The role of polymorphonuclear neutrophil granulocytes (PMN) in antitumoral immune responses displays a striking dichotomy. Under inflammatory conditions, PMN may promote tumor growth and progression. In contrast, especially in the context of therapeutic interventions, PMN can exert important antitumor functions. However, until now, the mechanisms of PMN-mediated activation of tumor immunity are poorly defined. Based on a murine model of Bacillus Calmette-Guérin (BCG) immunotherapy of bladder cancer, we provide evidence for a novel immunoregulatory role of this leukocyte subset. PMN immigrate into the bladder after intravesical BCG instillation and depletion of PMN from tumor-bearing mice completely abrogated antitumor efficacy of BCG. PMN stimulated with BCG in vitro as well as PMN isolated from the urine of BCG-treated patients were a major source of the chemokines interleukin-8, growth-related oncogene-α, macrophage inflammatory protein-1α and of the inflammatory cytokine migration inhibitory factor. In vitro, BCG-stimulated PMN indirectly induced T-cell chemotaxis via the accessory function of activated monocytes. In vivo, depletion of PMN from BCG-treated mice significantly impaired CD4+ T-cell trafficking to the bladder. These data show that PMN direct the migration of effector cells to the bladder and by this means are indispensable for effective tumor immunotherapy. Thus, our findings provide evidence for a novel early immunoregulatory role of these innate immune cells in local antitumor immunity. (Cancer Res 2006; 66(16): 8250-7)

Bladder cancer is the fifth most common malignant disease, with an incidence that reached >63,000 new cases in the United States alone in 2005 (1). On initial diagnosis, 75% to 80% of cancers are nonmuscle invasive tumors of stages pTa, pT1, or pTis. Depending on grade and stage, these tumors display highly variable rates of recurrence and progression at 5 years following transurethral resection (TUR) of up to 78% and 45%, respectively (2). In 1976, Morales et al. (3) reported for the first time on the successful intravesical use of Mycobacterium bovis Bacillus Calmette-Guérin (BCG) as an adjuvant treatment option following TUR. BCG immunotherapy consists of 6 weekly intravesical instillations. In high-risk tumors, an additional maintenance schedule is recommended (4). The mycobacteria induce a strong local immune response within the urinary bladder, characterized by the massive secretion of cytokines into the urine as well as inflammatory cellular infiltration (5). Today, BCG immunotherapy has been shown to be the only efficient adjuvant treatment option for preventing or delaying progression after TUR of nonmuscle invasive bladder cancers (6, 7) and represents one of the most successful immunotherapies of solid cancer.

Since the first use, many aspects of the complex immune response to BCG have been revealed. Following intravesical instillation, BCG provokes an early influx of innate immune cells, such as polymorphonuclear neutrophil granulocytes (PMN) and a proinflammatory cytokine milieu of predominantly TH1 type (5, 8, 9). After repeated instillations, the bladder wall is infiltrated by mononuclear cells, consisting mainly of monocytes/macrophages, CD4+ and CD8+ T lymphocytes, as well as natural killer (NK) cells, which form chronic granuloma-like cellular infiltrates in the suburothelial stroma (10). In addition to descriptive studies in humans, an orthotopic murine bladder cancer model was established to do functional investigations (11). Using this model, it was shown that TH1-polarized cell-mediated immunity plays a key role during BCG immunotherapy in vivo and that IFN-γ, NK cells, and CD8+ as well as CD4+ T cells are major cellular mediators of this antitumor action (1214).

Although T-cell–mediated cell lysis and the release of regulatory cytokines, such as IFN-α and IFN-γ, have been shown to represent late acquired immune events of the antitumor effector phase, thus far little attention has been paid to the early innate immune response that immediately follows BCG instillation (5). PMN are such early innate immune cells and the predominant subpopulation of leukocytes in the urine after BCG instillation (8, 9). Having long been regarded simply as phagocytic effector cells and potential mediators of BCG side effects, some studies have of late challenged this opinion. Today, PMN have been shown to be of critical importance in the early inflammatory response against mycobacterial infections, probably by providing an important link between innate and acquired immune responses (15, 16). Furthermore, their role in malignant tumor growth and tumor immunology is currently a matter of intensive debate. Although most studies suggest a tumor-promoting “nursing” function of PMN in advanced disease (17), under certain circumstances, they also seem to exert tumor-inhibitory actions during early cancer cell growth (18, 19).

Previously, we have been able to show extensive alterations in gene expression, enhanced interleukin (IL)-8 secretion and prolongation of survival in PMN following BCG stimulation (20). These data are in concordance with recent evidence that activated PMN act as an important source of chemokines of the CXC [e.g., IL-8 and growth-related oncogene (GRO)-α], as well as CC (e.g., macrophage inflammatory protein (MIP)-1α and MIP-1β), subfamily in response to mycobacterial stimulation in vitro (21). Yet, the implications of these observations in vivo are unknown. Functional studies on immune cell activation, chemotaxis, and trafficking in response to Mycobacterium-stimulated PMN are lacking. Clinical observations support a potential functional role for PMN during BCG immunotherapy. Local concentrations of IL-8 in the bladder were shown to predict clinical responses to BCG (22). The extent of leukocyturia as well as local irritative symptoms and fever correlate with a favorable outcome after BCG administration (23, 24). Stimulated by these observations, we hypothesized that via secretion of chemotactic factors, PMN might exert essential immunoregulatory functions during BCG immunotherapy by directing leukocyte traffic in the early inflammatory response in the bladder of patients. In fact, we show here for the first time that PMN isolated from the urine of BCG-treated patients are a relevant source of CC as well as CXC chemokines. In vitro and in a well-defined orthotopic murine bladder tumor model, activated PMN attract monocytes, which in turn results in BCG-induced CD4+ T-cell migration. Our findings define PMN as indispensable immunoregulatory cells during BCG immunotherapy.

