Nonresolving inflammation is a hallmark of many types of tumors and the molecular mechanisms maintaining this inflammation are still largely unknown. In a two-stage carcinogenesis model, we observed here that the lack of IFN-γ receptor or neutralization of IFN-γ accelerated spontaneous papilloma regression in mice. The impaired maintenance of local inflammation was associated with reduced IFN-γ and enhanced biosynthesis of proresolution lipid mediator lipoxin A4 (LXA4). Interestingly, blocking LXA4 eliminated the effect of anti-IFN-γ, whereas treatment of mice with a therapeutic dose of LXA4 accelerated papilloma regression in an IFN-γ–independent manner. These results link for the first time a cytokine-dependent maintenance of inflammation with a downregulated production of proresolution lipid mediators. Strategies promoting spontaneous resolution of chronic inflammation by blocking IFN-γ and/or increasing LXA4 may be useful for the treatment of inflammation-associated tumors. Cancer Res; 73(6); 1742–51. ©2012 AACR.

Persistent inflammation failing to resolve is an inherent risk in tumor pathogenesis (1). Inflammatory cells and soluble mediators such as cytokines and chemokines significantly contribute to an environment aiding tumor-cell survival and proliferation (2). Recently, tumor-related inflammation has become a target in tumor therapy (3, 4). In clinical epidemiologic studies, nonsteroidal anti-inflammatory drugs such as aspirin are found to prevent tumor occurrence and recurrence. This subsequently reduces the risk of inflammation-related tumors bearing a specific risk of malignant transformation, such as colorectal adenomas (5, 6). Determining key modulators in inflammation resolution and understanding their role during tumorigenesis will help to define new therapeutic targets.

Papilloma development in mice after application of 7,12-dimethylbenz(a)anthracene (DMBA) and subsequently repeated 12-O-tetradecanoylphorbol-13-acetate (TPA) is a well-characterized animal model for analyzing the effect of inflammation on tumor development (7). Interestingly, without the continuous support of TPA-induced inflammation, a good percentage of papillomas can regress spontaneously, but so far little is known about the underlying molecular mechanisms of inflammation resolution and spontaneous papilloma regression (8, 9).

Inflammation resolution has been recognized as an active process involving lipid mediators such as lipoxin A4 (LXA4; ref. 10). LXA4 can reduce and stop neutrophil infiltration in inflamed tissues, stimulate nonphlogistic recruitment and phagocytosis of macrophages, inhibit inflammatory molecule production by leukocytes and fibroblasts, restore vascular permeability to homeostasis, and limit endothelial cell migration (11, 12). The conversion of arachidonic acid to LXA4 is catalyzed by transcellular pathways involving sequential action of lipoxygenases (LOX), such as 15- and 5-LOX mainly in human tissues, whereas 5- and 12-LOX during leukocyte–platelet interaction (13). Investigating the regulatory relationship between LXA4 and inflammatory cytokines can help us to understand the inflammation resolution.

Many studies have shown that LXA4 or its analogs can suppress the expression of IFN-γ or IFN-γ–induced genes in vitro and in vivo (14–17). On the other hand, almost 20 years ago, it has been reported that cytokine interleukin (IL)-4 or -13 mediates upregulation of 15-LOX expression in human blood monocytes in vitro and IFN-γ inhibits this process (18, 19). However, until now it is still not clear how proinflammatory cytokines interact with the proresolution lipid mediators in vivo.

We previously showed that IFN-γ promoted papilloma development via upregulation of local IL-17 and inflammatory response in the DMBA/TPA model (20). However, the role of IFN-γ during inflammation resolution and spontaneous papilloma regression is still to be investigated. Here, we specifically asked how IFN-γ acts during inflammation resolution and spontaneous papilloma regression. We report that accelerated papilloma regression and inflammation resolution after neutralization of IFN-γ were related to enhanced generation of proresolving lipid mediator LXA4. Inhibiting LXA4 eliminated the effect of IFN-γ neutralization on papilloma regression. In accordance, LXA4 treatment attenuated the effect of IFN-γ in the maintenance of papillomas and nonresolving inflammation. These results shed new light on understanding the dynamic balance between proinflammatory factors and proresolution mediators and may help to develop novel strategies for the treatment of inflammation related diseases, such as cancer.

Mice

Wild-type 129/Sv/Ev (WT) and IFN-γ receptor knockout (IFN-γR-KO) mice on the 129/Sv/Ev background were inbred strains purchased from The Jackson Laboratory. All mice were bred in a specific pathogen-free environment at the Institute of Biophysics, Chinese Academy of Sciences (Beijing, China). Sex- and age-matched mice (6- to 8-week old) were used. Animal experiments were carried out with the approval of the Institutional Laboratory Animal Care and Use Committee.

Skin tumorigenesis

Papillomas and local inflammation were induced as described before (20). In brief, an area of 200 mm2 dorsal skin was shaved in every mouse and 25 μg DMBA (Sigma) in 100 μL acetone was administered to the shaved skin. One week after single DMBA application, 4 μg TPA (Sigma) in 100 μL acetone was painted to this area twice a week for 14 weeks. Lesions with a minimum diameter of 1 mm for at least 2 weeks were defined as a papilloma. The size of a papilloma was determined by its length × width in mm2. The sum of sizes of all single papillomas was recorded as the total area of papillomas in a mouse (21). At different time points after TPA withdrawal, the papilloma number and total area of each mouse was compared with that at week 0. The mice whose papilloma number or total area was getting smaller for at least 2 weeks were recognized as regressors and others as nonregressors. Percentage reduction of papilloma number at every time point = [(papilloma number at each week – papilloma number at week 0)/papilloma number at week 0] × 100%. Percentage shrinkage of papilloma area at every time point = [(papilloma area at each week – papilloma area at week 0)/papilloma area at week 0] × 100%.

Treatment regimen

After TPA withdrawal, papilloma-bearing mice were randomly divided into groups with similar average tumor numbers per mouse. For neutralizing endogenous IFN-γ, mice were injected with the rat anti-mouse IFN-γ monoclonal antibody (mAb; R4-6A2; ref. 20) or with an isotype-matched control mAb [rat immunoglobulin G (IgG)] as control. Antibodies (0.3 mg per mouse) were administered once a week by intraperitoneal injection. For LXA4 treatment, mice were injected intraperitoneally with 100 ng LXA4 (Cayman) in 200 μL PBS twice a week. LXA4 was stored at −80°C and in dark. The structural integrity of LXA4 was confirmed before and after use (Supplementary Fig. S1). For t-butoxycarbonyl-Phe-Leu-Phe-Leu-Phe (BOC2) treatment blocking the LXA4 receptor (LXA4R; ref. 22), mice were injected intraperitoneally with 10 μg BOC2 (GeneScript) in 200 μL PBS twice a week. Mice treated with 200 μL PBS served as controls.

