Inhibition of Mac-1 (CD11b/CD18) enhances tumor response to radiation by reducing myeloid cell recruitment

Abstract

Despite recent advances in radiotherapy, loco-regional failures are still the leading cause of death in many cancer patients. We have previously reported that bone marrow-derived CD11b⁺ myeloid cells are recruited to tumors grown in irradiated tissues, thereby restoring the vasculature and tumor growth. In this study, we examined whether neutralizing CD11b monoclonal antibodies could inhibit the recruitment of myeloid cells into irradiated tumors and inhibit their regrowth. We observed a significant enhancement of antitumor response to radiation in squamous cell carcinoma xenografts in mice when CD11b antibodies are administered systemically. Histological examination of tumors revealed that CD11b antibodies reduced infiltration of myeloid cells expressing S100A8 and matrix metalloproteinase-9. CD11b antibodies further inhibited bone marrow-derived cell adhesion and transmigration to C166 endothelial cell monolayers and chemotactic stimuli, respectively, to levels comparable to those from CD11b knockout or CD18 hypomorphic mice. Given the clinical availability of humanized CD18 antibodies, we tested two murine tumor models in CD18 hypomorphic or CD11b knockout mice and found that tumors were more sensitive to irradiation when grown in CD18 hypomorphic mice but not in CD11b knockout mice. When CD18 hypomorphism was partially rescued by reconstitution with the wild-type bone marrow, the resistance of the tumors to irradiation was restored. Our study thus supports the rationale of using clinically available Mac-1 (CD11b/CD18) antibodies as an adjuvant therapy to radiotherapy.

Radiotherapy plays a crucial role in cancer treatment, especially for inoperable tumors. Recent advances, including image-guided and intensity-modulated radiotherapy, leading to higher and more precise dose delivery, has achieved superior treatment outcomes (1). However, despite these advances, recurrence of the primary tumors still remains the leading cause of death of patients treated with radiotherapy (2). This finding highlights the fact that we need an improved understanding of the reasons for treatment failure.

We have previously shown that irradiated tumors, or tumors grown in previously irradiated tissues (thereby mimicking recurrent primary tumors), recruit large numbers of bone marrow-derived CD11b⁺ myeloid cells expressing matrix metalloproteinase-9 (MMP-9) (3). We further demonstrated that these MMP-9-expressing myeloid cells restored tumor vasculature and allowed tumor growth in the irradiated tissues of MMP-9 knockout (KO) mice, suggesting that these cells could be an important target in radiotherapy (3). There is strong evidence to suggest that tumor-infiltrating CD11b⁺ myeloid cells promote angiogenesis, and do so by expressing various proangiogenic and chemoattractant molecules, including VEGF (4), Bv8 (5), and S100A8 (6). However, despite their tumor-promoting roles, targeting CD11b⁺ myeloid cells as a cancer therapy has proven difficult.

CD11b (Mac-1, α_Mβ₂) is the α-subunit of the predominant β₂ (CD18) integrin expressed on monocytes/macrophages and granulocytes (7). This subunit has been shown to mediate many functions of myeloid cells, including adhesion, migration, chemotaxis, phagocytosis, and respiratory burst activity (7). Studies have reported that antibodies to CD11b or CD18 inhibit these functions (8) and in vivo administration of the antibodies reduce leukocyte recruitment into various sites of inflammation (9). Given the promising preclinical activities of CD11b/CD18 antibodies in inhibiting myeloid cell adhesion onto endothelium, humanized antibodies [LeukArrest; 23F2G (10)] have been developed and tested in patients with stroke, multiple sclerosis, or myocardial infarction (11). However, although the antibodies showed excellent safety profiles, they lacked therapeutic efficacy (10, 11).

In this study, we used CD11b-neutralizing monoclonal antibodies as a means to inhibit the recruitment of myeloid cells to irradiated tumors. With systemic administration of these CD11b antibodies following local tumor irradiation, we observed a significant enhancement of tumor response to radiation accompanied by a reduced infiltration of myeloid cells expressing MMP-9 and S100A8 into the tumors. We also observed that CD18 hypomorphism, which had lowered CD11b surface expression myeloid cells, significantly associated with the sensitivity of tumors to radiation. Together, these results suggest that clinically available humanized antibodies against CD11b/CD18 could be useful as an adjuvant therapy to radiotherapy.