Reagents and cell lines. Viable mycobacteria (BCG, Connaught substrain; kindly provided by Cytochemia, Ihringen, Germany) from the logarithmic growth phase were used for stimulation experiments. The number of colony-forming units (CFU) was routinely determined by plating and incubation at 37°C for 4 weeks on Löwenstein-Jensen agar. The murine bladder tumor cell line MB49 [kind gift of E.C. de Boer (University of Amsterdam, Amsterdam, the Netherlands) and M.A. O'Donnell (University of Iowa, Iowa City, IA)] was maintained in RPMI 1640 culture medium (Biochrom, Berlin, Germany) supplemented with 10% low-endotoxin FCS, penicillin, and streptomycin, hereafter referred to as complete medium, at 37°C in 5% CO2. The rat hybridoma cell line RB6-8C5 producing the anti-Gr-1 antibody was a kind gift from T. Laskay (Institute for Medical Microbiology and Hygiene, Lübeck University, Lübeck, Germany) and maintained in complete medium containing 1% FCS. For antibody preparation, supernatants were purified on sepharose G columns.

Murine bladder cancer model and PMN depletion. We used a well-defined murine syngeneic orthotopic MB49 bladder cancer model to evaluate the role of PMN for effective BCG immunotherapy in vivo (11). In brief, 6- to 8-week-old female C57BL/6 mice (The Jackson Laboratory, Bar Harbor, ME) were catheterized to receive an intravesical inoculate of 105 MB49 bladder tumor cells on day 0 following 5 seconds of electrocautery of the bladder wall. On days 1, 8, 15, and 22 following tumor implantation, mice were treated intravesically with either 3 × 106 CFU BCG or PBS control for a total of four instillations. Groups of eight mice each were followed daily for survival or terminal morbidity. To do immunohistochemical bladder wall stainings, groups of six mice each were sacrificed on day 9 and the bladders were processed for cryosections. PMN were depleted from BCG- or PBS-treated mice using an anti-Gr-1 monoclonal antibody (mAb) or a respective isotype control according to a modified protocol as described (25). RB6-8C5 (250 μg) or the isotype control antibody (250 μg) was given i.p. every 2 to 3 days as indicated in the respective figure legends. Efficacy of depletion was constantly monitored during experiments by counting the relative numbers of PMN on Giemsa-stained whole mouse blood smears. RB6-8C5 treatment reduced PMN blood counts to <10% of pretreatment levels. Neutropenia persisted for 10 days after which repletion to levels equal to or even greater than pretreatment levels occurred. In control experiments, 100% of tumor-free mice survived the antibody injections (data not shown). All animal experiments done during this study were approved by the ministry for environmental, agricultural, and forestry affairs of the county of Schleswig-Holstein, Germany.

Immunohistochemistry. Mouse bladders were resected, shock frozen in liquid nitrogen, and stored at −80°C. Immunohistochemical detection of various immune cell subpopulations was done on 5-μm cryostat sections following a routine staining protocol using peroxidase-conjugated secondary antibodies as published previously (12, 13). Antibodies and their respective isotype controls against the listed mouse antigens were as follows: CD4 (clone RM4-5, 500 ng/mL; BD Biosciences, Heidelberg, Germany), Gr-1 (clone RB6-8C5, 700 ng/mL; see above), and CD11c (clone N418, 1 μg/mL; Serotec, Düsseldorf, Germany). To quantify the cellular bladder infiltrate, groups of six mice were used. The number of infiltrating immune cells was determined by examining 7 to 12 randomly selected nonoverlapping microscopic fields (0.126 mm2) at ×250 magnification. Counting of cells was done in blinded fashion by two independent individuals.

Patients, healthy donors, and urine and peripheral blood samples. Ten urine samples from BCG-treated patients were collected 20 to 24 hours past the first, second, or third instillation of the induction cycle. Simultaneously, peripheral venous blood was drawn from the respective patients as well as from healthy volunteers. Urinalysis of pre-BCG urine samples from all individuals revealed clean urine devoid of cellular or bacterial components. Urine samples were immediately processed by centrifugation and washed twice in complete medium. These cell suspensions contained >95% PMN as confirmed by Giemsa stainings and flow cytometry. Cell suspensions of peripheral blood PMN from patients and healthy volunteers were obtained according to a standard protocol described recently (20). The experiments were approved by the ethics committee of the Lübeck University Medical School (#02-137) after obtaining written informed consent from all individuals.

