Rejection of intradermally injected syngeneic tumor cells from mice by specific elimination of tumor-associated macrophages with liposome-encapsulated dichloromethylene diphosphonate, followed by induction of CD11b⁺/CCR3⁻/Gr-1⁻ cells cytotoxic against the tumor cells

Abstract

Tumor cell expansion relies on nutrient supply, and oxygen limitation is central in controlling neovascularization and tumor spread. Monocytes infiltrate into tumors from the circulation along defined chemotactic gradients, differentiate into tumor-associated macrophages (TAMs), and then accumulate in the hypoxic areas. Elevated TAM density in some regions or overall TAM numbers are correlated with increased tumor angiogenesis and a reduced host survival in the case of various types of tumors. To evaluate the role of TAMs in tumor growth, we here specifically eliminated TAMs by in vivo application of dichloromethylene diphosphonate (DMDP)-containing liposomes to mice bearing various types of tumors (e.g., B16 melanoma, KLN205 squamous cell carcinoma, and 3LL Lewis lung cancer), all of which grew in the dermis of syngeneic mouse skin. When DMDP-liposomes were injected into four spots to surround the tumor on day 0 or 5 after tumor injection and every third day thereafter, both the induction of TAMs and the tumor growth were suppressed in a dose-dependent and injection number-dependent manner; and unexpectedly, the tumor cells were rejected by 12 injections of three times-diluted DMDP-liposomes. The absence of TAMs in turn induced the invasion of inflammatory cells into or around the tumors; and the major population of effector cells cytotoxic against the target tumor cells were CD11b⁺ monocytic macrophages, but not CCR3⁺ eosinophils or Gr-1⁺ neutrophils. These results indicate that both the absence of TAMs and invasion of CD11b⁺ monocytic macrophages resulted in the tumor rejection.

Introduction

Macrophages are highly versatile immune effector cells that are derived from bone marrow progenitors, which continually proliferate and release promonocytes into the bloodstream. These promonocytes circulate briefly, differentiate into monocytes, and then migrate from the blood into tissues, where they differentiate further into resident macrophages. When monocytes are recruited into malignant tumors, they rapidly differentiate into tumor-associated macrophages (TAMs). In this location, they are stimulated by various kinds of cytokines and promote tumor differentiation, invasion, metastasis, and angiogenesis; and, therefore, tumor levels of chemoattractant proteins (e.g., M-CSF and CCL2) often correlate positively with TAM numbers in human tumors such as endometrial [25], breast [5, 17], prostate [3], and ovarian [23] carcinomas. New blood vessels in tumors are usually disorganized and prone to collapse, resulting in areas of inadequate perfusion and hypoxia (low oxygen tension). Additionally, rapid tumor cell proliferation in some areas may outpace the rate of new blood vessel growth, causing hypoxic areas to form [31, 42]. The presence of multiple areas of hypoxia is a hallmark feature of experimental tumors.

In the present study, we intradermally (i.d.) transplanted syngeneic tumor cells and explored the effects of specific removal of TAMs by in vivo application of dichloromethylene diphosphonate (DMDP)-containing liposomes on the tumor growth. The results revealed that the DMDP-liposome treatment suppressed the tumor growth in a dose- and injection number-dependent manner and that 12 injections of three times-diluted DMDP-liposomes resulted in rejection of the tumor. As expected, TAMs were not detected in the tumors; however, quite unexpectedly, an enormous number of cells including F4/80⁺, Ly-6C⁺ or CD11b^low+ macrophages infiltrated into the rejection site: the % cell numbers of these infiltrating cells were CD11b⁺ ≡ CD11a⁺ > F4/80⁺ ≫ Ly6C⁺ ≡ CCR3⁺ ≡ Gr-1⁺ ≫ CD11c⁺ > Ly-6G⁺ > NK-1.1⁺ > CD4⁺ > CD8⁺. The major population of effector cells cytotoxic against the target tumor cells comprised CD11b⁺ monocytic macrophages, but not CCR3⁺ eosinophils or Gr-1⁺ neutrophils. These results suggest that specific elimination of TAMs followed by induction of CD11b⁺ monocytic effector cells might cause the tumor rejection.