Results

Radiation Inhibits Local Angiogenesis but the Vasculature Is Restored in Recurrent Tumors Accompanied by Infiltrating Myeloid Cells.

To determine the effects of local irradiation on angiogenesis, we first examined the histology of FaDu human head and neck squamous cell carcinoma xenografts grown in immune-deficient mice that were either unirradiated (control, 0 Gy), harvested shortly after 20 Gy of irradiation (IR 20 Gy), or had recurred after irradiation with 20 Gy, which took ≈2 months following irradiation [recurrent (2 mo)] (Fig. 1A). Tumor volumes were ≈25 and 50% of the unirradiated control for irradiated tumors and recurrent tumors, respectively (Fig. 1A). By staining for endothelial cells and pericytes with CD31 and α-smooth muscle actin (α-SMA) antibodies, respectively, we found that the irradiated tumors had significantly fewer endothelial cells and pericytes compared with the control tumors (Fig. 1 A and B). However, when the tumors had recurred after irradiation, the number of endothelial cells had returned to control levels, although the pericyte coverage was partially restored (Fig. 1 A and B). To determine the myeloid cell contribution to the changes in the tumor vasculature, we stained the tumor sections with CD11b for myeloid cells and CD45 for all leukocytes. We observed that the levels of myeloid cells were increased in the irradiated tumors, and this increase was maintained in the recurrent tumors (Fig. 1 A and B). Leukocyte levels were transiently increased in the irradiated tumors but in the recurrent tumors had decreased to levels approaching that in the control tumors (Fig. 1B). To separate the effects of irradiation on existing versus new vasculature derived from angiogenesis, we implanted matrigel plugs in mice and irradiated them with 20 Gy. In unirradiated (control) matrigels, there was extensive penetration of CD31⁺ endothelial cells into the plugs, some of which were associated with α-SMA⁺ pericytes (Fig. 1C). Infiltrating blood vessels were functional, as demonstrated by Hoechst 33342 positivity, injected immediately before the matrigel harvest. However, there were significantly fewer endothelial cells in the irradiated matrigel plugs (Fig. 1 C and D), demonstrating that irradiation inhibits the development of de novo vasculature. Overall, the results indicate that although radiation produces a major depletion of the tumor vasculature (presumably because of killing of endothelial cells in the tumor and inhibition of local angiogenesis from surrounding normal tissues), tumors that regrow following irradiation have a restored vasculature accompanied by an increased accumulation of CD11b⁺ myeloid cells.

Fig. 1.

Local irradiation inhibits angiogenesis. (A) Staining of FaDu tumors grown in immunodeficient mice that were not irradiated (control, 0 Gy), irradiated with 20 Gy and harvested at day 14 postirradiation [IR 20 Gy (d14)], or recurrent after 20 Gy of irradiation [recurrent (2 mo)]. The tumors were stained for CD31 (red) and α-SMA (green) (Upper), and for CD11b (green) (Lower). Nuclei were stained with DAPI and are shown in blue. (Scale bars, 100 μm.) The bar graph shows the mean tumor volume. (B) Quantification of CD31, α-SMA, CD11b, and CD45 shown in A. Symbols and error bars in A and B are the mean ± SEM for n ≥ 5 animals per group. *, **, and *** denote P < 0.05, < 0.01, and < 0.001 by one-way ANOVA, respectively. (C) Immunostaining of matrigel implanted in mice as in A, stained with CD31 (red) and α-SMA (green) antibodies. Hoechst 33342, a diffusion dye injected immediately before the matrigel harvest, is shown in blue. (Scale bar, 100 μm.) (D) Quantification of the matrigel section from C as in B. Symbols and error bars are the mean ± SEM for n = 4 animals per group. ***, P < 0.001 by Student's t test.