PMN purification protocol. To remove contaminating monocytes and lymphocytes from PMN in patients' urine and peripheral blood before doing ELISAs, we developed an antibody-based purification protocol. Cell suspensions containing peripheral blood PMN and urine PMN were labeled at a concentration of 3 × 107/mL with 1 μg/mL mAb directed against CD66b (clone 80H3; BD/PharMingen, Heidelberg, Germany) and subsequently coated with goat anti-mouse immunomagnetic microbeads (Miltenyi; Bergisch Gladbach, Heidelberg, Germany). CD66b is constitutively expressed on PMN, induced on activation, and has been identified as a major galectin-3 receptor (26). Cells were processed on LS VarioMACS separation columns (Miltenyi) and washed thrice before eluting them in MACS buffer containing 2 mmol/L EDTA and 0.5% bovine serum albumin. Purity of PMN in the eluate was >99.9% as determined by extensive flow cytometric analyses. Using FITC-conjugated anti-CD66b mAbs, we did not detect any differences in the expression of CD66b on peripheral blood PMN and urine PMN before and after separation. These data are in concordance with a recent publication, which showed effective PMN isolation with anti-CD15 immunobeads without cell activation (27).

Human IL-8, GRO-α, MIP-1α, and migration inhibitory factor ELISA. Following purification, peripheral blood PMN and urine PMN were analyzed for chemokine secretion using commercially available assay kits supplied by R&D systems (Minneapolis, MI) according to the manufacturer's instructions.

Monocyte and T-cell chemotaxis assay. A method to assay monocyte and T-cell chemotaxis was developed using a modification of the endogenous component chemotaxis assay described previously (28). Monocytes were isolated from peripheral blood mononuclear cells (PBMC) from healthy human donors by counterflow centrifugation (elutriation). This method yields monocyte suspensions with >99% purity. T cells were generated by culture of PBMC in complete medium supplemented with 2.5 μg/mL phytohemagglutinin for 3 days and subsequently for 6 days in complete medium supplemented with 200 units/mL recombinant human IL-2. After this procedure, cultures contained T cells of >98% purity as determined by flow cytometry. Monocytes (107) or T cells (107) were radioactively labeled in complete medium with 200 μCi Na251CrO4 for 2 hours. Labeled cells (0.5 × 106) in 200 μL complete medium were placed into the upper compartments of a 24-Transwell Costar plate (Corning, New York, NY) with 5-μm pore membranes. The lower compartments contained the various chemoattractants in 500 μL complete medium as indicated in figure legends. At different time points, cells in the lower compartments and cells detached from the lower side of the pore membranes were lysed in 0.1% Triton X-100 and radioactive release in supernatants was determined in a gamma counter (Berthold LB2111, Bad Wildbach, Germany). Percentage of migrated cells was determined in relation to the maximal radioactive release of 0.5 × 106 lysed labeled cells (representing 100% migration).

Statistical analyses. Survival of mice was evaluated using Kaplan-Meier plots and the log-rank test. Significance of differences in leukocyte bladder wall infiltration and PMN chemokine secretion was determined using a Mann-Whitney U test. Results are shown as mean with error bars representing ± SD. Statistical analyses are given as two-sided Ps with a P < 0.05 being considered a statistically significant difference.

PMN are essential for effective BCG immunotherapy in vivo. Clinical observations support a potential functional role for PMN during BCG immunotherapy of bladder cancer (2224). We used a well-defined murine preclinical model of orthotopic bladder cancer (11) to test this hypothesis in vivo and evaluated the influx of PMN into the bladder wall following intravesical BCG instillation. In this model, C57BL/6 mice were intravesically treated with BCG, and the bladders were subjected to anti-Gr-1 immunohistochemistry to visualize neutrophils. Whereas untreated control bladders are moderately infiltrated by PMN (Fig. 1A), BCG-treated tumor-bearing mice show a dense neutrophilic infiltrate (Fig. 1B). To determine whether these PMN actually contribute to the antitumor effect mediated by BCG, we compared efficacy of BCG therapy in normal control mice and mice depleted for PMN. Successful depletion of PMN was verified by counting Giemsa-stained blood smears (Fig. 2A) and by immunohistology of bladder specimens (data not shown). Reduced PMN counts were observed for up to 11 days. Kaplan-Meier survival analyses show an expected therapeutic benefit of BCG treatment in control mice. Importantly, depletion of PMN abrogated the therapeutic effect of BCG and reduced survival to levels of untreated control mice (hazard ratio, 3.55; 95% confidence interval, 1.42-8.88; Fig. 2B; P = 0.007). These results show that BCG therapy is ineffective in mice devoid of PMN and suggest an important functional role of PMN for the antitumor efficacy of BCG.