Materials and methods

Chemicals

RPMI 1640 medium was obtained from Nissui Seiyaku (Tokyo, Japan). Fetal calf serum (FCS) from Filtron Pty Ltd. (Brooklyn, Australia) was used after heat inactivation. Penicillin and streptomycin were obtained from GIBCO Labs. (Grand Island, NY, USA). OTC compound was purchased from International Equipment Co. (Needham, MA, USA). Paraformaldehyde was obtained from Taab Labs. (Berkshire, UK). Trypan blue, Turk’s solution, and 3-3′ diamino-benzidine tetrahydrochloride were the products of Wako Pure Chemicals (Osaka, Japan). Biotin-conjugated monoclonal antibody (mAb) against mouse F4/80 (clone MCA497B) and a Vectorstain-ABC kit were purchased from Serotec (Oxford, UK) and Vector (Burlingame, CA, USA), respectively. Other Abs used in this study were the following: fluorescein isothiocyanate-labeled anti-mouse Gr-1 (RB6-8C5), CD11c (HL3), and Ly-6G (1A8) Abs, phycoerythrin (PE)-conjugated anti-mouse CD4 (RM4-5), CD8 (53-6.7), CD11a (2D7), CD11b (M1/70), and NK-1.1 (PK136) Abs, and purified anti-mouse CD16/CD32 (2.4G2) Ab, all from PharMingen (San Diego, CA, USA). PE-labeled anti-mouse CCR3 (83101), F4/80 (A3-1), and Ly-6C (HK1.4) Abs were the products of R&D Systems (Minneapolis, MN, USA), Serotec, and Beckman Coulter (Fullerton, CA, USA), respectively. All other chemicals were of reagent grade.

Animals

Specific pathogen-free male C57BL/6 mice and DBA/2 mice (7 weeks of age) were purchased from Japan SLC (Hamamatsu, Japan). All experiments were carried out in accordance with The Guidelines on Animal Experiments of Osaka Medical College and the Japanese Government Notification on Feeding and Safekeeping of Animals (Notification No.6 of the Prime Minister’s Office); and the experimental protocol was approved by The Review Committee for Animal Experiments of Osaka Medical College.

Tumor cells

3LL cells (Lewis lung carcinoma cell line of C57BL/6 origin), B16 cells (melanoma cell line of C57BL/6 origin), and KLN205 cells (squamous cell carcinoma cell line of DBA/2 origin) were purchased from American Type Culture Collection (Manassas, VA, USA). The tumor cells were maintained by in vitro culturing in RPMI 1640 medium supplemented with 10% FCS and antibiotics (100 U/ml of penicillin and 100 μg/ml of streptomycin).

Preparation of DMDP-containing liposomes

Macrophages ingest liposomes but not small molecules including DMDP [28]; and therefore, DMDP can be ingested by macrophages when this compound is given encapsulated in liposomes, but not when given free in solution. Multilamellar liposomes containing either the cytolytic agent DMDP or phosphate-buffered saline (PBS) were prepared as described previously [22, 38], according to the method of Van Rooijen et al. [40], who showed that the i.v. injection of DMDP encapsulated in liposomes (“DMDP-liposomes”) almost completely depleted macrophages from spleen and liver, but not those in other organs including the bone marrow. Similarly, subcutaneous or intratracheal administration of the liposome-encapsulated DMDP eliminates macrophages in draining lymph nodes [6] or from lung tissues [37], respectively. The DMDP-liposomes are ingested by macrophages via endocytosis; and after fusion with lysosomes containing phospholipases the latter disrupt the bilayers of the liposomes. Therefore, the cytotoxic effect of DMDP-liposomes is specific for macrophages, because monocytes are not active phagocytes and because granulocytes do not have phospholipases to disrupt the bilayers of the liposomes [6, 37, 38, 40, 41]. The more concentric bilayers that are disrupted, the greater, the DMDP release within the cell, revealing that DMDP-liposomes specifically and dose-dependently eliminate macrophages migrating from neighboring tissues. In the present study, the original suspension or three to ten times-diluted suspension was i.d. injected into four spots (50 μl/spot) in the dorsal region to surround the tumors (total of 0.2 ml/mouse); and the distance between the sites where the tumor cells and DMDP-liposomes were injected was 6~7 mm. The DMDP is membrane impermeable and has an extremely short half life (a few minutes) in circulation and body fluids [10]. Therefore, it is unlikely that the locally injected DMDP-liposomes affect directly the survival of tumor cells.

In vivo measurement of tumor growth

3LL tumor cells, KLN205 tumor cells or B16 tumor cells (each 5 × 10⁵ cells/50 μl PBS/mouse) were i.d. injected into the dorsal region of a syngeneic strain of mouse. At appropriate intervals after the treatment, the tumor growth was determined by measuring two diameters (perpendicular to each other) of the tumor with Vernier calipers [44]. The surface area of the tumor was previously shown to positively correlate precisely with the tumor weight [12].

Immunohistochemistry

Specimens were dissected and fixed with 4% paraformaldehyde in PBS. After having been washed sequentially with 5, 15, and 29% sucrose-PBS, the specimens were embedded in OCT compound and quick-frozen in a mixture of acetone and dry ice. Frozen sections (8-μm thickness) were cut, placed on poly-l-lysine-coated glass slides, and air-dried. One section was stained with hematoxylin–eosin, and the others were used for immunohistochemical analyses. A two-step immunoperoxidase method [7] was applied: After incubation in 0.3% H₂O₂ for 30 min, the sections were incubated with a biotin-conjugated mAb in a solution containing 5% normal goat serum in PBS, 5% bovine serum albumin, and 0.025% Triton X-100. Biotin-bound sections were then reacted with avidin–peroxidase complex for 30 min and visualized after incubation with 0.06% 3-3′ diaminobenzidine tetrahydrochloride and 0.03% H₂O₂ in 0.1 M Tris–HCl buffer at pH 7.6. Nuclei were counterstained with methyl green. As control experiments, normal rat serum or PBS was applied instead of the primary Ab.