Cell proliferation and apoptosis assays

To label proliferating cells, mice were injected intraperitoneally with 2 mg bromodeoxyuridine (BrdUrd; Sigma) in 200 μL PBS, 6 hours before sacrifice. Paraffin sections of papillomas in 6 μm were simultaneously incubated with mouse anti-BrdUrd (ZBU30; Zhongshan Goldenbridge) and rabbit anti-cytokeratin (CK; Abcam) antibodies. Primary Ab binding was assessed using Alexa Fluor 488–conjugated goat anti-mouse IgG or Alexa Fluor 555-conjugated goat anti-rabbit IgG (Invitrogen). Apoptotic papilloma cells were identified by cytokeratin staining and terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) using the DeadEnd TUNEL Systems (Promega). Nuclei were counterstained with 4′-6-diamidino-2-phenylindole (DAPI; Sigma). At least 3 sections per papilloma, 2 papillomas per mouse were observed, and images of 2 high-power fields (HPF) from 1 section were taken by fluorescence microscopy (Olympus).

Real-time PCR

Total RNA was extracted from papillomas (less than 1 cm in length, noninvasive, and not ulcerated) as well as skin from untreated mice for control and cDNA was synthesized from 2 μg total RNA. Real-time reverse transcription-PCR was done with a RealMasterMix SYBRGreen Kit (Tiangen Biotech) using a Corbett Rotor-Gene real-time PCR detection system (Qiagen). The following primers were designed using PrimerBank or published sequences (23): (i) glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5′-CATCAAGAAGGTGGTGAAGC-3′ and 5′-CCTGTTGCTGTAGCCGTATT-3′; (ii) IFN-γ, 5′-ACAGCAAGGCGAAAAAGGATG-3′ and 5′-TGGTGGACCACTCGGATGA-3′; (iii) 15-LOX, 5′-GCGACGCTGCCCAATCCTAATC-3′ and 5′-CATATGGCCACGCTGTTTTCTACC-3′; (iv) 12-LOX, 5′-AGACAACAGACCTACTGCTGG-3′ and 5′-TCCTTACAGTCCGCTGCTATT-3′; and (v) 5-LOX, 5′-GGGCTGTAGCGAGAAGCATC-3′ and 5′-CACGGTGACATCGTAGGAGT-3′. Specific mRNA level of IFN-γ or LOXs of each sample was normalized to the respective GAPDH mRNA.

Cytokine analysis

The lysates of papillomas (less than 1 cm in length, noninvasive, and not ulcerated; 200 mg/mL) or cell supernatants were centrifuged at 12,000 rpm for 30 minutes at 4°C (20). The protein levels of cytokines were detected by ELISA or Cytometric Bead Arrays (CBA), all according to the manufacturers' instructions (BD Pharmingen).

LXA4 analysis

Tissue lysates (50 mg in 500 μL) prepared as described earlier were mixed with 1 mL methanol, stored at −80°C overnight, and centrifuged at 12,000 rpm for 30 minutes at 4°C. Clear lysates diluted with 7.5 mL deionized water were acidified to pH 3.5 with 1 N HCl and immediately loaded onto preconditioned C18 Sep-Pak VAC columns (Waters Corporation). Columns were washed with 5 mL deionized water followed by 5 mL hexane. Eluted methyl formate fractions (total 2 mL) were evaporated under a gentle N2 stream, reconstituted and assayed for LXA4 by ELISA as described in the manufacturers' instructions (Neogen).

Immunohistological staining

For inflammatory cell detection, cryostat tissue sections were stained with rat-derived anti-CD4 (H129.19), anti-CD8 (53-6.7), anti-Gr1 (RB6-8C5), anti-CD11b (M1/70), anti-B220 (RA3-6B2), and isotype-matched control mAbs (all from BD Pharmingen) in combination with Alexa Fluor 555–conjugated goat anti-rat IgG (Invitrogen). Sheep anti-mouse 15-LOX antibody (Cayman) was combined with rhodamine-conjugated donkey anti-sheep IgG (Jackson Immunoresearch). All sections were counterstained with DAPI. At least 3 sections per papilloma, 2 papillomas per mouse were observed, and images of 2 HPFs from 1 section were taken by fluorescence microscopy (Olympus).

Flow cytometry

Single-cell suspensions were prepared from spleens and stained for Th17 cells, regulatory T cells (Treg) or Gr1+CD11b+ cells. For Th17 detection, cells were stimulated with 50 ng/mL TPA, 1 μg/mL ionomycin, and 10 μg/mL brefeldin A (all from Sigma) for 6 hours. The following directly conjugated mAb were used: anti-CD4 (RM4-5), anti-CD25 (7D4), anti-Gr1 (RB6-8C5), anti-CD11b (M1/70), anti-IL-17A (TC11-18H10), and isotype control (all from BD Biosciences) as well as anti-Foxp3 (3G3; Miltenyi Biotec). Before intracellular staining of IL-17 or Foxp3, cells were fixed and permeabilized according to the manufacturers' instructions (BD Pharmingen and eBioscience). Antibody-binding was analyzed using a FACSCalibur flow cytometer and the CellQuest Pro software (BD Biosciences).

Statistical analysis

The comparisons between groups were analyzed with GraphPad Prism 5 software (GraphPad) using one-way ANOVA in Fig. 1D and E, and two-tailed, unpaired Student t test in other figures. Differences were considered significant when P values were less than 0.05.

Figure 1.

Spontaneous regression of papillomas was accompanied by a decrease of IFN-γ expression. A, experiment design for analysis of inflammation resolution and papilloma regression. Short arrows represent DMBA/TPA treatments in mice. Papilloma number and area were observed after TPA withdrawal. Eight weeks after TPA withdrawal, the relative mRNA (B) and protein (C) expression of IFN-γ in papillomas from regressors and nonregressors were detected. Papilloma-bearing mice at week 0 and regressors at week 5 (5 weeks without TPA) or week 8 (8 weeks without TPA) were randomly sacrificed and all papillomas were homogenized. The dynamic change of relative IFN-γ mRNA (D) and protein (E) levels in papillomas was determined over time. The mRNA expression was detected by real-time PCR and protein level was measured by ELISA. Shown are mean fold increase (±SEM) in IFN-γ expression from more than 3 mice per group, with skin tissue from untreated mice as control (defined as 1). *, P < 0.05.

Figure 1.

Spontaneous regression of papillomas was accompanied by a decrease of IFN-γ expression. A, experiment design for analysis of inflammation resolution and papilloma regression. Short arrows represent DMBA/TPA treatments in mice. Papilloma number and area were observed after TPA withdrawal. Eight weeks after TPA withdrawal, the relative mRNA (B) and protein (C) expression of IFN-γ in papillomas from regressors and nonregressors were detected. Papilloma-bearing mice at week 0 and regressors at week 5 (5 weeks without TPA) or week 8 (8 weeks without TPA) were randomly sacrificed and all papillomas were homogenized. The dynamic change of relative IFN-γ mRNA (D) and protein (E) levels in papillomas was determined over time. The mRNA expression was detected by real-time PCR and protein level was measured by ELISA. Shown are mean fold increase (±SEM) in IFN-γ expression from more than 3 mice per group, with skin tissue from untreated mice as control (defined as 1). *, P < 0.05.