PMN direct T-cell trafficking to bladders of BCG-treated mice. Two potential mechanisms on how PMN could contribute to the antitumor effect of BCG are the following: (a) direct antitumor cytotoxicity (29) and (b) indirect orchestration and modulation of the subsequent immune response. We tested whether the second mechanism is operative during BCG immunotherapy. At first, we determined whether recruitment of cellular mediators of adaptive immunity, such as dendritic and T cells, depends on the presence of PMN. Mice injected with anti-Gr1 or control antibody were intravesically treated with BCG or PBS. Mice were sacrificed and bladder walls were immunostained for CD11c (expressed on the majority of murine dendritic cells) and CD4 antigen expression. CD4+ T cells are of central importance for effective BCG immunotherapy and represent the predominant immune cell subpopulation in BCG-induced granuloma-like bladder infiltrates (8, 10, 14). They exert crucial accessory functions, such as IFN-γ production or activation of cytotoxic effector lymphocytes, like CD8+ T and NK cells (5, 30). As depicted in Fig. 3A, intravesical BCG instillation strongly induced CD4+ T-cell infiltration in normal control mice (P < 0.05). At the same time, infiltration with CD11c+ dendritic cells was less pronounced in BCG-treated mice (Fig. 3B). Strikingly, BCG-induced influx of CD4+ T cells was completely abolished in mice depleted of PMN. These results clearly suggest that T-cell migration to BCG-treated bladders depends on the presence of PMN in vivo.

BCG-activated PMN indirectly induce T-cell chemotaxis. We used an in vitro Transwell-based two chamber system to obtain direct evidence for and mechanistic insight into BCG-induced PMN-mediated T-cell migration. In a first experimental setting, we analyzed whether BCG-stimulated PMN directly induce T-cell migration. For this purpose, PMN were stimulated with BCG in the lower compartments of Transwell plates, and T-cell chemotaxis through the separating membranes was assayed. Stimulated monocytes served as a positive control in our system and induced substantial T-cell migration. However, there was clearly no increase in T-cell chemotaxis toward supernatants of stimulated PMN compared with culture medium or unstimulated PMN controls even if high amounts of PMN were used (Fig. 4A).

We and others have shown previously that the induction of cytotoxic antitumor effector cells during BCG immunotherapy requires a complex interaction of various immune cell subpopulations, including monocytes (5, 9, 30, 31). Monocytes are a potent source of CC chemokines with a strong chemotactic effect on T cells (32, 33). Therefore, we tested whether PMN-induced T-cell migration requires the accessory function of monocytes. First, we investigated monocyte migration toward stimulated PMN (Fig. 4B) and then analyzed whether monocytes, after priming in conditioned medium derived from BCG-stimulated PMN, would acquire the potential to induce T-cell migration (Fig. 4C). In the first approach, PMN in the lower compartments of Transwell plates were stimulated with BCG and tested for the induction of monocyte chemotaxis. Compared with cell-free culture medium and unstimulated PMN, BCG-activated PMN clearly induced migration of monocytes. Monocyte migration was dependent on the time of PMN stimulation and a 3-hour stimulation period was found to be optimal in this system (Fig. 4B). These results show for the first time the direct induction of monocyte chemotaxis following mycobacterial stimulation of PMN.

Although chemokines primarily induce chemotaxis, they additionally synergize with proinflammatory cytokines to activate immune cells, such as monocytes, to undergo functional changes in inflammatory disease (32, 33). We reasoned that BCG-activated PMN in combination with activated monocytes induce T-cell chemotaxis. To test this hypothesis, monocytes were conditioned for 24 hours with the culture supernatant of unstimulated and BCG-stimulated PMN. Thereafter, T-cell migration toward PMN-conditioned monocytes was measured. Interestingly, following priming in culture medium of BCG-stimulated PMN, monocytes induced substantial T-cell migration reaching >50% of the T-cell migration induced by lipopolysaccharide (LPS)–stimulated monocytes. Again, induction of migration was dependent on the time of PMN stimulation with an optimum between 3 and 18 hours (Fig. 4C). To our knowledge, these are the first functional studies providing evidence for the fact that stimulated PMN have the capacity to indirectly induce T-cell chemotaxis via the accessory function of activated monocytes.

BCG-activated peripheral blood PMN and urinary PMN of bladder cancer patients are a major source of inflammatory chemokines and cytokines. In a recent report, we have shown that BCG stimulation prolonged the life span of peripheral blood PMN from healthy donors and induced gene expression of a set of inflammatory and regulatory cytokines and chemokines (20). In the present report, using a preclinical murine model together with in vitro migration assays, we have now established an important immunoregulatory role for PMN in directing subsequent local antitumor immune responses during BCG immunotherapy. Although certain chemokines have already been detected in the urine of patients undergoing BCG therapy, the cellular source of these molecules has not yet been identified (4, 22). Consequently, the aim of the final part of our study was to investigate expression of inflammatory chemokines and cytokines by urinary PMN from bladder cancer patients undergoing BCG immunotherapy and to compare this expression with the expression observed in peripheral blood PMN.

Following BCG instillation, PMN are the predominant immune cell population in the urine (9). However, because even small numbers of contaminating monocytes or lymphocytes are likely to also produce chemokines, we first developed a purification protocol to obtain PMN cultures close to homogeneity. This protocol is based on the positive selection of PMN using magnetic bead-coupled antibodies directed against CD66b and lead to the successful depletion of contaminating cells (mostly monocytes) from urinary PMN (Fig. 5A) and peripheral blood PMN (Fig. 5B).