CD11b Antibody Treatment Enhances Tumor Response to Radiation.

Given that recurrent tumors after radiation have an increased infiltration of CD11b⁺ myeloid cells, we determined whether antibodies to CD11b could reduce the myeloid cell recruitment and sensitize the tumors to irradiation. We harvested CD11b monoclonal antibodies (IgG2b isotype) from M1/70 hybridoma and first determined the dose and timing of the antibodies for in vivo administration. We similarly studied Gr-1 antibodies (IgG2b isotype) harvested from RB6-8C5 hybridoma. We first observed that Gr-1 antibodies efficiently depleted granulocytes (CD11b⁺Gr-1⁺ cells) at 24 h after a single i.p. administration; CD11b antibodies did not affect the leukocyte composition (Fig. S1A and Table S1). CD11b antibodies exhibited a complete epitope blockage with 100 μg at 24 h (Fig. S1 C and D), which was partially reversed at 72 h after administration (Fig. S1E). Therefore, to maintain constant epitope blocking of myeloid cells, we treated the mice with CD11b antibodies at 100 μg per mouse every 2 days. Gr-1 antibodies were administered to a separate group of animals in a similar manner. To monitor epitope blockage by CD11b antibodies or granulocyte depletion by Gr-1 antibodies, we sampled peripheral blood from the treated animals once every four days for FACS analysis (Figs. S1H and S2D).

When FaDu tumor-bearing immunodeficient mice were treated with CD11b antibodies following either 12 or 20 Gy of a single dose of irradiation given locally to the tumors, we observed a significant enhancement of tumor response to radiation (Fig. 2A). Macroscopically, the CD11b antibody-treated tumors shrank dramatically, leading to nearly nonpalpable tumors by the end of the study (Fig. 2B) (12 of 16 tumor cures in CD11b antibody treatment and 7 of 19 in the control groups in three independent experiments, P < 0.05 by two-tailed Student's t test). In contrast, Gr-1 antibodies exhibited little or no effect (Fig. 2A), indicating that CD11b⁺Gr-1⁺ cells play a less important role than CD11b⁺Gr-1⁻ cells in influencing tumor regrowth after irradiation. Further analysis of CD11b⁺Gr-1⁺ and CD11b⁺Gr-1⁻ cells by FACS revealed that the former population of cells had slightly higher expression of mature myeloid lineage markers, including macrophage colony-stimulating factor-receptor (MCSF-R), 7/4, and F4/80 (Fig. S1 F and G). We further tested CD11b antibodies in immunocompetent C3H/HeJ mice bearing SCCVII tumors and observed that tumor regrowth after irradiation was also inhibited (Fig. 2C).

Fig. 2.

CD11b monoclonal antibody treatment enhances tumor response to radiation. (A) Growth of irradiated FaDu tumors with 12 Gy (Left) or 20 Gy (Right) in mice treated with isotype control antibodies (control), Gr-1 antibodies (Gr-1 Ab), or CD11b antibodies (CD11b Ab) at 100 μg per mouse from the fourth day following irradiation for every 2 days. (B) Photographs of mice bearing FaDu tumors that had been irradiated with 20 Gy and treated with isotype control antibodies (control, Left) or CD11b antibodies (CD11b Ab, Right) for up to 2 months. Tumors had regrown in the control group (black arrowheads), whereas they became not palpable in the CD11b Ab group (black arrowheads indicate where the tumor had been originally implanted). (C) Growth of irradiated SCCVII tumors in C3H/HeJ mice with 15 Gy, followed by the control or CD11b antibodies. Symbols and error bars in A and C are the mean ± SEM for n ≥ 7 per group.

We further tested the possibility that CD11b antibodies might sensitize normal tissues to irradiation to a similar extent as that observed with the tumors. To do this, we irradiated a region of normal skin on the back of nontumor-bearing mice with 20, 25, or 30 Gy and treated the animals with either saline or CD11b antibodies on the same schedule as that used for the tumor-bearing mice. We found that CD11b antibodies did not sensitize, but rather protected the normal skin from the radiation-induced skin reaction (P < 0.01 for median scores between the control and CD11b antibody treated groups for all radiation doses, by two-tailed Mann-Whitney t test) (Fig. S2A). This finding suggests that the enhanced tumor response to radiation by CD11b antibodies is not a result of sensitizing normal tissues to radiation but rather of direct effects caused by CD11b antibodies to the irradiated tumors.