Close modal

Spontaneous regression of papillomas was accompanied by a decrease of local IFN-γ

To investigate the role of IFN-γ in maintaining sterile inflammation and tumor persistence, we established papillomas by DMBA/TPA treatment of WT 129/Sv/Ev mice as described before (20). Spontaneous papilloma regression was observed after 14-week TPA application (Fig. 1A). We artificially separated tumor-bearing mice into 2 groups, namely, mice with regressing tumors (regressors) and mice with nonregressing tumors (nonregressors) as described in the Materials and Methods. Analysis showed that IFN-γ expression in papillomas of regressors was significantly lower than that of nonregressors both at the mRNA and protein levels (Fig. 1B and C). Furthermore, in the regressors, the expression of IFN-γ in papillomas decreased over time (Fig. 1D and E). The results suggested an important role of local IFN-γ in the maintenance of papillomas.

IFN-γR deficiency promoted papilloma regression

We then asked if the lack of IFN-γ signaling favored spontaneous regression of papillomas. Upon the treatment with DMBA/TPA, all 14 of 14 WT 129/SV/EV control mice and 9 of 14 IFN-γR-KO mice developed papillomas. The first mouse with regressing papilloma number appeared at week 4 and belonged to the IFN-γR-KO group, 2 weeks earlier than that in WT group. At week 8, more than 78% of IFN-γR-KO, whereas only 33% of WT mice had regressing papilloma numbers (Fig. 2A). Furthermore, the speed of papilloma number regression in IFN-γR-KO group was approximately 0.68 papillomas per week, higher than 0.37 of WT (Fig. 2B). Among mice with papilloma area regression, the papilloma area shrunk by 86.5% in IFN-γR-KO group 8 weeks after TPA withdrawal, whereas only by 34.7% in WT group (Fig. 2C). These results showed that IFN-γR deficiency promoted papilloma regression.

Figure 2.

Papilloma regression was accelerated in IFN-γR-KO mice. The number and area of papillomas in WT and IFN-γR-KO mice were observed over time after TPA withdrawal. Shown are percentages of mice with decreased papilloma number (A), average number of papillomas per mouse (B), and percentage shrinkage of papilloma area in mice with decreased papilloma area (C) at different time points. In B, the black arrow illustrates the time point of TPA withdrawal; the gray beeline in each curve shows the papilloma regression trend; the accordant slope value KWT or KKO indicates the regression rate of WT or IFN-γR-KO mice, respectively. Data are representative of 2 independent experiments (mean ± SEM in B and C). *, P < 0.05 between WT and IFN-γR-KO mice.

Figure 2.

Papilloma regression was accelerated in IFN-γR-KO mice. The number and area of papillomas in WT and IFN-γR-KO mice were observed over time after TPA withdrawal. Shown are percentages of mice with decreased papilloma number (A), average number of papillomas per mouse (B), and percentage shrinkage of papilloma area in mice with decreased papilloma area (C) at different time points. In B, the black arrow illustrates the time point of TPA withdrawal; the gray beeline in each curve shows the papilloma regression trend; the accordant slope value KWT or KKO indicates the regression rate of WT or IFN-γR-KO mice, respectively. Data are representative of 2 independent experiments (mean ± SEM in B and C). *, P < 0.05 between WT and IFN-γR-KO mice.

Close modal

Blocking endogenous IFN-γ accelerated papilloma regression

The enhanced spontaneous papilloma regression in IFN-γR-KO mice may be due to other inherent or developmental defect in these mice (20, 24). Therefore, the WT mice with established papillomas were treated with an anti-IFN-γ mAb after the stop of TPA application. The average papilloma number began to decrease significantly at week 6 in the anti-IFN-γ group, but did not for at least 10 weeks in the control group. At week 10, the mean papilloma number in control group was 9.3 ± 1.9; however, anti-IFN-γ treatment accelerated the tumor regression significantly to 3.4 ± 1.3 (Fig. 3A). As for tumor size, the speed of area shrinkage in mice with decreased papilloma area of anti-IFN-γ group was also significantly higher than that of control (Fig. 3B). Consistently, more mice with complete papilloma regression were observed in anti-IFN-γ group than that in control (Fig. 3C). Taken together, the results suggested that IFN-γ maintained the persistence of papillomas.

Figure 3.

Blocking IFN-γ promoted papilloma regression. Papilloma-bearing WT mice were treated with anti-IFN-γ or the isotype-matched control mAb directly after TPA withdrawal as described in Materials and Methods. Papilloma regression in mice was followed. Shown are average number of papillomas per mouse (A), percentage shrinkage of papilloma area in mice with decreased papilloma area (B), and percentage of tumor-bearing mice (C) at different time points after TPA withdrawal. Data are representative of 3 independent experiments (mean ± SEM in A and B). D, proliferation and apoptosis detection of epithelial cells in papillomas at week 10. Mice were injected with BrdUrd 6 hours before sacrifice. Then, tumor sections were stained by anti-cytokeratin antibody for epithelial cells (red), anti-BrdUrd antibody for proliferating cells (green; left), TUNEL for apoptotic cells (green; right), and DAPI (blue) for nuclei. Shown are proliferating epithelial cells (BrdUrd+CK+) and apoptotic epithelial cells (TUNEL+CK+) in papillomas from control (top) or anti-IFN-γ (bottom) group. The stroma (S) and outer layers (O) of papillomas are indicated. Representative images of immunofluorescence in situ (bar, 50 μm) from more than 3 mice per group are shown, as well as mean proportion (±SEM) of BrdUrd+CK+ or TUNEL+CK+ cells in relation to CK+ cells per HPF. *, P < 0.05 compared with control mice.

Figure 3.

Blocking IFN-γ promoted papilloma regression. Papilloma-bearing WT mice were treated with anti-IFN-γ or the isotype-matched control mAb directly after TPA withdrawal as described in Materials and Methods. Papilloma regression in mice was followed. Shown are average number of papillomas per mouse (A), percentage shrinkage of papilloma area in mice with decreased papilloma area (B), and percentage of tumor-bearing mice (C) at different time points after TPA withdrawal. Data are representative of 3 independent experiments (mean ± SEM in A and B). D, proliferation and apoptosis detection of epithelial cells in papillomas at week 10. Mice were injected with BrdUrd 6 hours before sacrifice. Then, tumor sections were stained by anti-cytokeratin antibody for epithelial cells (red), anti-BrdUrd antibody for proliferating cells (green; left), TUNEL for apoptotic cells (green; right), and DAPI (blue) for nuclei. Shown are proliferating epithelial cells (BrdUrd+CK+) and apoptotic epithelial cells (TUNEL+CK+) in papillomas from control (top) or anti-IFN-γ (bottom) group. The stroma (S) and outer layers (O) of papillomas are indicated. Representative images of immunofluorescence in situ (bar, 50 μm) from more than 3 mice per group are shown, as well as mean proportion (±SEM) of BrdUrd+CK+ or TUNEL+CK+ cells in relation to CK+ cells per HPF. *, P < 0.05 compared with control mice.