Subsequent analysis of a broad range of chemokines revealed that the inflammatory chemokines IL-8, GRO-α, and MIP-1α were indeed consistently produced by urinary PMN isolated from patients in the course of BCG immunotherapy. Expression levels were in the same range as expression levels of BCG-stimulated peripheral blood PMN, suggesting substantial in vivo activation of urinary PMN (Fig. 5C). The chemokines MCP-1, IP-10, RANTES, and MDC were not expressed by urinary PMN and peripheral blood PMN, whereas MIP-1β could be induced in peripheral blood PMN but was absent in urinary PMN (data not shown). Interestingly, high levels of macrophage migration inhibitory factor (MIF), a still poorly investigated inflammatory cytokine that might play a role in bacterial sepsis and innate immunity to different microbial pathogens, were detectable in both peripheral blood PMN and urinary PMN (Fig. 5C).

Circulating human PMN are a short-lived subpopulation of innate immune cells with a half-life of only 6 to 10 hours after being released from the bone marrow and undergo rapid apoptosis followed by elimination in the liver and spleen (34). On tissue injury, they home to the site of trauma or infection and migrate from the bloodstream into the inflamed tissue. These activated, infiltrating PMN have a considerably longer life span than their circulating counterparts (35). They can phagocytose microorganisms, generate reactive oxygen intermediates, and release lytic enzymes with antimicrobial potential (21). PMN are the first immune cells to leave the blood vessels and infiltrate tissue that has been invaded by mycobacteria (36). Although recent evidence from in vivo studies emphasized the importance of PMN in the immune response to mycobacterial infections, their functional role in this process remains ill defined (15, 16). Similarly, in bladder cancer patients treated with BCG, PMN are the first and most abundant immune cell subpopulation immediately following BCG instillation (8, 9). Yet, until today, they have been largely ignored as playing any role in the BCG-induced antitumor action. Studies by various groups, including ours, showed secretion of a large array of proinflammatory cytokines and chemokines from PMN following mycobacterial stimulation in vitro (20, 37, 38). The present study was designed to define a potential functional role of PMN during BCG immunotherapy of bladder cancer. Our results clearly show that PMN are an indispensable immune cell subpopulation for effective BCG therapy. Using the same in vivo bladder tumor model, CD4+ and CD8+ T cells as well as NK cells have been identified as the major immune cells mediating the late antitumor effector phase (1214). We hypothesized that early PMN-derived chemotactic factors are necessary to secondary attract effector cell populations. Following BCG instillation, high levels of proinflammatory cytokines and also chemokines, such as IL-8, can be detected in the urine of treated patients, yet the cellular origin of these mediators had thus far been unknown (5, 22). In the present report, we show de novo synthesis of GRO-α as well as an increase in IL-8, MIP-1α, and MIF secretion in activated PMN from the urine of BCG patients and from the peripheral blood of healthy donors. With respect to the vast numbers of PMN in the urine after BCG administration, we reason that these cells are the dominating source of early urinary chemokines, such as IL-8. Interestingly and in contrast to most other cytokines, IL-8 has been shown to appear in the urine already after the first BCG instillation. Additionally, high urinary IL-8 concentrations are associated with the subsequent development of a strong immune response and low disease recurrence rate (22, 39). To our knowledge, GRO-α, MIP-1α, and MIF have thus far not been detected in the urine of BCG-treated patients in larger studies. We now report the active secretion of CXC as well as CC chemokines by BCG-stimulated PMN in vitro and ex vivo and provide evidence for a functional role of PMN-derived chemotactic factors in monocyte and T-cell migration. Specifically, we can show that BCG-activated PMN directly induce monocyte chemotaxis and indirectly regulate T-cell chemotaxis via the accessory function of monocytes. Thus far, we could not attribute this effect to a single chemokine. However, considering the diverse biological effects of chemokines, they are likely to act in concert (32, 33). Notably, this activity seems to be subject to tightly controlled counterregulation as prolonged PMN stimulation leads to a secondary down-regulation of monocyte and T-cell migration.

Although this is the first functional study to describe monocyte and T-cell migration following mycobacterial stimulation of PMN, there is some evidence concerning T-cell chemotaxis in response to mycobacterial monocyte/macrophage activation. Ferrero et al. (40) reported on the induction of TH1 cell and γδ-T-cell chemotaxis after pulsing alveolar macrophages isolated from patients with various pulmonary diseases with Mycobacterium tuberculosis cell wall components. Interestingly, mycobacterial stimulation of resting monocytes alone was not sufficient, whereas stimulation of preactivated alveolar macrophages induced strong T-cell migration. This is in concordance with our experiments, in which T-cell attraction was due to the combined action of BCG-activated PMN and primed monocytes. Additionally, the crucial role of neutrophils for T-cell attraction was shown in vivo. PMN depletion from BCG-treated mice significantly reduced the influx of CD4+ T cells into the bladder wall. We also tried to stain BCG-treated mouse bladders in control and PMN-depleted mice for macrophages. Unfortunately, all antibodies (e.g., anti-Mac-1 and F4/80) that are available to date either display strong cross-reactivity against the rat antibody required for in vivo PMN depletion or show coexpression on PMN and macrophages (e.g. CD11b).