CD11b or Gr-1 antibodies alone, on the other hand, showed no effects on the growth of nonirradiated tumors (Fig. S2B). Histology of the antibody-treated tumors showed similar area densities for CD31⁺ endothelial cells and α-SMA⁺ pericytes in all groups (Fig. S2 E and F). However, we observed less CD45⁺ leukocyte infiltration in the CD11b antibody-treated tumors, whereas the Gr-1 antibody-treated tumors showed significantly increased levels of CD45⁺ cells compared with the control tumors (Fig. S2 E and F).

CD11b Antibodies Reduce Infiltration of Myeloid Cells Expressing S100A8 and MMP-9.

To determine whether the enhanced tumor response to radiation by CD11b antibodies was associated with reduced myeloid cell infiltration in the irradiated tumors, we examined the histology of FaDu tumors in mice treated with CD11b antibodies or isotype control antibodies at 7 and 14 days after irradiation. We found reduced levels of both CD11b⁺ myeloid cells and CD45⁺ leukocyte levels in the CD11b antibody-treated tumors on day 7 after 20 Gy of irradiation (Fig. 3 A and B). However, there was no significant difference in CD31⁺ endothelial cells between the two groups at day 7, although α-SMA⁺ pericytes were lower in the irradiated tumors treated with CD11b antibodies (Fig. 3 A and B). At day 14, CD11b antibody-treated tumors showed significantly reduced levels of CD45⁺ leukocytes, although CD11b⁺ myeloid cells, CD31⁺ endothelial cells, and α-SMA⁺ pericytes were similar to the control antibody treated tumors (Fig. S3B).

Fig. 3.

CD11b monoclonal antibodies inhibit tumor infiltrating myeloid cells expressing S100A8 and MMP-9. (A) Immunostaining of FaDu tumors from Fig. 2A (20 Gy) harvested at 7 days (d7) after irradiation, stained for CD11b⁺ cells using CD11b (control tumors) or anti-rat (CD11b Ab-treated tumors) antibodies (Upper). (Lower) CD31 (red) and α-SMA (green) staining. Nuclei are shown in blue with DAPI staining. (B) Quantification of immunostaining in A as area densities for CD11b, CD45, CD31, and α-SMA. (C) Quantification of S100A8 immunostaining (Upper) and colocalization with CD11b (Lower) in unirradiated (No IR) or recurrent (IR 20 Gy) tumors in Fig. 1A. (D) Immunostaining of S100A8 (red) and MMP-9 (green) in unirradiated or recurrent tumors as shown in C. Quantification of colocalization between S100A8⁺ cells and MMP-9 is shown in the bar graph. (E) Immunostaining of d 7 FaDu tumors as shown in A for S100A8 (red) and CD11b (green; for control tumors), or anti-rat (green; for CD11b Ab-treated tumors) antibodies. The bar graph shows quantification of S100A8 immunostaining. (Scale bars for A, D, and E: 100 μm.) Symbols and error bars in B to E are the mean ± SEM for n ≥ 3 mice per group. *, **, and *** denote P < 0.05, < 0.01, and < 0.001 by Student's t test, respectively.

To further identify the myeloid subsets affected by CD11b antibody treatment in irradiated tumors, we first stained unirradiated (No IR) or recurrent (IR 20 Gy) FaDu tumors with various myeloid markers including S100A8 and F4/80 in conjunction with CD11b. We found a strong colocalization (85–99%) between CD11b and S100A8 markers (Fig. 3C), whereas F4/80 was poorly colocalized with CD11b in the tumors (Fig. S3A). S100A8 showed significantly elevated levels in irradiated compared to unirradiated tumors (Fig. 3C), although F4/80 levels did not significantly change (Fig. S3A). We further observed that S100A8⁺ myeloid cells strongly expressed MMP-9, which revealed ≈80% of colocalization (Fig. 3D). Irradiated tumors treated with CD11b antibodies exhibited a significant reduction of S100A8⁺ myeloid cells at day 7 (Fig. 3E) and day 14 (Fig. S3B), consistent with the above results with CD11b staining (Fig. 3B and Fig. S3B).