Close modal

Decreased proliferation and increased apoptosis of epithelial cells in the absence of IFN-γ

To understand the possible mechanism underlying the accelerated papilloma regression in the absence of IFN-γ, mice were injected with BrdUrd before sacrifice at week 10. The proliferation of local epithelial cells was analyzed by immunohistochemistry and at the same time, apoptosis of these cells was detected by TUNEL. As shown in Fig. 3D, papillomas in mice treated with anti-IFN-γ mAb had significantly less proliferating (BrdUrd+CK+) epithelial cells and more apoptotic (TUNEL+CK+) epithelial cells than control mice. Interestingly, proliferating epithelial cells were located in the middle of papillomas and apoptotic epithelial cells were found on the outer layer of papillomas. This suggested that the proliferation of papillomas required the tumor stroma–derived factors.

IFN-γ maintained inflammation in papilloma tissues

Nonresolving inflammation in tumor stroma supports viability and proliferation of tumor cells (4). Upon 10-week neutralization of endogenous IFN-γ, several proinflammatory cytokines, such as TNF-α, IL-6, -17, and the monocyte chemotactic protein-1 (MCP1) were significantly reduced in papilloma tissues. In contrast, the local protein levels of IL-10, -2, and -4 did not respond significantly to anti-IFN-γ treatment (Fig. 4A). As for local infiltrating leukocytes, Gr1+, CD11b+, or CD4+ cells were significantly reduced in papillomas of anti-IFN-γ–treated group (Fig. 4B). However, there was no significant difference in the numbers of CD8+ T cells (Fig. 4B) and B220+ cells (data not shown) between these 2 groups. The cellular response in spleen toward anti-IFN-γ–mediated acceleration of papilloma regression was analyzed at week 10. Although the percentages of inflammatory cells such as Gr1+CD11b+ bone marrow–derived suppressor cells (MDSC), CD4+Foxp3+ Tregs, and CD4+IL-17+ Th17 cells increased in spleens of tumor-bearing mice compared with untreated mice, but there was no significant difference between control and anti-IFN-γ groups, indicating the number of these immune regulatory cells in spleen did not make contributions to the accelerated papilloma regression (Fig. 4C). Rather the local inflammatory microenvironment established by the infiltrating immune cells and increased proinflammatory cytokines in papilloma tissue contributed to the proliferation of epithelial cells.

Figure 4.

IFN-γ maintained local inflammation in papilloma tissue. Tumor-bearing WT mice were treated with anti-IFN-γ or control mAb for 10 weeks. A, protein levels of TNF-α, MCP1, IL-6, -10, -2, -4, and -17 in the lysates of papillomas at week 10 were detected by CBA. The cytokine concentrations in papilloma tissue (mean ± SEM) of at least 3 mice per group are shown. B, infiltration of leukocytes in papillomas at week 10 from control (top) or anti-IFN-γ (bottom) group. Papilloma sections were stained with anti-Gr1, CD11b, CD4, or CD8 antibody (red) and nuclei were counterstained with DAPI (blue). Dotted lines delineate the border between epithelium and mesenchymal stroma. Representative images of immunofluorescence in situ (bar, 200 μm) from at least 3 mice per group are shown, as well as mean relative infiltration of inflammatory cells (±SEM) per HPF normalized to control group (defined as 1). *, P < 0.05. C, composition of immune regulatory cells in spleen at week 10 was analyzed by fluorescence-activated cell sorting (FACS). Representative density plots and percentages of Gr1+CD11b+, CD4+Foxp3+, and CD4+IL17+ for at least 3 mice per group are given.

Figure 4.

IFN-γ maintained local inflammation in papilloma tissue. Tumor-bearing WT mice were treated with anti-IFN-γ or control mAb for 10 weeks. A, protein levels of TNF-α, MCP1, IL-6, -10, -2, -4, and -17 in the lysates of papillomas at week 10 were detected by CBA. The cytokine concentrations in papilloma tissue (mean ± SEM) of at least 3 mice per group are shown. B, infiltration of leukocytes in papillomas at week 10 from control (top) or anti-IFN-γ (bottom) group. Papilloma sections were stained with anti-Gr1, CD11b, CD4, or CD8 antibody (red) and nuclei were counterstained with DAPI (blue). Dotted lines delineate the border between epithelium and mesenchymal stroma. Representative images of immunofluorescence in situ (bar, 200 μm) from at least 3 mice per group are shown, as well as mean relative infiltration of inflammatory cells (±SEM) per HPF normalized to control group (defined as 1). *, P < 0.05. C, composition of immune regulatory cells in spleen at week 10 was analyzed by fluorescence-activated cell sorting (FACS). Representative density plots and percentages of Gr1+CD11b+, CD4+Foxp3+, and CD4+IL17+ for at least 3 mice per group are given.

Close modal

LXA4 was a critical mediator in the accelerated papilloma regression

Inflammation resolution has been recognized as an active process induced by a series of lipid mediators that can inhibit the accumulation of leukocytes, the generation of inflammatory cytokines, and promote the clearance of apoptotic cells (11). Interestingly, the expression of LXA4 was upregulated in papilloma tissues compared with skin tissues from untreated mice and a significantly much more higher level of LXA4 was detected in papillomas after 8-week anti-IFN-γ treatment than that of isotype control (11.92 ± 0.83 ng/g vs. 6.88 ± 0.47 ng/g; Fig. 5A). However, the role of LXA4 in papilloma regression, especially in the accelerated papilloma regression induced by anti-IFN-γ treatment is largely unknown. By in vivo study, we found that administration of LXA4 in papilloma-bearing mice facilitated a faster regression of papilloma numbers compared with control group that did not receive LXA4 (Fig. 5B). Yet, injection of anti-IFN-γ mAb together with LXA4 did not lead to an additive effect on accelerating papilloma regression (Fig. 5B). To further confirm the critical role of LXA4 in the accelerated papilloma regression induced by neutralization of IFN-γ, we inhibited the LXA4 pathway in IFN-γ–neutralized papilloma-bearing mice. Blocking LXA4R with the antagonist BOC2 abrogated the acceleration of papilloma regression induced by neutralizing IFN-γ as shown by average numbers of papillomas (Fig. 5C). After 8-week BOC2 treatment, the elevated LXA4 generation induced by anti-IFN-γ treatment was not influenced in papillomas, despite the reduced rate of papilloma regression (Supplementary Fig. S2). The histologic analysis by hematoxylin and eosin (H&E) staining also showed an accelerated papilloma regression in mice with 8-week anti-IFN-γ and/or LXA4 treatment (Supplementary Fig. S3). These results indicated that IFN-γ inhibited papilloma regression by downregulating LXA4-mediated inflammation resolution and therapeutic LXA4 treatment counteracted the suppression of papilloma regression by endogenous IFN-γ.