PMN depletion from mice with the RB6-8C5 antibody is a commonly used and well-established procedure (25, 41). Recently, however, plasmacytoid IFN-α-producing dendritic cells have been described to express low levels of the Gr-1 antigen together with CD11c (42, 43). Interestingly, despite BCG-induced recruitment of T cells to the bladder, CD11c counts were only marginally induced on BCG stimulation. In anti-Gr-1-treated mice, a minimal reduction in CD11c counts was observed, possibly reflecting a codepletion of Gr-1/CD11c double-positive cells. Although we have not formally addressed the role of CD11c+/Gr-1+ cells in our model, these data suggest that CD4+ T-cell recruitment is largely independent of CD11c+ cells and almost exclusively depends on the activity of CD11c-/Gr-1+ PMN.

Whereas, with the exception of IL-8 (see above), there are thus far no studies investigating the role of chemokines in BCG immunotherapy, there are some studies addressing the role of chemokines during infections with M. tuberculosis in humans. IL-8, GRO-α, and MIP-1α together with MCP-1 and IP-10 are up-regulated in granuloma formation following M. tuberculosis infection (40, 44, 45). MIP-1α, MCP-1, and IP-10 are major factors for the chemoattraction of TH1 and CD8+ T cells, and IL-8, GRO-α, MIP-1α, and MCP-1 are important in monocyte recruitment (32, 33, 40, 44, 45). However, the cellular source of the respective chemokines has thus far not been revealed in vivo. Nevertheless, these data support our initial hypothesis that PMN-derived chemokines are important for initiating the BCG-induced immune response by attracting other immune cells to the bladder. Thus, it is tempting to speculate that both in the immune response following M. tuberculosis infection and in BCG therapy of cancer BCG-activated PMN might induce a two-step process resulting in direct monocyte and indirect T-cell recruitment.

In addition to IL-8, GRO-α, and MIP-1α, we also detected a significant increase in MIF secretion from BCG-stimulated peripheral blood PMN as well as from urine PMN compared with baseline levels. MIF is an interesting and still somewhat mysterious cytokine with diverse actions within the immunoneuroendocrine system that are yet to be fully understood. Recently, it has been described to serve as a central regulator of innate and inflammatory immune responses. Thus, it controls the “set point” and magnitude of immune reponses by overriding the immunosuppressive effects of glucocorticoids. Even more important, accumulating experimental evidence suggests a crucial role of MIF in the pathogenesis of different inflammatory diseases and in the innate immune response to a broad range of microbial pathogens (46, 47). PMN have only been identified recently as a significant source of MIF in the induction and onset of bacterial sepsis in vivo (48). Our data now suggest that PMN-derived MIF might also be involved in the control of the BCG-induced inflammatory antitumor response.

Our identification of PMN as a powerful source of proinflammatory chemokines and cytokines in vivo also adds further evidence to the current discussion about the role of PMN in tumor immunology. In advanced malignant diseases, PMN have been described to exert a nursing function on tumor cells by providing growth and proangiogenic factors as well as matrix-degrading enzymes (17, 18). Recently, however, an anticancer potential of PMN during early tumor growth was shown in preclinical animal models (18). In murine models, the application of tumor cells genetically engineered to release various proinflammatory cytokines results in a massive local immune reaction characterized by numerous PMN, which results in the rejection of the tumor cells (49). In these studies, PMN additionally cooperate with antigen-presenting cells and T cells to establish significant antitumor immune memory against the wild-type parental tumor (50). Considering our results, it could well be that the major role of PMN also in those experimental systems lies in the early recruitment and activation of adaptive immune cells, such as T cells.

In summary, we describe for the first time that BCG-stimulated PMN are an indispensable subset of immunoregulatory cells and orchestrate T-cell chemotaxis to the bladder during BCG immunotherapy. Using functional chemotaxis assays, it could be shown that activated PMN additionally induce monocyte migration and exploit the accessory function of these immune cells to attract T cells. In concordance with previous studies, our results provide strong evidence for a mechanism of PMN-mediated tumor cell rejection, in which PMN shape the development of the antitumor immune response by providing early chemotactic and activating factors for subsequent adaptive immunity.

Grant support: Deutsche Forschungsgemeinschaft Collaborative Research Center grant SFB 367.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Kirsten Bruderek and the laboratory technician trainees of the Division of Immunotherapy (Borstel, Germany) for their technical assistance.