Isolation of cells infiltrating into the transplantation site of tumor cells

Infiltrates into or around transplanted tumor cells were recovered by the method of Krist et al. [15] with some modifications [45]. At various intervals after transplantation of the cells and PBS- or DMDP-liposome treatment, the tumor and the surrounding tissues of recipients were removed en block and kept on ice. The harvested tissue was cut into small blocks (≈ 1 mm × 1 mm) with scissors. The blocks in a siliconized (Surfa Sil™; Pierce, Rockford, IL, USA) glass tube were digested at 37°C with continuous shaking in a water bath. The digestion medium was RPMI 1640 medium containing 0.15% protease (type IV; Sigma), 0.075% collagenase (Wako Pure Chemicals), and 0.001% DNase I (Boehringer-Mannheim, GmbH, Germany). After a 20-min incubation, the blocks were allowed to settle, and the supernatant was then passed through a 67-μm nylon mesh. The cells in the siliconized glass tube were washed in RPMI 1640 medium containing 10% FCS, suspended in Ca²⁺, Mg²⁺-free PBS, and kept on ice. The residual blocks were treated with the digestion medium as described above, and the enzyme treatment was repeated totally three times. All the digested cells were centrifuged and resuspended in PBS containing 2% FCS. The infiltrating leukocytes (<20 μm in a diameter) that had infiltrated into the tumor cells were separated from the tumor cells (>25 μm in a diameter) by using a fluorescence-activated cell sorter (FACS; FACSAria^®, Becton Dickinson, Mountain View, CA, USA) in a mode of forward scattering/side scattering and used as a source of effector cells for cytotoxic activity assays. The protease/collagenase/DNase treatment did not affect either CD4, CD8, and CD3 molecules of T cells or the MHC class I molecules of recovered cells. Neither did the treatment invalidate the cytotoxic activity of allografted Meth A tumor-induced CTL against BALB/c lymphoblasts [before: 65.0 ± 7.8% specific lysis in a 4-h incubation with an effector/target (E/T) ratio of 20 (n = 4); after a 20-min enzyme treatment of CTL: 64.3 ± 5.8%; after a 40-min enzyme treatment: 66.7 ± 8.5%; after a 60-min enzyme treatment: 69.7 ± 7.8%].

Cell number and viability

Cell number was determined by counting the cells in Turk’s solution with a hemocytometer. The viability of cells was assessed by the trypan blue exclusion method.

Flow cytometric analysis

Cells (1–5 × 10⁵ cells) suspended in 25 μl of cold PBS containing 2% FCS were stained with a fluorescein-labeled Ab at 4°C for 20 min in the presence (for staining) or absence (for sorting) of 0.1% sodium azide, washed, and analyzed by FACS, as described previously [48].

Cytotoxicity assay

B16 melanoma cells were used as target cells. The target cells were labeled with ⁵¹Cr (Na⁵¹₂CrO₄; Dupont New England Nuclear, Boston, MA, USA) for 2 h at 37°C in 5% CO₂. Effector cells (5 × 10³ to 2 × 10⁵ cells) in 200 μl of culture medium were mixed with 20 μl of ⁵¹Cr-labeled targets (10³ to 5 × 10³ cells) in 96-well U-bottomed microtest plates (3799; Corning Incorporated, Corning, NY, USA). After a 12-h incubation at 37°C in 5% CO₂, the amount of released ⁵¹Cr in an aliquot (100 μl) of supernatant was measured by a gamma counter (Cobra 5550; Hewlett-Packard, Meriden, CT, USA); and percentage of specific lysis was calculated as described previously [1].

Transmission electron microscopy

Cells were fixed for 1 h at 4°C in 2% glutaraldehyde solution and then postfixed in 1% osmium tetroxide. The cells were dehydrated through a series of graded alcohols (10 min for each alcohol) and then permeated with propylene oxide. The specimens were next embedded in epoxy resin, sectioned with a Reichert-Nissei ultramicrotome (Reichert-Nissei, Vienna, Austria), stained with 4% uranyl acetate, and treated with Reynolds’ lead acetate, as described previously [43]. Finally they were examined with a Hitachi H-7650 electron microscope (Hitachi Ltd., Tokyo) at 80 kV.

Results

Growth of tumor cells in the dermis of skin

When B16 melanoma cells, KLN205 squamous cell carcinoma cells or 3LL Lewis lung cancer cells (each 5 × 10⁵ cells/50 μl PBS/mouse) were transplanted into the dermis of syngeneic mouse skin, all of them continued to grow there with a similar growth rate and killed the hosts approx. 4 weeks after the injection (Fig. 1).

Fig. 1.