Figure 5.

IFN-γ maintained papilloma persistence by downregulation of LXA4. A, concentration of LXA4 in papillomas of mice with 8-week isotype control or anti-IFN-γ mAb treatment and in skin from untreated mice was detected by ELISA. Mean concentration of LXA4 (±SEM) from 3 to 6 mice per group is shown. *, P < 0.05. B, after TPA application was stopped, randomly grouped tumor-bearing mice were treated by LXA4 in combination with isotype control mAb (LXA4 group), LXA4 together with anti-IFN-γ mAb (LXA4 and anti-IFN-γ group), or isotype control mAb alone (control group). Shown are average numbers of papillomas per mouse (mean ± SEM) representative for 2 independent experiments. *, P < 0.05, compared with LXA4 group. C, after TPA painting was stopped, randomly grouped tumor-bearing mice were administered anti-IFN-γ mAb in the presence of BOC2 (BOC2 group) or anti-IFN-γ mAb alone (control group). Shown are average numbers of papillomas per mouse (mean ± SEM) representative for 2 independent experiments. *, P < 0.05, compared with control group.

Figure 5.

IFN-γ maintained papilloma persistence by downregulation of LXA4. A, concentration of LXA4 in papillomas of mice with 8-week isotype control or anti-IFN-γ mAb treatment and in skin from untreated mice was detected by ELISA. Mean concentration of LXA4 (±SEM) from 3 to 6 mice per group is shown. *, P < 0.05. B, after TPA application was stopped, randomly grouped tumor-bearing mice were treated by LXA4 in combination with isotype control mAb (LXA4 group), LXA4 together with anti-IFN-γ mAb (LXA4 and anti-IFN-γ group), or isotype control mAb alone (control group). Shown are average numbers of papillomas per mouse (mean ± SEM) representative for 2 independent experiments. *, P < 0.05, compared with LXA4 group. C, after TPA painting was stopped, randomly grouped tumor-bearing mice were administered anti-IFN-γ mAb in the presence of BOC2 (BOC2 group) or anti-IFN-γ mAb alone (control group). Shown are average numbers of papillomas per mouse (mean ± SEM) representative for 2 independent experiments. *, P < 0.05, compared with control group.

Close modal

The lipoxygenase 15-LOX was involved in IFN-γ–mediated downregulation of LXA4

It is known that LXA4 can be synthesized through 15- and 5-LOX pathway or 5- and 12-LOX pathway (13). Here, we analyzed the expression of the three LOXs important for lipid mediator generation after 8-week anti-IFN-γ treatment in papillomas. The mRNA expression of 5- and 12-LOX had no difference between control and anti-IFN-γ groups (Fig. 6A). However, the mRNA expression of 15-LOX in anti-IFN-γ group was more than 3 times higher than that in control group (Fig. 6A). In situ, significantly more 15-LOX+ cells were found in the tumor stroma of anti-IFN-γ group compared with the control (Fig. 6B). All these suggested that IFN-γ could downregulate LXA4 synthesis by inhibiting 15-LOX expression.

Figure 6.

IFN-γ inhibited the expression of 15-LOX. After 8-week anti-IFN-γ or isotype control mAb administration in papilloma-bearing WT mice, mRNA levels of 5-, 12-, and 15-LOX in papillomas were detected by real-time PCR (A). Shown are mean fold increase (±SEM) in expression for 3 to 8 mice per group, with isotype control mAb-treated group as control (defined as 1). *, P < 0.05. B, 15-LOX+ cells in papillomas from control (left) or anti-IFN-γ (right) group. Thin sections of papillomas were stained by anti-15-LOX antibody (red) and counterstained with DAPI (blue). Dotted lines delineate the border between epithelium and mesenchymal stroma. Inserts display enlarged images of the areas in gray rectangles (bar, 50 μm). Representative images of immunofluorescence in situ (bar, 200 μm) from at least 3 mice per group were shown, as well as mean relative numbers of 15-LOX+ cells (±SEM) per HPF normalized to control group (defined as 1).

Figure 6.

IFN-γ inhibited the expression of 15-LOX. After 8-week anti-IFN-γ or isotype control mAb administration in papilloma-bearing WT mice, mRNA levels of 5-, 12-, and 15-LOX in papillomas were detected by real-time PCR (A). Shown are mean fold increase (±SEM) in expression for 3 to 8 mice per group, with isotype control mAb-treated group as control (defined as 1). *, P < 0.05. B, 15-LOX+ cells in papillomas from control (left) or anti-IFN-γ (right) group. Thin sections of papillomas were stained by anti-15-LOX antibody (red) and counterstained with DAPI (blue). Dotted lines delineate the border between epithelium and mesenchymal stroma. Inserts display enlarged images of the areas in gray rectangles (bar, 50 μm). Representative images of immunofluorescence in situ (bar, 200 μm) from at least 3 mice per group were shown, as well as mean relative numbers of 15-LOX+ cells (±SEM) per HPF normalized to control group (defined as 1).

Close modal

All together, IFN-γ maintained the sterile inflammation by regulating 15-LOX–LXA4 synthesis pathway. Blocking IFN-γ or increasing LXA4 could break the dominance of proinflammatory factors and promote inflammation resolution and papilloma regression.

The spontaneous regression of DMBA/TPA–induced papilloma is an ideal model to analyze the effects of inflammation resolution on tumorigenesis. Here, we showed that blocking IFN-γ accelerated the resolution of sterile inflammation and therefore enhanced the spontaneous regression of papillomas by upregulation of the proresolving molecule LXA4. This study provides new evidence for tumor therapy by promoting inflammation resolution.

Papilloma regression is not likely due to the enhanced antitumor immune response. (i) As a key cytokine in activating antitumor immunity, IFN-γ should promote papilloma regression, but the results here were quite contrary (Figs. 2 and 3). (ii) In this work, IFN-γ maintained the infiltration of CD4+ but not CD8+ T cells in papillomas (Fig. 4B), and CD4+ T cells impaired the regression of papillomas (Supplementary Fig. S4). (iii) IFN-γ did not influence the composition of immune regulatory cells (4), such as MDSCs and Tregs in spleen (Fig. 4C), but it increased the infiltration of inflammatory cells, especially the Gr1+/CD11b+ cells (Fig. 4B) in local papilloma tissues. Furthermore, studies from other laboratories show no evidence that papilloma regression is induced by antitumor immunity. Antigen-specific immunity toward mutant Ras or P53 protein fails to eradicate mutant oncogene-expressing epidermal cells and even induces a remarkable enhancement of tumor growth (25, 26). T cells can promote papilloma development (21, 27, 28) and deficiency of CXCR3, an important chemokine receptor for T-cell recruitment, inhibits skin tumorigenesis (29). Antithymocyte serum has no effect on papilloma regression (30). Therefore, increased papilloma regression was not a process mediated by antitumor immunity.