1
Jemal A, Murray T, Ward E, et al. Cancer statistics, 2005.
CA Cancer J Clin
2005
;
55
:
10
–30.
2
Sylvester RJ, van der Meijden APM, Oosterlinck W, et al. Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials.
Eur Urol
2006
;
49
:
466
–74.
3
Morales A, Eidinger D, Bruce AW. Intracavitary Bacillus Calmette-Guérin in the treatment of superficial bladder tumors.
J Urol
1976
;
116
:
180
–3.
4
Alexandroff A, Jackson AM, O'Donnell M, James K. BCG immunotherapy of bladder cancer: 20 years on.
Lancet
1999
;
353
:
1689
–94.
5
Böhle A, Brandau S. Immune mechanisms in bacillus Calmette-Guerin immunotherapy for superficial bladder cancer.
J Urol
2003
;
170
:
964
–9.
6
Sylvester RJ, van der Meijden AP, Lamm DL. Intravesical bacillus Calmette-Guerin reduces the risk of progression in patients with superficial bladder cancer: a meta-analysis of the published results of randomized clinical trials.
J Urol
2002
;
168
:
1964
–70.
7
Herr HW, Schwalb DM, Zhang ZF, et al. Intravesical bacillus Calmette-Guerin therapy prevents tumor progression and death from superficial bladder cancer: ten-year follow-up of a prospective randomized trial.
J Clin Oncol
1995
;
13
:
1404
–8.
8
Prescott S, James K, Hargreave TB, Chisholm GD, Smyth JF. Intravesical Evans strain BCG therapy: quantitative immunohistochemical analysis of the immune response within the bladder wall.
J Urol
1992
;
147
:
1636
–42.
9
de Boer EC, de Jong WH, van der Meijden AP, et al. Leukocytes in the urine after intravesical BCG treatment for superficial bladder cancer. A flow cytofluorometric analysis.
Urol Res
1991
;
19
:
45
–50.
10
Böhle A, Gerdes J, Ulmer AJ, Hofstetter AG, Flad H-D. Effects of local Bacillus Calmette-Guérin therapy in patients with bladder carcinoma on immunocompetent cells of the bladder wall.
J Urol
1990
;
144
:
53
–8.
11
Günther JH, Jurczok A, Wulf T, et al. Optimizing syngeneic orthotopic murine bladder cancer (MB49).
Cancer Res
1999
;
59
:
2834
–7.
12
Brandau S, Suttmann H, Riemensberger J, et al. Perforin-mediated lysis of tumor cells by Mycobacterium bovis Bacillus Calmette-Guerin-activated killer cells.
Clin Cancer Res
2000
;
6
:
3729
–38.
13
Brandau S, Riemensberger J, Jacobsen M, et al. NK cells are essential for effective BCG immunotherapy.
Int J Cancer
2001
;
92
:
697
–702.
14
Ratliff TL, Ritchey JK, Yuan JJJ, Andriole GL, Catalona WJ. T-cell subsets required for intravesical BCG immunotherapy for bladder cancer.
J Urol
1993
;
150
:
1018
–23.
15
Appelberg R, Castro AG, Gomes S, Pedrosa J, Silva MT. Susceptibility of beige mice to Mycobacterium avium: role of neutrophils.
Infect Immun
1995
;
63
:
3381
–7.
16
Fulton SA, Reba SM, Martin TD, Boom WH. Neutrophil-mediated mycobacteriocidal immunity in the lung during Mycobacterium bovis BCG infection in C57BL/6 mice.
Infect Immun
2002
;
70
:
5322
–7.
17
Coussens LM, Werb Z. Inflammation and cancer.
Nature
2002
;
420
:
860
–7.
18
Di Carlo E, Forni G, Musiani P. Neutrophils in the antitumoral immune response. In: Cassatella MA, editor. Basel: The neutrophil; 2003. p. 182–203.
19
Vakkila J, Lotze MT. Inflammation and necrosis promote tumour growth.
Nat Rev Immunol
2004
;
4
:
641
–8.
20
Suttmann H, Lehan N, Bohle A, Brandau S. Stimulation of neutrophil granulocytes with Mycobacterium bovis Bacillus Calmette-Guerin induces changes in phenotype and gene expression and inhibits spontaneous apoptosis.
Infect Immun
2003
;
71
:
4647
–56.
21
Cassatella MA. Neutrophil-derived proteins: selling cytokines by the pound.
Adv Immunol
1999
;
73
:
369
–509.
22
Thalmann GN, Sermier A, Rentsch C, Mohrle K, Cecchini MG, Studer UE. Urinary Interleukin-8 and 18 predict the response of superficial bladder cancer to intravesical therapy with bacillus Calmette-Guerin.
J Urol
2000
;
164
:
2129
–33.
23
Lüftenegger W, Ackermann DK, Futterlieb A, et al. Intravesical versus intravesical plus intradermal bacillus Calmette-Guerin: a prospective randomized study in patients with recurrent superficial bladder tumors.
J Urol
1996
;
155
:
483
–7.
24
Saint F, Patard JJ, Irani J, et al. Leukocyturia as a predictor of tolerance and efficacy of intravesical BCG maintenance therapy for superficial bladder cancer.
Urology
2001
;
57
:
617
–21.
25
Jensen J, Warner T, Balish E. Resistance of SCID mice to Candida albicans administered intravenously or colonizing the gut: role of polymorphonuclear leukocytes and macrophages.
J Infect Dis
1993
;
167
:
912
–9.
26
Feuk-Lagerstedt E, Jordan ET, Leffler H, Dahlgren C, Karlsson A. Identification of CD66a and CD66b as the major galectin-3 receptor candidates in human neutrophils.
J Immunol
1999
;
163
:
5592
–8.
27