CD11b Antibodies Inhibit Adhesion and Transmigration of Bone Marrow-Derived Cells.

To determine how CD11b antibodies reduced the infiltration of myeloid cells in the irradiated tumors, we first examined whether irradiation increases the expression of intercellular adhesion molecule-1 (ICAM-1), a ligand of CD11b integrin receptor (12), on endothelial cells. We observed that surface expression of ICAM-1 (Fig. 4A), but not vascular cell adhesion molecule (VCAM-1) (Fig. S3C) was increased in irradiated C166 endothelial cells in a dose- and time-dependent manner. We then examined the effect of CD11b antibodies on adhesion of carboxyfluorescein succinimidyl ester (CFSE)-labeled bone marrow-derived cells onto C166 endothelial cells expressing endogenous levels of ICAM-1. We observed that CD11b antibodies significantly reduced bone marrow-derived cell adhesion onto the endothelial cells compared with the isotype control antibodies (Fig. 4B). Furthermore, this effect was comparable to that of bone marrow-derived cells isolated from CD11b KO mice or CD18 hypomorphic mice (Fig. 4B), the latter resembling moderate levels of human leukocyte adhesion deficiency (13). We further investigated whether CD11b antibodies influence transmigration efficiency of the bone marrow-derived cells by using a modified Boyden chamber assay. We observed that pretreatment of the bone marrow-derived cells with CD11b antibodies essentially abolished the increased migration of the cells toward all of the chemotactic stimuli, resulting in the number of migrated cells similar to the control group with no chemotaxis (Fig. 4C). A similar lack of migration toward the chemotactic stimuli was seen in the bone marrow-derived cells isolated from CD11b KO mice or CD18 hypomorphic mice (Fig. 4C).

Fig. 4.

CD11b antibodies inhibit adhesion and transmigration of bone marrow-derived cells. (A) Irradiated C166 endothelial cells analyzed by FACS showed up-regulation of ICAM-1 expression in a dose- (0–20 Gy) and time- (24 h, Left; 48 h, Right) dependent manner. (B) Fluorescent images of adhered CFSE-labeled bone marrow cells (green) onto C166 endothelial cell monolayers (nuclei of the endothelial cells are shown in blue by DAPI staining) that were either pretreated with isotype control (control) or CD11b antibodies (CD11b Ab; Upper), or isolated from CD11b KO mice or CD18 hypomorphic (CD18 hypo) mice (Lower). (Scale bar, 100 μm.) Quantification of the number of CFSE-positive cells per field is shown on the right. (C) CFSE-labeled bone marrow derived cells as shown in B were incubated in modified Boyden chambers containing the culture media supplemented with no chemokine (no treatment), 10% serum, VEGF, or M-CSF. Symbols and error bars in B and C are the mean ± SEM for triplicate determinations in three independent experiments. ***, P < 0.001 by one-way ANOVA.

CD18 Hypomorphism Influences Tumor Sensitivity to Radiation.

