (1989) Neutrophil Mac-1 and MEL-14 adhesion proteins inversely regulated by chemotactic factors Science 245,1238-1241[Abstract/Free Full Text]
39. Kishimoto, T. K., Jutila, M. A., Butcher, E. C. (1990) Identification of a human peripheral lymph node homing receptor: a rapidly down-regulated adhesion molecule Proc. Natl. Acad. Sci. USA 87,2244-2248[Abstract/Free Full Text]
40. Luscinskas, F. W., Kiely, J. M., Ding, H., Obin, M. S., Hebert, C. A., Baker, J. B., Gimbrone, M. A., Jr (1992) In vitro inhibitory effect of IL-8 and other chemoattractants on neutrophil-endothelial adhesive interactions J. Immunol. 149,2163-2171[Abstract]
41. Bruehl, R. E., Moore, K. L., Lorant, D. E., Borregaard, N., Zimmerman, G. A., McEver, R. P., Bainton, D. F. (1997) Leukocyte activation induces surface redistribution of P-selectin glycoprotein ligand-1 J. Leukoc. Biol. 61,489-499[Abstract]
42. Luster, A. D. (1998) Chemokines-chemotactic cytokines that mediate inflammation N. Engl. J. Med. 338,436-445[Free Full Text]
43. Rottman, J. B. (1999) Key role of chemokines and chemokine receptors in inflammation, immunity, neoplasia, and infectious disease Vet. Pathol. 36,357-367[Abstract]
44. Murdoch, C., Finn, A. (2000) Chemokine receptors and their role in inflammation and infectious diseases Blood 95,3032-3043[Abstract/Free Full Text]
45. Dustin, M. L., Rothlein, R., Bhan, A. K., Dinarello, C. A., Springer, T. A. (1986) Induction by IL-1 and interferon-: tissue distribution, biochemistry, and function of a natural adherence molecule (ICAM-1) J. Immunol. 137,245-254[Abstract]
46. Nortamo, P., Li, R., Renkonen, R., Timonen, T., Prieto, J., Patarroyo, M., Gahmberg, G. C. (1991) The expression of human intercellular adhesion molecule-2 is refractory to inflammatory cytokines Eur. J. Immunol. 21,2629-2632[Medline]
47. de Fourgerolles, A. R., Stacker, S. A., Schwarting, R., Springer, T. A. (1991) Characterization of ICAM-2 and evidence for a third counter-receptor for LFA-1 J. Exp. Med. 174,253-267[Abstract/Free Full Text]
48. Albelda, S. M., Muller, W. A., Buck, C. A., Newman, P. J. (1991) Molecular and cellular properties of PECAM-1 (endoCAM/CD31): a novel vascular cell-cell adhesion molecule J. Cell Biol. 114,1059-1068[Abstract/Free Full Text]
49. Wagner, D. D. (1993) The Weibel-Palade Body: the storage granule for von Willebrand factor and P-selectin Thrombosis Haemost 70,105-110
50. Bevilacqua, M. P., Pober, J. S., Mendrick, D. L., Cotran, R. S., Gimbrone, M. A., Jr (1987) Identification of an inducible endothelial-leukocyte adhesion molecule Proc. Natl. Acad. Sci. USA 84,9238-9242[Abstract/Free Full Text]
51. Carlos, T. M., Schwartz, B. R., Kovach, N. L., Yee, E., Rosso, M., Osborn, L., Chi-Rosso, G., Newman, B., Lobb, R., Harlan, J. M. (1990) Vascular cell adhesion molecule-1 mediates lymphocyte adherence to cytokine-activated cultured human endothelial cells Blood 76,965-970[Abstract/Free Full Text]
52. Salmi, M., Jalkanen, S. (1992) A 90-kilodalton endothelial cell molecule mediating lymphocyte binding in humans Science 257,1407-1409[Abstract/Free Full Text]
53. Salmi, M., Kalimo, K., Jalkanen, S. (1993) Induction and function of vascular cell adhesion molecule-1 at sites of inflammation J. Exp. Med. 178,2255-2260[Abstract/Free Full Text]
54. Salmi, M., Tohka, S., Berg, E. L., Butcher, E. C., Jalkanen, S. (1997) Vascular adhesion protein-1 (VAP-1) mediates lymphocyte subtype-specific, selectin-independent recognition of vascular endothelium in human lymph nodes J. Exp. Med. 186,589-600[Abstract/Free Full Text]
55. Stocker, C. J., Sugars, K. L., Harari, O. A., Landis, R. C., Morley, B. J., Haskard, D. O. (2000) TNF-, IL-4, and IFN- regulate differential expression of P- and E-selectin by porcine aortic endothelial cells J. Immunol. 164,3309-3315[Abstract/Free Full Text]
56. Leeuwenberg, J. F., von Asmuth, E. J., Jeunhomme, T. M., Buurman, W. A. (1990) IFN-gamma regulates the expression of the adhesion molecule ELAM