Growth of i.d. transplanted tumor cells (each 5 × 10⁵ cells/mouse). B16 melanoma cells (C57BL/6 origin), KLN205 squamous cell carcinoma cells (DBA/2 origin) or 3LL Lewis lung cancer cells (C57BL/6 origin) were transplanted into the dorsal dermis of syngeneic mice. At appropriate intervals after tumor transplantation, the size of the tumor was assessed. Each value for the tumor size represents the mean ± SD of four mice. Open circles indicate 3LL cells; open triangles indicate B16 cells; open squares indicate KLN205 cells

Inhibition of tumor growth by specific elimination of macrophages infiltrating into the tumors

To eliminate macrophages infiltrating into the tumors, we i.d. injected a PBS solution of DMDP- or PBS-liposomes (0.2 ml/mouse) into four spots (50 μl/spot) in the dorsal region of the skin to surround the tumor on days 0 and 2 after tumor (5 × 10⁵ cells/50 μl PBS/mouse) injection (Fig. 2). The distance between the sites where the tumor cells and DMDP- or PBS-liposomes were injected was 6~7 mm (Fig. 2a); and the free DMDP is membrane impermeable and has an extremely short half life (a few minutes) in circulation and body fluids [10]. In fact, there was no damage of tissues (e.g., hair follicles and mucosal muscle) near the injection sites; and the tumor cells were clearly out of contact with the DMDP-liposomes (Fig. 2a), revealing no direct toxic effects of the DMDP-liposomes on the tumor cells. Surprisingly, however, the tumor growth was significantly inhibited for several days after specific elimination of macrophages by DMDP-liposomes, whereas PBS-liposomes were almost ineffective. Among the three kinds of tumor cells so far tested, the B16 melanoma cells appeared to be the most resistant to the liposomal treatment (Fig. 2b–d).

Fig. 2.

Inhibition of tumor growth by specific elimination of macrophages infiltrating into the tumors. For elimination of macrophages infiltrating into the tumors, undiluted DMDP- or PBS-liposome solution (0.2 ml/mouse) was i.d. injected into four spots in the dorsal region of the skin to surround the tumor on days 0 and 2 after tumor (5 × 10⁵ cells/mouse) injection. a Histological features of skin on day 6. Arrow represents tumor injection site. Arrow head represents liposome injection site. Scale bar = 250 μm. b 3LL lung cancer cells; c B16 melanoma cells; d KLN205 squamous carcinoma cells. Open circles indicate PBS-liposomes; closed circles indicate DMDP-liposomes. Each value represents the mean ± SD of five mice. The difference between sample (DMDP-liposomes) and control (PBS-liposomes) is significant (*P < 0.05; **P < 0.02; ***P < 0.01; ****P < 0.005; *****P < 0.002; ******P < 0.001) according to Student’s t-test. n.s. not significant

When we i.d. injected a ten times-diluted suspension of DMDP- or PBS-liposomes (each 0.2 ml/mouse) on day 0 and every third day thereafter (total of six injections) to surround B16 melanoma cells (5 × 10⁵ cells/mouse), the DMDP-liposome treatment significantly suppressed the tumor growth (Fig. 3a). Treatment of tumor bearing mice with five times-diluted suspension of DMDP-liposomes (total of six injections) was more effective; and unexpectedly, 4 of 12 treated mice rejected the tumor cells (Fig. 3b). In contrast, the PBS-liposome treatment had no effect on the tumor growth. In the case of five times-diluted DMDP-liposomes (total of 12 injections), the DMDP-liposome treatment was much more effective; and the tumor cells were rejected from three of five DMDP-liposome-treated animals (Fig. 3c). Finally, after 12 injections of three times-diluted DMDP-liposomes all of the treated animals rejected the tumor cells, whereas 12 injections of three times-diluted PBS-liposomes had no effect on the tumor growth (Fig. 3d).

Fig. 3.

Tumor rejection by specific elimination of macrophages infiltrating into B16 melanoma tumors. B16 melanoma cells (5 × 10⁵ cells/mouse) were i.d. transplanted. On day 0 and every third day thereafter, three to ten times-diluted DMDP- or PBS-liposome solution was i.d. injected into four spots to surround the tumor. a six injections of ten times-diluted DMDP- or PBS-liposome solution into mice. Open circles indicate PBS-liposomes; closed circles indicate DMDP-liposomes (each n = 12). b six injections of five times-diluted DMDP- or PBS-liposome solution into mice. Open circles indicate PBS-liposomes (n = 3); closed symbols indicate DMDP-liposomes (circles, n = 8; triangles, n = 4). The difference between circles and triangles appeared to be due to the difference in the fine movement of i.d. injected tumor cells to the site(s) suitable for their growing, since the five times-diluted DMDP-liposomes almost completely suppressed the growth of B16 melanoma tumors that had grown to some extent in all of five mice (see Fig. 4a). c 12 injections of five times-diluted DMDP- or PBS-liposome solution into mice. Open circles indicate PBS-liposomes (n = 3); closed symbols indicate DMDP-liposomes (circles, n = 2; triangles, n = 3). d 12 injections of three times-diluted DMDP- or PBS-liposome solution into mice. Open circles indicate PBS-liposomes (n = 3); closed triangles indicate DMDP-liposomes (n = 8). Each value represents the mean ± SD of 3–12 mice. In the case of closed circles in experiment “C,” each value represents the mean of two mice