Because humanized CD18 antibodies are clinically available and have shown similar efficacy in inhibiting myeloid cell functions to CD11b antibodies (14), we tested CD18 hypomorphic mice and CD11b KO mice for tumor response to radiation by using Lewis lung carcinomas (LLC) or MC38 colon adenocarcinomas, which are syngeneic to these mouse strains. We observed that tumors were more sensitive to radiation when grown in CD18 hypomorphic mice, but not in CD11b KO mice compared with the WT mice (Fig. 5A). As the latter result was unexpected in view of the above data with CD11b antibodies, we examined the histology of the tumors. We found that there were infiltrating S100A8⁺ myeloid cells in the irradiated tumors from CD11b KO mice, despite of the genetic absence of CD11b (Fig. 5B). The levels of S100A8⁺ myeloid cells were similar between CD11b KO and WT mice (Fig. 5B). We then examined CD11b levels in CD18 hypomorphic mice by subjecting the CFSE-labeled bone marrows isolated from CD18 hypomorphic mice to FACS analysis. CD18 hypomorphic mice exhibited a significantly lower surface expression of CD11b on their bone marrow-derived cells compared with the WT mice (Fig. 5C). To determine whether we could modulate tumor radiosensitivity by rescuing CD18 hypomorphism, we transplanted the WT bone marrow into nonwhole-body-irradiated CD18 hypomorphic mice once a week for 4 weeks. At 6 weeks we examined the bone marrow reconstitution by peripheral blood analyses and found that CD18 hypomorphism was partially rescued by the WT bone marrow (Fig. 5 D and E). We further observed that the response of MC38 tumors to radiation in the CD18 hypomorphic mice was reversed to the WT level in the CD18 hypomorphic mice reconstituted with the WT bone marrow (Fig. 5F). Histology of the tumors showed that infiltrating CD11b⁺ myeloid cells (Fig. 5G) and CD18⁺ cells (Fig. 5H) were dramatically increased in the CD18 hypomorphic mice partially reconstituted with the WT bone marrow, which otherwise was very low in the CD18 hypomorphic mice. These results are consistent with the antibody neutralization study, suggesting that inhibition of Mac-1 (by CD11b antibodies or by CD18 hypomorphism) enhances tumor response to radiation through inhibiting the infiltration of myeloid cells expressing S100A8.

Fig. 5.

CD18 hypomorphism influences tumor response to radiation. (A) Growth of irradiated LLC (Left) or MC38 (Right) tumors with 15 Gy in the WT, CD11b KO (CD11b KO), or CD18 hypomorphic (CD18 hypo) mice. (B) Immunostaining of irradiated LLC in WT or CD11b KO as in A for S100A8 (red) and CD11b (green). DAPI shows nuclear staining in blue. Quantification of S100A8 area densities is shown in the bar graph. (C) FACS plots showing CFSE-labeled bone marrow cells isolated from WT or CD18 hypo mice for CD11b surface expression. (D) FACS analyses of the peripheral blood obtained from WT, CD18 hypo, or CD18 hypomorphic mice reconstituted with the WT bone marrow cells (CD18 hypo + WT). (E) Quantification of propidium iodide-negative, live CD18⁺ cells (as highlighted in the red boxes in D) in the WT, CD18 hypo, or CD18 hypo + WT mice. (F) MC38 tumor growth after irradiation with 15 Gy in WT, CD18 hypo, or CD18 hypo + WT mice shown in D and E. (G and H) Immunostaining of MC38 tumors in F for CD11b (red, G), CD18 (red, H), and CD45 (green, G). Nuclei are shown in blue with DAPI staining. (Scale bars in B, G, and H: 100 μm.) The symbols and error bars represent the mean ± SEM for n ≥ 5 per group (for A, E, and F) or n ≥ 3 per group (for B, G, and H). ** and *** denote for P < 0.01 and < 0.001, respectively, determined by one-way ANOVA.

Discussion

In this study, we show that CD11b monoclonal antibodies reduced the radiation-induced infiltration of myeloid cells into squamous cell carcinoma xenografts in mice (Fig. 3). Furthermore, this inhibition led to a significant enhancement of tumor response to irradiation (Fig. 2), with no significant effect on unirradiated tumors (Fig. S2). These results are consistent with our earlier hypothesis (3) that under normal conditions, tumors support their growth by local angiogenesis (by proliferation and migration of endothelial cells from nearby blood vessels) with little or no requirement for vasculogenesis from the bone marrow-derived circulating cells. However, as we have shown earlier with transplanted murine tumors (3), when the tumor and immediately surrounding normal tissue are irradiated (which severely inhibits local angiogenesis, as we show in Fig. 1), tumors become highly dependent on the vasculogenesis pathway to restore the tumor vasculature following its depletion by irradiation, thereby supporting tumor recurrence. Therefore, both these and our prior data (3) demonstrate the importance of vasculogenesis, and in particular CD11b cells, for restoring the tumor vascular and allowing tumor regrowth following irradiation.