Papilloma regression is the result of impaired maintenance of sterile inflammation. Papilloma growth strictly depends on growth factors, cytokines, and reactive oxygen species provided by keratinocytes and infiltrating inflammatory cells activated by TPA (31). If TPA treatment is stopped, TPA-dependent papillomas tend to regress, whereas TPA-independent papillomas can form autonomous benign lesions and some of them might progress to carcinoma (9). IFN-γ is a proinflammatory cytokine in papilloma environment. Acting on tumor epithelial cells, IFN-γ can provoke proinflammatory cytokine production and leukocyte adhesion (32, 33). We found that IFN-γ was effective to promote TNF-α and IL-6 expression and enhance infiltrations of CD4+, Gr1+, and CD11b+ cells in papilloma tissues (Fig. 4A and B). In accordance, IFN-γ is elevated in many human inflammatory skin diseases, including UVA–induced squamous cell carcinoma (34, 35). Besides, local IFN-γ administration on human skin induces epidermal hyperproliferation concomitant with T-cell–rich inflammatory dermal reaction (36, 37). Anti-IFN-γ treatment has shown its therapeutic efficacy in many inflammatory skin diseases (35, 38). All together, by mediating the talk between tumor, innate, and adoptive immune systems, IFN-γ may play a central role in maintaining the inflammation loop during the development of skin tumors. However, blocking IFN-γ did not affect the potential of malignant progression as shown in Supplementary Table S1, which indicated IFN-γ–independent mechanisms for tumor malignant procession.

The inflammation resolution can be induced by regulating the balance between proinflammatory and proresolving factors (1). Proresolving LXA4 can induce the expression of the suppressor of cytokine signaling-2 (SOCS-2), subsequently suppressing the cytokine-activated JAK–STAT signaling pathway and impair cytokine-induced inflammation (22). The therapeutic role of LXA4 has been reported in many acute inflammation models, including skin inflammation in mouse ear treated with TPA-containing croton oil (39). As for chronic inflammation, a reduction of LXA4 generation was found in human ulcerative colitis (40), cirrhosis (41), and chronic myelogenous leukemia (42, 43). Oral treatment with LXA4 analog effectively promoted colitis resolution in a hapten-induced mouse model of Crohn's disease (17). In accordance, the anti-inflammatory drug aspirin reduces colorectal cancer incidence and death rate of several common cancers (5) and a stable analog of LXA4 (carbon 15-epimeric lipoxin A4, 15-epi-LXA4) is an endogenous metabolic product of aspirin in vivo. With a higher proresolving efficacy compared with LXA4, this analog might be an important mediator for aspirin's protective effect in inflammation-associated tumors (11, 44). Earlier studies from Claria and colleagues have shown that LXA4 and especially 15-epi-LXA4 have an antiproliferative effect on human lung adenocarcinoma cell line A549 in vitro (45). Other studies indicated that these lipoxins may also inhibit tumor growth by suppressing angiogenesis (46). Here, we found that LXA4 administration promoted the regression of DMBA/TPA–induced papillomas (Fig. 5B) and decreased the protein levels of several proinflammatory cytokines including IFN-γ in papilloma tissue (Supplementary Fig. S5A). This effect could be retarded by the antagonist BOC2 (Supplementary Fig. S5). Mice with only BOC2 treatment even showed a slightly higher level of TNF-α and MCP1 expression and an inhibited papilloma regression, compared with that treated with only PBS (Supplementary Fig. S5). This implied some potential effect of the low level LXA4 on the slow papilloma regression of control mice. Meanwhile, 15-epi-LXA4 was also efficient in accelerating inflammation resolution and eventually regression of papillomas (Supplementary Fig. S5). Therefore, LXA4 and its analogs are important modulators in balancing the proinflammatory cytokines, providing a potential for the treatment of inflammation-associated tumors.

Here, for the first time, we found that blocking IFN-γ can promote the generation of LXA4 in papilloma tissues (Fig. 5A). Meanwhile, we cocultured skin fibroblasts and splenocytes to mimic the inflammatory environment in papillomas and found that by acting on fibroblasts, IFN-γ could decline LXA4 generation in the coculture system, whereas blocking JAK–STAT-1 pathway could reverse the effect of IFN-γ (Supplementary Fig. S6A). Consistently, of the 3 important LOXs in LXA4 generation, the 15-LOX expression was increased in papilloma tissues in response to IFN-γ inhibition (Fig. 6A). 15-LOX can switch eicosanoid production from proinflammatory leukotrienes and prostaglandins to proresolving lipoxins in human leukocytes (47). Others have reported an important role of macrophage with increased 15-LOX expression for inflammation resolution (48, 49). Our in vitro study showed that IFN-γ repressed expression of 15-LOX not only in macrophages but also in skin fibroblasts from neonatal mice (Supplementary Fig. S6B–S6D). Studies from other laboratory showed that IFN-γ can attenuate IL-4–dependent mRNA stabilization of 15-LOX (50). According to these results, we proposed that the persistent expression of IFN-γ suppressed the expression of 15-LOX in fibroblasts/macrophages, and therefore, reduced the production of LXA4 both in vitro and in vivo.

In summary, nonresolving inflammation is a major driver of many diseases, especially tumors. The uncontrolled balance between proinflammation and proresolution factors can lead to nonresolving inflammation and tumor persistence (1). Here, we show that blocking IFN-γ accelerated regression of DMBA/TPA–induced skin papillomas by upregulating lipid mediator LXA4 and this throws new light on tumor therapy with proresolving strategies.

No potential conflicts of interest were disclosed.

Conception and design: C. Wang, M. Xiao, C. Ni, Z. Qin

Development of methodology: C. Wang, J. Liu

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C. Wang, M. Xiao, X. Liu, J. Liu

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C. Wang, C. Ni, J. Liu, U. Erben

Writing, review, and/or revision of the manuscript: C. Wang, M. Xiao, C. Ni, U. Erben, Z. Qin

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Wang, M. Xiao, Z. Qin

Study supervision: Z. Qin

The authors thank Qinghong Meng, Ning Tao, Xiao Li, Xia Xu, Jinhua Zhang, Jingjing Deng, Lin Chen, Lijie Rong, Jie Li, Liwei Qu, Pan Gao, Chengliang Dai, Shuna Cui, Hao Wu, and Liang Cheng for technical assistance and helpful discussion; Junying Jia for flow cytometry analysis; Zhenwei Yang for quantity PCR analysis; Junfeng Hao and Jianhua Wang for pathology analysis; Tanxi Cai for MS/MS; Tao Su for reversed phase high-performance liquid chromatography (RP-HPLC); Xinyi Wu for animal experiments; and Jenny Anne Cook for article revision.

This work was supported by the Ministry of Science and Technology of China (2012CB917103), the National Natural Science Foundation of China (81030049 and 91229203), the grant of Plan For Scientific Innovation Talent of Henan Province (104200510001), and the Special Fund on Scientific Research for Excellent Doctoral Dissertation and dean prize winner of CAS in first half year of 2010.