Growth inhibition of melanoma tumors that had grown to some extent by specific elimination of macrophages

Next, we examined the effect of specific elimination of macrophages by DMDP-liposome treatment on the growth of B16 melanoma tumors that had grown to some extent before macrophage elimination (Fig. 4). We made six i.d. injections of five times- or ten times-diluted DMDP- or PBS-liposome suspension (each 0.2 ml/mouse) on day 5 and every third day thereafter to surround tumors that had been transplanted on day 0. The average size of the tumors on day 5 after tumor (5 × 10⁵ cells/mouse) transplantation was 2.6 mm × 2.6 mm (n = 20). Six injections of five times-, but not of ten times-, diluted DMDP-liposome suspension were sufficient to obtain almost complete suppression of tumor growth as long as we continued the liposomal treatment (Fig. 4a). After 12 i.d. injections of three times-diluted DMDP-liposomes on day 5 and every third day thereafter to surround tumors, the tumor growth was suppressed for several days; and the tumor cells were rejected from all of the DMDP-liposome-treated mice (Fig. 4b). In contrast, 12 injections of three times-diluted PBS-liposomes had no effect on the tumor growth. These results suggest that the tumor growth was suppressed by DMDP-liposome treatment in a dose-dependent and injection number-dependent manner and that the tumor cells were rejected possibly by mechanisms other than specific elimination of TAM by the DMDP-liposome treatment.

Fig. 4.

Growth inhibition of B16 melanoma tumors that had grown to some extent prior to specific elimination of macrophages. B16 melanoma cells (5 × 10⁵ cells/mouse) were i.d. injected into C57BL/6 mice. a On day 5 and every third day thereafter (total of six injections) after the tumor injection, five or ten times-diluted DMDP- or PBS-liposome suspension (each 0.2 ml/mouse) was injected into five mice to surround their tumors. Open circles indicate ten times-diluted PBS-liposomes; open triangles indicate five times-diluted PBS-liposomes; closed circles indicate ten times-diluted DMDP-liposomes; closed triangles indicate five times-diluted DMDP-liposomes. b 12 injections of three times-diluted DMDP (closed triangles)- or PBS (open circles)-liposome suspension into five mice. Each value represents the mean ± SD of five mice. The difference between sample (DMDP-liposomes) and control (PBS-liposomes) is significant (*P < 0.05; **P < 0.01; ***P < 0.005; ****P < 0.002; *****P < 0.001) according to Student’s t-test. n.s. not significant

Immunohistochemical examination of PBS- and DMDP-liposome-treated skin

The results mentioned above imply the possibility that specific elimination of macrophages infiltrating into i.d. transplanted tumor cells not only suppressed the tumor growth but also induced cytotoxic activity against the tumor cells. To test this possibility, we immunohistochemically examined the morphological features of the skin on day 6 after PBS- or DMDP-liposome treatment on days 0 and 3. Figure 5a shows clear evidence of tumor formation in the PBS-liposome-treated skin; Fig. 5b partial tumor degeneration with a marked cell infiltration into or around tumors in the DMDP-liposome-treated skin. There was consistent positive staining with anti-F4/80 Ab in the tumors [188.7 ± 25.8 cells (mean ± SE; 17 fields)] of PBS-liposome-treated mice (Fig. 5c), indicating infiltration of macrophages (i.e., TAMs) into the tumors. As expected, however, there was a dose of DMDP-liposome-dependent decrease [100.6 ± 8.0 cells (mean ± SE; 47 fields), 38.4 ± 6.4 cells (mean ± SE; 21 fields) and 48.5 ± 4.5 cells (mean ± SE; 32 fields)] in the number of macrophages into the tumors of ten, five, and three times-diluted DMDP-liposome-treated mice, respectively; and macrophages rather infiltrated around tumors [117.5 ± 12.3 cells (mean ± SE; 32 fields)] (Fig. 5d). In contrast, Fig. 5e, f also shows that a few CD3⁺ cells infiltrated into or around the tumors after treatment with PBS-liposomes [8 cells; 1 field] or DMDP-liposomes [5 cells; 1 field], although T cells in the epidermis were clearly stained (inset). Instead, there was massive infiltration of Gr-1⁺ cells [in tumors: 21.7 ± 2.3 cells (mean ± SD; 3 fields); around tumors: 137.3 ± 11.7 cells (mean ± SD; 3 fields)] in the rejection site (Fig. 5h). The infiltration of Gr-1⁺ cells after PBS-liposome treatment was almost negligible in and around tumors [6.5 ± 2.1 cells (mean ± SD; n = 3)] (Fig. 5g), revealing infiltration of the F4/80⁺ or Gr-1⁺ cells mainly around the tumors of DMDP-liposome-treated mice.