There is currently no therapy available to selectively deplete CD11b⁺ myeloid cells in humans. A number of preclinical studies have reported that Gr-1 antibodies deplete myeloid cells and that this depletion leads to an inhibition of tumor growth (15). However, although we observed an efficient depletion of Gr-1⁺ cells, which were consistent with these reports, Gr-1 antibodies did not inhibit the growth of FaDu tumors in mice with (Fig. 2A) or without irradiation (Fig. S2). This result suggests that CD11b⁺Gr- 1⁻ cells, rather than CD11b⁺Gr-1⁺ cells, may play an important role in restoring the vasculature after irradiation. Recent studies have suggested that CD11b⁺Gr-1⁻ cells are the major proangiogenic myeloid cells promoting tumor vascularization (16) and progression to metastases (17).

Studies have shown that tumors recruit myeloid cells by secreting cytokines, including VEGF and M-CSF to facilitate neovascularization (18, 19). VEGF and M-CSF signal through VEGF receptor-1 and M-CSF receptor, respectively, which are expressed on myeloid cells, hence modulating their migrating abilities (20, 21). Once recruited to tumors, myeloid cells may produce various proangiogenic cytokines, including stromal-derived factor-1, TGF-β, and VEGF, as shown by macrophages cocultured with conditioned media derived from tumor cells (22). The present findings that CD11b antibodies inhibit transmigration of bone marrow-derived cells toward chemotactic stimuli thus suggest a mechanism by which the antibodies could attenuate the myeloid cell recruitment to tumors, thereby inhibiting neovascularization by vasculogenesis.

CD11b monoclonal antibodies are known to modify many functions of myeloid cells. Here, we showed that the antibodies inhibited adhesion and migration of the bone marrow-derived cells to the endothelial monolayers and chemotactic stimuli, respectively (Fig. 4). Myeloid cells adhere on the endothelium via an interaction between CD11b (on the myeloid cells) and ICAM-1 (on the endothelium) (12). In this study, we observed that radiation directly up-regulated ICAM-1 expression on the endothelial cells (Fig. 4), consistent with a study by Hallahan et al., who showed that irradiation increased ICAM-1 expression on endothelial cells and that leukocyte adhesion occurs concurrently with ICAM-1 expression (23). Hence, we speculate that the enhanced ICAM-1 expression on the irradiated endothelium may be the first step in recruiting myeloid cells into the irradiated tumors. Consistent with this hypothesis, Handschel et al. have reported that ICAM-1 expression in the endothelium and myeloid cell infiltration were significantly increased in radiation-induced inflammation in oral mucosa of head and neck cancer patients treated with radiotherapy (24).

Our study also shows that CD18 hypomorphism, but not genetic deficiency of CD11b, affects tumor sensitivity to radiation (Fig. 5). The enhanced radiosensitivity of tumors in CD18 hypomorphism was associated with lowered levels of CD11b surface expression on the bone marrow cells and subsequently lowered infiltration of CD11b⁺ myeloid cells into irradiated tumors (Fig. 5). The apparent paradox of the lack of radiosensitivity of tumors in CD11b KO mice was explained by the infiltration of S100A8⁺ myeloid cells to the irradiated tumors in these mice. S100A8 is a myeloid-specific intracellular calcium binding protein (25) and has been reported to participate in various inflammatory responses, including vascular injury (26), by regulating myeloid cell chemotaxis and adhesion (27). Furthermore, it has been demonstrated that S100A8 increases CD11b expression thereby facilitating adhesion of CD11b⁺ myeloid cells onto its ligand (28). Our data demonstrating that CD11b antibodies, but not the genetic deficiency of CD11b, could inhibit infiltration of S100A8⁺ myeloid cells, therefore indicate that S100A8 is required for CD11b cell surface expression in myeloid cells and this is important for contributing to tumor resistance to radiation.