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.

1.
Nathan
C
,
Ding
A
. 
Nonresolving inflammation
.
Cell
2010
;
140
:
871
82
.
2.
Mantovani
A
,
Allavena
P
,
Sica
A
,
Balkwill
F
. 
Cancer-related inflammation
.
Nature
2008
;
454
:
436
44
.
3.
Balkwill
F
,
Mantovani
A
. 
Cancer and inflammation: implications for pharmacology and therapeutics
.
Clin Pharmacol Ther
2010
;
87
:
401
6
.
4.
Grivennikov
SI
,
Greten
FR
,
Karin
M
. 
Immunity, inflammation, and cancer
.
Cell
2010
;
140
:
883
99
.
5.
Burn
J
,
Gerdes
AM
,
Macrae
F
,
Mecklin
JP
,
Moeslein
G
,
Olschwang
S
, et al
Long-term effect of aspirin on cancer risk in carriers of hereditary colorectal cancer: an analysis from the CAPP2 randomised controlled trial
.
Lancet
2011
;
378
:
2081
7
.
6.
Coussens
LM
,
Werb
Z
. 
Inflammation and cancer
.
Nature
2002
;
420
:
860
7
.
7.
Yuspa
SH
. 
Cutaneous chemical carcinogenesis
.
J Am Acad Dermatol
1986
;
15
:
1031
44
.
8.
Ishikawa
TO
,
Jain
NK
,
Herschman
HR
. 
Cox-2 gene expression in chemically induced skin papillomas cannot predict subsequent tumor fate
.
Mol Oncol
2010
;
4
:
347
56
.
9.
Balmain
A
,
Ramsden
M
,
Bowden
GT
,
Smith
J
. 
Activation of the mouse cellular Harvey-ras gene in chemically induced benign skin papillomas
.
Nature
1984
;
307
:
658
60
.
10.
Serhan
CN
. 
Lipoxins and aspirin-triggered 15-epi-lipoxins are the first lipid mediators of endogenous anti-inflammation and resolution
.
Prostaglandins Leukot Essent Fatty Acids
2005
;
73
:
141
62
.
11.
Serhan
CN
. 
Resolution phase of inflammation: novel endogenous anti-inflammatory and proresolving lipid mediators and pathways
.
Annu Rev Immunol
2007
;
25
:
101
37
.
12.
Serhan
CN
,
Chiang
N
,
Van Dyke
TE
. 
Resolving inflammation: dual anti-inflammatory and pro-resolution lipid mediators
.
Nat Rev Immunol
2008
;
8
:
349
61
.
13.
Serhan
CN
. 
Lipoxins and aspirin-triggered 15-epi-lipoxin biosynthesis: an update and role in anti-inflammation and pro-resolution
.
Prostaglandins Other Lipid Mediat
2002
;
68–69
:
433
55
.
14.
Schwab
JM
,
Chiang
N
,
Arita
M
,
Serhan
CN
. 
Resolvin E1 and protectin D1 activate inflammation-resolution programmes
.
Nature
2007
;
447
:
869
74
.
15.
Ohse
T
,
Ota
T
,
Kieran
N
,
Godson
C
,
Yamada
K
,
Tanaka
T
, et al
Modulation of interferon-induced genes by lipoxin analogue in anti-glomerular basement membrane nephritis
.
J Am Soc Nephrol
2004
;
15
:
919
27
.
16.
Aliberti
J
,
Serhan
C
,
Sher
A
. 
Parasite-induced lipoxin A4 is an endogenous regulator of IL-12 production and immunopathology in Toxoplasma gondii infection
.
J Exp Med
2002
;
196
:
1253
62
.
17.
Fiorucci
S
,
Wallace
JL
,
Mencarelli
A
,
Distrutti
E
,
Rizzo
G
,
Farneti
S
, et al
A beta-oxidation-resistant lipoxin A4 analog treats hapten-induced colitis by attenuating inflammation and immune dysfunction
.
Proc Natl Acad Sci U S A
2004
;
101
:
15736
41
.
18.
Conrad
DJ
,
Kuhn
H
,
Mulkins
M
,
Highland
E
,
Sigal
E
. 
Specific inflammatory cytokines regulate the expression of human monocyte 15-lipoxygenase
.
Proc Natl Acad Sci U S A
1992
;
89
:
217
21
.
19.
Nassar
GM
,
Morrow
JD
,
Roberts
LJ
 II
,
Lakkis
FG
,
Badr
KF
. 
Induction of 15-lipoxygenase by interleukin-13 in human blood monocytes
.
J Biol Chem
1994
;
269
:
27631
4
.
20.
Xiao
M
,
Wang
C
,
Zhang
J
,
Li
Z
,
Zhao
X
,
Qin
Z
. 
IFNgamma promotes papilloma development by up-regulating Th17-associated inflammation
.
Cancer Res
2009
;
69
:
2010
7
.
21.
Roberts
SJ
,
Ng
BY
,
Filler
RB
,
Lewis
J
,
Glusac
EJ
,
Hayday
AC
, et al
Characterizing tumor-promoting T cells in chemically induced cutaneous carcinogenesis
.
Proc Natl Acad Sci U S A
2007
;
104
:
6770
5
.
22.
Machado
FS
,
Johndrow
JE
,
Esper
L
,
Dias
A
,
Bafica
A
,
Serhan
CN
, et al
Anti-inflammatory actions of lipoxin A4 and aspirin-triggered lipoxin are SOCS-2 dependent
.
Nat Med
2006
;
12
:
330
4
.
23.
Gronert
K
,
Maheshwari
N
,
Khan
N
,
Hassan
IR
,
Dunn
M
,
Laniado
Schwartzman M
. 
A role for the mouse 12/15-lipoxygenase pathway in promoting epithelial wound healing and host defense
.
J Biol Chem
2005
;
280
:
15267
78
.
24.
Huang
S
,
Hendriks
W
,
Althage
A
,
Hemmi
S
,
Bluethmann
H
,
Kamijo
R
, et al
Immune response in mice that lack the interferon-gamma receptor
.
Science
1993
;
259
:
1742
5
.
25.
Siegel
CT
,
Schreiber
K
,
Meredith
SC
,
Beck-Engeser
GB
,
Lancki
DW
,
Lazarski
CA
, et al
Enhanced growth of primary tumors in cancer-prone mice after immunization against the mutant region of an inherited oncoprotein
.
J Exp Med
2000
;
191
:
1945
56
.
26.
Remenyik
E
,
Wikonkal
NM
,
Zhang
W
,
Paliwal
V
,
Brash
DE
. 
Antigen-specific immunity does not mediate acute regression of UVB-induced p53-mutant clones
.
Oncogene
2003
;
22
:
6369
76
.
27.
Girardi
M
,
Glusac
E
,
Filler
RB
,
Roberts
SJ
,
Propperova
I
,
Lewis
J
, et al
The distinct contributions of murine T cell receptor (TCR)gammadelta+ and TCRalphabeta+ T cells to different stages of chemically induced skin cancer