Fig. 5.

Immunohistochemical examinations of PBS-liposome- and DMDP-liposome-treated skin. B16 melanoma cells (5 × 10⁵ cells/mouse) were i.d. injected into C57BL/6 mice (day 0). On days 0 and 3 after tumor injection, three times-diluted PBS- or DMDP-liposome suspension (each 0.2 ml/mouse) was injected to surround the tumors. On day 6, the dorsal skin was excised to examine the histological features of the skin. a, c, e, and g PBS-liposome treatment. b, d, f, and h DMDP-liposome treatment. c and d staining with anti-F4/80 Ab. e and f staining with anti-CD3 Ab. Insets, T cells in the epidermis. g and h staining with anti-Gr-1 Ab

Time-dependent changes in leukocyte infiltration into DMDP-liposome-treated skin and in the cytotoxic activity against target tumor cells

Figure 5 showed that tumor growth was suppressed by specific elimination of TAMs; and in addition, Figs. 3 and 4 implied the possibility that leukocytes cytotoxic against tumor cells might be induced in the transplantation site. In particular, the results shown in Figs. 3 and 4 suggest that cells cytotoxic against the tumor cells might have infiltrated around day 8 after DMDP-liposome treatment. To test this possibility, we determined the time-dependent changes in the cytotoxic activities of bulk infiltrates against B16 melanoma cells. As shown in Fig. 6, the leukocytes infiltrated into or around tumors and showed increased cytotoxicity against B16 melanoma cells with time after DMDP-liposome treatment. The cytotoxic activity reached a peak on day 8 [i.e., on day 8 after simultaneous injection with tumor (Fig. 6a); and on day 13 in the case of the first DMDP-liposome treatment on day 5 after tumor transplantation (Fig. 6b)], and decreased to the original level on day 11.

Fig. 6.

Induction of cytotoxic activity by DMDP-liposome treatment. B16 melanoma cells (5 × 10⁵ cells/mouse) were i.d. injected into C57BL/6 mice. a On day 0, 3, 6 or 9 after tumor injection, three times-diluted DMDP-liposome suspension (each 0.2 ml/mouse) was injected into 15 mice (5 mice/group) to surround their tumors. b On day 5, 8, 11 or 14 after tumor injection (day 0), three times-diluted DMDP-liposome suspension (each 0.2 ml/mouse) was injected into 20 mice (5 mice/group) to surround their tumors. Open circles indicate cytotoxic activities against the tumor cells. Each value represents the mean ± SD of 12 cultures from three different experiments. Closed circles indicate numbers of infiltrating cells. Each value represents the mean of three different experiments. Arrow represents first DMDP-liposome injection. Arrow head represents tumor injection

CD11b^low+ monocytic macrophage as a candidate of effector cells cytotoxic against tumor cells

To identify the type of leukocytes cytotoxic against tumor cells, we i.d. injected three times-diluted suspension of DMDP-liposomes on days 0, 3, and 6 to surround B16 melanoma cells, explored the phenotypes of the infiltrating cells on day 8 by flow cytometry, isolated the antigen⁺ cells, and determined the cytotoxic activity against the tumor cells (Table 1). The major population in the infiltrates was CD11b⁺ cells (67.6%) including CD11b^high+ cells (57.7%), CD11b^low+ cells (9.9%), F4/80⁺ cells (56.2%), CCR3⁺ cells (39.6%), Ly-6C⁺ cells (33.9%), Gr-1⁺ cells (29.0%), CD11c⁺ cells (6.1%), and Ly-6G⁺ cells (5.3%). The cytotoxic activity was mainly recovered in the CD11b^low+, but not CD8⁺, NK-1.1⁺, CD11b^high+, F4/80⁺, Ly-6C⁺, Gr-1⁺, CCR3⁺ or Ly-6G⁺, cell fraction; and they appeared to express CD11a on their surface. Although the CD11a⁺ cells mainly consisted of two populations, monocytic macrophages and eosinophils (Fig. 7), the CCR3⁺ eosinophils in the FACS-purified CD11a⁺ cells had not yet degranulated (Fig. 7) but were inactive toward B16 melanoma cells (Table 1).

Table 1.

Identification of effector cells cytotoxic against B16 melanoma cells

Antigen	% Cell numbera	% Specific lysisb
Bulk	100	10.6 ± 2.8
Gr-1+	29.0	2.5 ± 1.3
CCR3+	39.6	0.9 ± 0.6
CD4+	1.9	n.d.
CD8+	1.0	1.1 ± 0.9c
Ly-6C+	33.9	1.3 ± 1.5
Ly-6G+	5.3	0.2 ± 0.2
NK-1.1	2.8	2.9 ± 0.4d
F4/80	56.2	0.9 ± 1.2
CD11blow+	9.9	9.1 ± 2.5
CD11bhigh+	57.7	1.1 ± 2.8
CD11a	74.4	12.0 ± 2.2
CD11c	6.1	n.d.

n.d. not determined

^aThe phenotypes of infiltrates were analyzed by FACS. Each value represents the mean of three different experiments

^bTwelve-hour assay at an E/T ratio of 40

^cTwelve-hour assay at an E/T ratio of 5

^dTwelve-hour assay at an E/T ratio of 10

Fig. 7.