Relevant to a possible use with radiotherapy, we found that CD11b antibodies did not sensitize normal skin to irradiation but rather protected it from radiation-induced damage (Fig. S2A). In agreement with this, Epperly et al. reported reduced levels of pulmonary damage in mice treated with CD11b antibodies or in mice transplanted with the bone marrow cells from CD18 hypomorphic mice, and these effects were accompanied by a reduction of macrophage migration to the irradiated lungs (29).

In conclusion, we present evidence that inhibiting the infiltration of myeloid cells into irradiated xenografts by CD11b monoclonal antibodies enhances their response to irradiation. Because of their excellent safety profiles demonstrated in clinical trials and even protection of irradiated normal tissues, we believe that CD11b antibodies are an attractive candidate for further evaluation as an adjunct therapy to radiotherapy.

Materials and Methods

Mice and Tumors.

All animal procedures were approved by Stanford's Administrative Panel on Laboratory Animal Care. Strains of the mice used are: nu/nu immunodeficient nude mice (Charles River), CD11b KO mice (B6.129S4-Itgam^tm1Myd/J; Jackson Laboratories), CD18 hypomorphic mutant mice (B6.129S7-Itgb2^tm1Bay/J; Jackson Laboratories), C57BL/6J (Jackson Laboratories), and C3H/HeJ (Jackson Laboratories). Female mice at 6 to 8 weeks of age were used. The mice were maintained in a germ-free environment and had access to food and water ad libitum.

Cells were maintained in Waymouth medium containing 15% FCS for FaDu cells or in DMEM plus 10% FCS for C166 endothelial cells [American Tissue Culture Collection (ATCC)], Lewis lung carcinoma (ATCC), MC38, and SCCVII cells.

FaDu cells (6–8 × 10⁶ cells per mouse), LLC (5 × 10⁵ cells per mouse), MC38 (5 × 10⁵ cells per mouse), and SCCVII (5 × 10⁵ cells per mouse) were inoculated and measured as described previously (3) in nude (for FaDu), C57BL/6J, CD11b KO, and CD18 hypomorphic mice (for LLC or MC38), or C3H/HeJ (for SCCVII) mice. Tumors were irradiated as described (3) when their volumes reached ∼200 mm³ for FaDu and ∼100 mm³ for LLC, MC38, or SCCVII tumors using a Phillips x-ray unit.

Immunostaining and Quantification.

The matrigel sections were fixed with 4% PFA in PBS for 30 min at room temperature before staining. Tumors were harvested as described earlier (3). Antibodies were purchased from BD Pharmingen for CD31 (biotin anti-mouse; clone Mec13.3), CD45 (biotin or FITC-conjugated; clone 30-F11), CD11b (biotin or PE-conjugated; clone M1/70), and Gr-1 (FITC-conjugated; clone RB6-8C5); Sigma for α-SMA antibodies (FITC-conjugated, clone 1A4); Serotec for F4/80 antibodies (488 conjugated, clone CI:A3-1) and 7/4 (647 conjugated, clone 7/4); e-Bioscience for MCSF-R (PE-conjugated; clone AFS98); and from R&D Systems for S100A8 (goat anti-mouse). The sections were stained as described previously (3) and examined and digital fluorescence microscopic images were taken using a Leica DM6000B microscope (with HC PL FLUOTAR 20× and 40× objective lenses with HC PLAN 10×/25 eyepieces) and Retiga Exi Q imaging camera. Images were developed with Image-Pro-6.2 software. For analysis, images at 20× objective were taken from nonnecrotic and viable tumor regions away from the edge of tumors, and analyzed at least four sections per tumor or matrigel and three to five animals per group by using ImageJ software for area densities as described elsewhere (30) or Image-Pro-6.2 software for colocalization.

Statistical Analysis.

Statistical comparisons of data sets were performed by two-tailed t test (Student's t test or Mann-Whitney t test) or one-way ANOVA with Tukey posttest using Prizm software (V4.00 GraphPad Inc.). The data were considered to be significantly different when P < 0.05.