Zahler S, Kowalski C, Brosig A, Kupatt C, Becker BF, Gerlach E. The function of neutrophils isolated by a magnetic antibody cell separation technique is not altered in comparison to a density gradient centrifugation method.
J Immunol Methods
1997
;
200
:
173
–9.
28
Creamer HR, Gabler WL, Bullock WW. Endogenous component chemotactic assay (ECCA).
Inflammation
1983
;
7
:
321
–9.
29
Kemp TJ, Ludwig AT, Earel JK, et al. Neutrophil stimulation with Mycobacterium bovis bacillus Calmette-Guerin (BCG) results in the release of functional soluble TRAIL/Apo-2L.
Blood
2005
;
106
:
3474
–82.
30
Thanhäuser A, Böhle A, Schneider B, et al. The induction of bacillus-Calmette-Guerin-activated killer cells requires the presence of monocytes and T-helper type-1 cells.
Cancer Immunol Immunother
1995
;
40
:
103
–8.
31
Suttmann H, Jacobsen M, Reiss K, Jocham D, Bohle A, Brandau S. Mechanisms of Bacillus Calmette-Guérin-mediated natural killer cell activation.
J Urol
2004
;
172
:
1490
–5.
32
Baggiolini M. Chemokines and leukocyte traffic.
Nature
1998
;
392
:
565
–8.
33
Luster AD. Chemokines—chemotactic cytokines that mediate inflammation.
N Engl J Med
1998
;
338
:
436
–45.
34
Savill JS, Wyllie AH, Henson JE, Walport MJ, Henson PM, Haslett C. Macrophage phagocytosis of aging neutrophils in inflammation. Programmed cell death in the neutrophil leads to its recognition by macrophages.
J Clin Invest
1989
;
83
:
865
–75.
35
Coxon A, Tang T, Mayadas TN. Cytokine-activated endothelial cells delay neutrophil apoptosis in vitro and in vivo. A role for granulocyte/macrophage colony-stimulating factor.
J Exp Med
1999
;
190
:
923
–34.
36
Kobayashi K, Allred C, Castriotta R, Yoshida T. Strain variation of bacillus Calmette-Guerin-induced pulmonary granuloma formation is correlated with anergy and the local production of migration inhibition factor and interleukin 1.
Am J Pathol
1985
;
119
:
223
–35.
37
Riedel DD, Kaufmann SH. Chemokine secretion by human polymorphonuclear granulocytes after stimulation with Mycobacterium tuberculosis and lipoarabinomannan.
Infect Immun
1997
;
65
:
4620
–3.
38
Kasahara K, Sato I, Ogura K, Takeuchi H, Kobayashi K, Adachi M. Expression of chemokines and induction of rapid cell death in human blood neutrophils by Mycobacterium tuberculosis.
J Infect Dis
1998
;
178
:
127
–37.
39
de Boer EC, Somogyi L, de Ruiter GJ, de Reijke TM, Kurth KH, Schamhart DH. Role of interleukin-8 in onset of the immune response in intravesical BCG therapy for superficial bladder cancer.
Urol Res
1997
;
25
:
31
–4.
40
Ferrero E, Biswas P, Vettoretto K, et al. Macrophages exposed to Mycobacterium tuberculosis release chemokines able to recruit selected leucocyte subpopulations: focus on γδ cells.
Immunology
2003
;
108
:
365
–74.
41
Wipke BT, Allen PM. Essential role of neutrophils in the initiation and progression of a murine model of rheumatoid arthritis.
J Immunol
2001
;
167
:
1601
–18.
42
Asselin-Paturel C, Boonstra A, Dalod M, et al. Mouse type I IFN-producing cells are immature APCs with plasmacytoid morphology.
Nat Immunol
2001
;
2
:
1144
–50.
43
Menges M, Baumeister T, Rossner S, et al. IL-4 supports the generation of a dendritic cell subset from murine bone marrow with altered endocytosis capacity.
J Leukoc Biol
2005
;
77
:
535
–43.
44
Ragno S, Romano M, Howell S, Pappin DJ, Jenner PJ, Colston MJ. Changes in gene expression in macrophages infected with Mycobacterium tuberculosis: a combined transcriptomic and proteomic approach.
Immunology
2001
;
104
:
99
–108.
45
Sadek MI, Sada E, Toossi Z, Schwander SK, Rich EA. Chemokines induced by infection of mononuclear phagocytes with mycobacteria and present in lung alveoli during active pulmonary tuberculosis.
Am J Respir Cell Mol Biol
1998
;
19
:
513
–21.
46
Calandra T, Roger T. Macrophage migration inhibitory factor: a regulator of innate immunity.
Nat Rev Immunol
2003
;
3
:
791
–800.
47
Baugh JA, Donnelly SC. Macrophage migration inhibitory factor: a neuroendocrine modulator of chronic inflammation.
J Endocrinol
2003
;
179
:
15
–23.
48
Riedemann NC, Guo RF, Gao H, et al. Regulatory role of C5a on macrophage migration inhibitory factor release from neutrophils.
J Immunol
2004
;
173
:
1355
–9.
49
Musiani P, Allione A, Modica A, et al. Role of neutrophils and lymphocytes in inhibition of a mouse mammary adenocarcinoma engineered to release IL-2, IL-4, IL-7, IL-10, IFN-α, IFN-γ, and TNF-α.
Lab Invest
1996
;
74
:
146
–57.
50
Giovarelli M, Cappello P, Forni G, et al. Tumor rejection and immune memory elicited by locally released LEC chemokine are associated with an impressive recruitment of APCs, lymphocytes, and granulocytes.
J Immunol
2000
;
164
:
3200
–6.