Electron microscopic examination of CD11a⁺ cells. B16 melanoma cells (5 × 10⁵ cells/mouse) were i.d. injected into C57BL/6 mice. On days 0, 3, and 6 after tumor injection, three times-diluted DMDP-liposome suspension (each 0.2 ml/mouse) was injected to surround the tumors. On day 8, whole infiltrates were stained with PE-labeled anti-mouse CD11a Ab and the CD11a⁺ cells were isolated by FACS. a Two monocytic cells and two polymorphonuclear cells positive for CD11a are seen. Bar, 5 μm. b CD11a⁺ cells in another area. Bar, 5 μm. c A higher power view of the area surrounded by the square in b shows the cell to be an eosinophil with typical intracellular granules. Bar, 2 μm

Discussion

Multiple injections of DMDP-liposomes to surround i.d. transplanted tumor cells (e.g., 3LL Lewis lung carcinoma cells, B16 melanoma cells, and KLN205 squamous carcinoma cells) not only inhibited the growth of tumor cells (Figs. 1, 2) but also induced tumor rejection (Figs. 3, 4). We assumed that in addition to specific elimination of TAMs from the tumor transplantation site, cytotoxic cells or molecules might have been induced in the tumor transplantation site. In fact, there were very few macrophages in the tumors; whereas there was marked invasion of inflammatory cells into or around the tumors (Fig. 5). At these inflammatory sites, CD11b⁺ cells were the major infiltrating cells; and CD11b^low+, but not CCR3⁺ or Gr-1⁺, cells were cytotoxic against the tumor cells (Fig. 6; Table 1). The CD11b^low+ cells also might have migrated from peripheral blood into skin to become tissue macrophages; but surprisingly, electron microscopic examinations showed them to be monocytic or non-phagocytic (Fig. 7), resulting in specific elimination of TAMs and infiltration of CD11b^low+ monocytic macrophages by DMDP-liposome treatment. These results taken together suggest that DMDP-liposomes induced specific elimination of macrophages and that the absence of these so-called TAMs, which promote tumor differentiation, invasion, metastasis, and angiogenesis [3, 17, 23, 31, 42], in turn caused inhibition of the tumor growth on the one hand and prompted CD11b⁺ monocytic macrophage infiltration to kill tumor cells on the other hand.

Over the last decade, cytokine gene transfer strategies in animal models have provided a tool with which to dramatically increase intratumoral cytokine availability, avoid the side effects of systemic administrations, and evaluate the antineoplastic potential of locally recruited inflammatory cells. Almost all the cytokines sustainedly released by engineered tumor cells, namely, IL-1, IL-2, IL-3, IL-4, IL-7, IL-10, IL-12, IFN-α, IFN-β, IFN-γ, G-CSF, and TNF-α [4, 24], quickly recruit a massive local reaction that leads to the rejection of engineered tumor cells and the establishment of a significant immunity against the wild-type parental tumor. Eosinophils are commonly found in infiltrates of many different cancers, and controversy exists as to whether they are passive bystanders or active cellular agents in host immune responses. Recently, however, it is assumed that Gr-1⁺ neutrophils play a key role in all of these cytokine-induced tumor rejections, often in cooperation with CD8⁺ T lymphocytes [21, 24]. In the present study, we revealed that the growth of non-engineered or natural tumor cells (e.g., B16 melanoma cells, 3LL Lewis lung cancer cells, and KLN205 squamous cell carcinoma) was inhibited by specific elimination of TAMs after multiple injections of DMDP-liposomes to surround i.d. transplanted tumor cells (Figs. 1, 2, 3, 4, 5) and that the tumor cells were rejected by the direct cytotoxic activity of CD11b⁺ monocytic macrophages, but not of CCR3⁺ eosinophils or Gr-1⁺ neutrophils, against the tumor cells (Fig. 6; Table 1). Among the cell types in the infiltrates, CD11a⁺ cells displayed higher cytotoxic activity against the tumor cells than CD11b^low+ cells (Table 1); and the CD11a⁺ population mainly contained monocytic macrophages and eosinophils (Fig. 7). Furthermore, preliminary experiments from our laboratory demonstrated that the antitumor effect of DMDP-liposomes was reduced by simultaneous injections of eotaxin receptor antagonists (e.g., YM344031 [32] and YM355179 [20]), suggesting that the cytotoxic activity might be mediated by CD11b^low+ monocytic macrophages with the help of eosinophils. In contrast, since the %cell number of Ly-6G⁺ neutrophils (5.3%) in inflammatory cells was much less than that (33.9%) of Ly-6C⁺ cells, Gr-1⁺ (=Ly-6G⁺ + Ly-6C⁺) cells in Fig. 5h appeared to be Ly-6C⁺ cells.