.
J Exp Med
2003
;
198
:
747
55
.
28.
Yusuf
N
,
Nasti
TH
,
Katiyar
SK
,
Jacobs
MK
,
Seibert
MD
,
Ginsburg
AC
, et al
Antagonistic roles of CD4+ and CD8+ T-cells in 7,12-dimethylbenz(a)anthracene cutaneous carcinogenesis
.
Cancer Res
2008
;
68
:
3924
30
.
29.
Winkler
AE
,
Brotman
JJ
,
Pittman
ME
,
Judd
NP
,
Lewis
JS
 Jr
,
Schreiber
RD
, et al
CXCR3 enhances a T-cell-dependent epidermal proliferative response and promotes skin tumorigenesis
.
Cancer Res
2011
;
71
:
5707
16
.
30.
Burns
FJ
,
Vanderlaan
M
,
Sivak
A
,
Albert
RE
. 
Regression kinetics of mouse skin papillomas
.
Cancer Res
1976
;
36
:
1422
7
.
31.
Rundhaug
JE
,
Fischer
SM
. 
Molecular mechanisms of mouse skin tumor promotion
.
Cancers (Basel)
2010
;
2
:
436
82
.
32.
Nickoloff
BJ
,
Lewinsohn
DM
,
Butcher
EC
. 
Enhanced binding of peripheral blood mononuclear leukocytes to gamma-interferon-treated cultured keratinocytes
.
Am J Dermatopathol
1987
;
9
:
413
8
.
33.
Teunissen
MB
,
Koomen
CW
,
de Waal Malefyt
R
,
Wierenga
EA
,
Bos
JD
. 
Interleukin-17 and interferon-gamma synergize in the enhancement of proinflammatory cytokine production by human keratinocytes
.
J Invest Dermatol
1998
;
111
:
645
9
.
34.
Bachelor
MA
,
Bowden
GT
. 
UVA-mediated activation of signaling pathways involved in skin tumor promotion and progression
.
Semin Cancer Biol
2004
;
14
:
131
8
.
35.
Skurkovich
S
,
Skurkovich
B
. 
Anticytokine therapy, especially anti-interferon-gamma, as a pathogenetic treatment in TH-1 autoimmune diseases
.
Ann N Y Acad Sci
2005
;
1051
:
684
700
.
36.
Barker
JN
,
Goodlad
JR
,
Ross
EL
,
Yu
CC
,
Groves
RW
,
MacDonald
DM
. 
Increased epidermal cell proliferation in normal human skin in vivo following local administration of interferon-gamma
.
Am J Pathol
1993
;
142
:
1091
7
.
37.
Nickoloff
BJ
,
Griffiths
CE
,
Barker
JN
. 
The role of adhesion molecules, chemotactic factors, and cytokines in inflammatory and neoplastic skin disease—1990 update
.
J Invest Dermatol
1990
;
94
:
151S
7S
.
38.
Skurkovich
B
,
Skurkovich
S
. 
Anti-interferon-gamma antibodies in the treatment of autoimmune diseases
.
Curr Opin Mol Ther
2003
;
5
:
52
7
.
39.
Takano
T
,
Fiore
S
,
Maddox
JF
,
Brady
HR
,
Petasis
NA
,
Serhan
CN
. 
Aspirin-triggered 15-epi-lipoxin A4 (LXA4) and LXA4 stable analogues are potent inhibitors of acute inflammation: evidence for anti-inflammatory receptors
.
J Exp Med
1997
;
185
:
1693
704
.
40.
Mangino
MJ
,
Brounts
L
,
Harms
B
,
Heise
C
. 
Lipoxin biosynthesis in inflammatory bowel disease
.
Prostaglandins Other Lipid Mediat
2006
;
79
:
84
92
.
41.
Claria
J
,
Titos
E
,
Jimenez
W
,
Ros
J
,
Gines
P
,
Arroyo
V
, et al
Altered biosynthesis of leukotrienes and lipoxins and host defense disorders in patients with cirrhosis and ascites
.
Gastroenterology
1998
;
115
:
147
56
.
42.
Stenke
L
,
Edenius
C
,
Samuelsson
J
,
Lindgren
JA
. 
Deficient lipoxin synthesis: a novel platelet dysfunction in myeloproliferative disorders with special reference to blastic crisis of chronic myelogenous leukemia
.
Blood
1991
;
78
:
2989
95
.
43.
Stenke
L
,
Nasman-Glaser
B
,
Edenius
C
,
Samuelsson
J
,
Palmblad
J
,
Lindgren
JA
. 
Lipoxygenase products in myeloproliferative disorders: increased leukotriene C4 and decreased lipoxin formation in chronic myeloid leukemia
.
Adv Prostaglandin Thromboxane Leukot Res
1991
;
21B
:
883
6
.
44.
Rothwell
PM
,
Fowkes
FG
,
Belch
JF
,
Ogawa
H
,
Warlow
CP
,
Meade
TW
. 
Effect of daily aspirin on long-term risk of death due to cancer: analysis of individual patient data from randomised trials
.
Lancet
2011
;
377
:
31
41
.
45.
Claria
J
,
Lee
MH
,
Serhan
CN
. 
Aspirin-triggered lipoxins (15-epi-LX) are generated by the human lung adenocarcinoma cell line (A549)-neutrophil interactions and are potent inhibitors of cell proliferation
.
Mol Med
1996
;
2
:
583
96
.
46.
Chen
Y
,
Hao
H
,
He
S
,
Cai
L
,
Li
Y
,
Hu
S
, et al
Lipoxin A4 and its analogue suppress the tumor growth of transplanted H22 in mice: the role of antiangiogenesis
.
Mol Cancer Ther
2010
;
9
:
2164
74
.
47.
Levy
BD
,
Clish
CB
,
Schmidt
B
,
Gronert
K
,
Serhan
CN
. 
Lipid mediator class switching during acute inflammation: signals in resolution
.
Nat Immunol
2001
;
2
:
612
9
.
48.
Schif-Zuck
S
,
Gross
N
,
Assi
S
,
Rostoker
R
,
Serhan
CN
,
Ariel
A
. 
Saturated-efferocytosis generates pro-resolving CD11b low macrophages: modulation by resolvins and glucocorticoids
.
Eur J Immunol
2011
;
41
:
366
79
.
49.
Stables
MJ
,
Shah
S
,
Camon
EB
,
Lovering
RC
,
Newson
J
,
Bystrom
J
, et al
Transcriptomic analyses of murine resolution-phase macrophages
.
Blood
2011
;
118
:
e192
208
.
50.
Chen
B
,
Tsui
S
,
Boeglin
WE
,
Douglas
RS
,
Brash
AR
,
Smith
TJ
. 
Interleukin-4 induces 15-lipoxygenase-1 expression in human orbital fibroblasts from patients with Graves disease. Evidence for anatomic site-selective actions of Th2 cytokines
.
J Biol Chem
2006
;
281
:
18296
306
.