Since the first observation of malignancy with marked blood eosinophilia described by Rheinbach in 1893, eosinophilia has been described in human cancers from a variety of organs [13, 26, 27]. Most of these studies have focused on whether the presence of a prominent eosinophilic infiltrate has a prognostic value or is an indicator of the response to treatment. Some authors have claimed that the presence of a marked or moderate eosinophilia is associated with a poor prognosis [16, 27, 39], whereas others have found eosinophilia to be a favorable prognostic feature [2, 9, 11, 14, 29]. In most patients, eosinophilia is asymptomatic [8]; and tumor-related peripheral eosinophilia rarely contributes to increased mortality except in unusual cases, such as that associated with endomyocardial fibrosis [36]. In fact, CD11a⁺ cells mainly consisting of CD11b^low+ cells and CCR3⁺ eosinophils (Fig. 7) displayed cytotoxic activity against B16 melanoma cells higher than CD11b^low+ cells alone (Table 1). Furthermore, our preliminary experiments revealed that specific induction of eosinophils around B16 melanoma cells by eotaxin injection clearly enhanced the tumor growth. Therefore, it is conceivable that eosinophil infiltration in the presence of TAMs is associated with a poor prognosis, whereas the infiltration in the presence of monocytic macrophages is a favorable prognostic feature.

Monocytes and macrophages are the most common cell populations of the innate immune system in mammals. Their role in tumor immunity is complex, because they either enhance or impair immune responses. In the case of allogeneic Meth A tumor rejection from C57BL/6 mice, two types of monocytic macrophages [allograft-induced macrophage-1 (AIM-1) and AIM-2] infiltrate into the tumors on days 4–14 after allografting and reject them. AIM-1, which expresses Ly-6C on its surface [33], is cytotoxic against target tumor cells through recognition of a major histocompatibility antigen complex on the allograft [34, 35]. AIM-2, which expresses CD11a, CD11b, and CD18 on its surface [49], is cytotoxic against allogeneic tumor cells [38, 45–48]. Unexpectedly, AIM-2, which is inducible by allografting, is cytotoxic against various kinds of syngeneic tumor cells [46, 50], indicating that the role of these monocytic macrophages in allogeneic tumor transplantation enhances immunity. On the other hand, macrophages respond to the levels of hypoxia found in syngeneic tumors by up-regulating such transcription factors as hypoxia-inducible factors 1 and 2, which in turn activate a broad array of mitogenic, proinvasive, proangiogenic, and prometastatic genes [18]. This could explain why high numbers of TAMs correlate with poor prognosis in various forms of cancer [3, 17, 23, 31, 42]. In the present study, we demonstrated that specific elimination of TAMs by multiple injections of DMDP-liposomes to surround i.d. transplanted syngeneic tumor cells clearly inhibited the tumor growth, revealing that the role of TAMs in cancer patients mainly impairs immunity. Of particular interest, the absence of TAMs in turn prompted the infiltration of two types of monocytic macrophages into or around tumors: as expected, Ly-6C⁺ AIM-1 was inactive to the syngeneic tumors, whereas CD11b^low+/CD11a⁺ AIM-2 was cytotoxic to reject them (Table 1). These results taken together indicate that these two types of monocytic macrophages infiltrate into allografts or tumors to reject them.

To obtain therapeutic effects of specific elimination of TAMs without any side effects, we i.d. injected DMDP-liposomes to surround i.d. transplanted tumor cells. A systemic (e.g., i.v.) injection of DMDP-liposomes, however, may be applied for tumors in other organs. DMDP-liposomes may be recruited from the tumor vasculature with monocytes by such chemoattractants released by tumors as M-CSF and CCL2 [30], whereas they are also ingested by spleen or liver macrophages [19, 41]. In the present study, we showed that the absence of TAMs in turn caused inhibition of the tumor growth on the one hand (Figs. 1, 2, 3, 4, 5) and prompted CD11b^low+ monocytic macrophage infiltration to kill tumor cells on the other hand (Figs. 6, 7; Table 1). Therefore, it may be crucial that targeting of liposomes to a macrophage subpopulation (e.g., TAM) is achieved by the insertion of target molecules (e.g., mAb specific toward TAMs but not toward AIM-2 or spleen or liver macrophages) in their outer (surface) bilayer. The specific elimination of TAMs may lead to intriguing and novel strategies to enable effective biologic therapy for cancer patients.

Abbreviations

AIM

Allograft-induced macrophage

DMDP

Dichloromethylene diphosphonate

E/T

Effector/target

FACS

Fluorescence-activated cell sorter

FCS

Fetal calf serum

i.d.

Intradermally

mAb

Monoclonal antibody

PBS

Phosphate-buffered saline

PE

Phycoerythrin

TAMs

Tumor-associated macrophages