Abstract
Increased recruitment of tumor-associated macrophages (TAM) to tumors following chemotherapy promotes tumor resistance and recurrence and correlates with poor prognosis. TAM depletion suppresses tumor growth, but is not highly effective due to the effects of tumorigenic mediators from other stromal sources. Here, we report that adoptive macrophage transfer led to a dramatically enhanced photodynamic therapy (PDT) effect of 2-(1-hexyloxyethyl)-2-devinyl pyropheophor-bide-alpha (HPPH)-coated polyethylene glycosylated nanographene oxide [GO(HPPH)-PEG] by increasing its tumor accumulation. Moreover, tumor treatment with commonly used chemotherapeutic drugs induced an increase in macrophage infiltration into tumors, which also enhanced tumor uptake and the PDT effects of GO(HPPH)-PEG, resulting in tumor eradication. Macrophage recruitment to tumors after chemotherapy was visualized noninvasively by near-infrared fluorescence and single-photon emission CT imaging using F4/80-specific imaging probes. Our results demonstrate that chemotherapy combined with GO(HPPH)-PEG PDT is a promising strategy for the treatment of tumors, especially those resistant to chemotherapy. Furthermore, TAM-targeted molecular imaging could potentially be used to predict the efficacy of combination therapy and select patients who would most benefit from this treatment approach. Cancer Res; 77(21); 6021–32. ©2017 AACR.
Introduction
Despite being one of the most widely used treatment strategies for cancer, the efficacy of chemotherapy is undermined by drug resistance and postchemotherapy tumor relapse, as well as serious side effects due to the systemic distribution and limited tumor cytotoxicity of drugs (1, 2). Photodynamic therapy (PDT), which is based on the delivery of photosensitizers, followed by light irradiation at a wavelength that can induce the generation of singlet oxygen or reactive oxygen species, is a safe treatment strategy that has been successfully used to treat cancer and other diseases (3, 4). Compared with chemotherapy, PDT has minimal side effects and higher specificity for tumor cells (4) and can also overcome tumor recurrence and drug chemoresistance (5–7). However, an unresolved challenge is the efficient delivery of photosensitizers to the tumor site to enhance tumor response (8, 9).
To improve the PDT effect, photosensitizers can be coated onto nanoparticles to increase tumor uptake via the enhanced penetration and retention (EPR) effect of tumors, which is attributable to their abnormal blood vessel structure and results in extravasation and macromolecule retention (10, 11). However, despite increased delivery efficiency, tumor targeting is suboptimal due to pharmacokinetics and physiologic barriers that lead to heterogeneous particle distribution (12). Moreover, EPR achieves a <2-fold increase in tumor uptake as compared with normal organs (13). Nanoparticle optimization strategies, such as altering size, shape, and composition, as well as surface modifications, such as polyethylene glycol (PEG) coating and conjugating of tumor-targeting ligands, have been proposed to overcome these limitations (14, 15). However, few studies have focused on exploiting the physiologic properties of tumors, such as tumor blood flow, vasculature permeability, and the microenvironment, to improve drug uptake and in vivo kinetics (16–18).
Tumor tissues comprise not only tumor cells but also the stroma that combine to create a unique microenvironment (19). Macrophages are a major component of tumor stroma that are recruited as monocytes from the peripheral blood and play important roles in tumor progression, invasion, and metastasis, as well as tumor-related inflammation and escape from immunosurveillance (20, 21). These tumor-associated macrophages (TAM) are induced by local signals to a state of classical (M1-polarized) or alternative (M2-polarized) activation (19). Inflammatory M1 macrophages are cytotoxic to tumor cells, whereas noninflammatory M2 macrophages are immunosuppressive and facilitate tumor progression and invasion (21, 22). Macrophages in the tumor microenvironment engulf nanoparticles by phagocytosis (23, 24) and promote tumor uptake, thereby increasing the efficacy of drugs delivered by nanoparticles (25, 26). In addition, TAMs contribute to angiogenesis and enhance tumor vascular permeability (27), suggesting an enhancement of the EPR effect. However, whether TAMs can be used to enhance the PDT effect by increasing tumor uptake of photosensitizer-coated nanomaterials remains unknown.
In this study, we hypothesized that TAM recruitment in tumors could increase the tumor uptake and PDT effect of 2-(1-hexyloxyethyl)-2-devinyl pyropheophor-bide-alpha (HPPH)-coated PEGylated nanographene oxide (GO-PEG). HPPH is a photosensitizer currently in clinical trials for PDT of lung, Barrett's esophageal, and head and neck cancers (4). GO(HPPH)-PEG generated by coating GO-PEG with HPPH via π–π stacking and hydrophobic interactions increases photosensitizer solubility and enhances tumor uptake of HPPH (28). Here, we investigated the improved PDT effect of GO(HPPH)-PEG in tumor mouse models treated with either adoptive transfer of F4/80+ macrophages or chemotherapy. In addition, we investigated whether macrophage-specific imaging probes enable sensitive and noninvasive visualization of postchemotherapy macrophage infiltration into tumors by near-infrared fluorescence (NIRF) imaging and single-photon emission CT (SPECT).
Materials and Methods
Preparation of GO(HPPH)-PEG
GO-COOH was prepared from graphite using the modified Hummers method (29). NH2-PEG5000 (JenKem Technology) was added to GO-COOH solution (1 mg/mL) at a weight ratio of 10:1 and sonicated for 15 minutes. N-(3-dimethylaminopropyl-N-ethylcarbodiimide)hydrochloride (EDC) was then mixed into the solution, followed by overnight stirring at room temperature and centrifugation at 21,000 × g for 30 minutes. The supernatant was filtered through a centrifugal filter with a 100-kDa molecular weight cutoff (Millipore) to remove excess NH2-PEG5000. The resultant GO-PEG was mixed with HPPH (Chembest) in 1% Tween-20 solution with overnight stirring, yielding GO(HPPH)-PEG that was purified by filtration, followed by washing with PBS. GO(HPPH)-PEG was characterized by atomic force microscopy and absorption analysis.
Cell lines and animal models
The 4T1 murine breast cancer and RAW264.7 murine macrophage (Abelson murine leukemia virus-induced tumor) cell lines were purchased from the ATCC in 2015 and authenticated using STR profile analysis before receipt. All cells used in this study were passaged fewer than 6 months after receipt or resuscitation, and no further genetic characterization was performed. Cells were grown in RPMI1640 medium supplemented with 10% FBS at 37°C in a humidified atmosphere containing 5% CO2 and were routinely screened for mycoplasma (Hoechst stain and PCR).
All animal experiments were performed in accordance with the Guidelines of Peking University Animal Care and Use Committee. BALB/c mice (5 weeks of age) were obtained from Department of Laboratory Animal Science of Peking University (Beijing, China). To establish the 4T1 tumor-bearing mouse model, 1 × 106 4T1 cells were subcutaneously injected into the hind legs of female BALB/c mice. To establish a dual tumor-bearing mouse model, 1 × 106 4T1 cells were subcutaneously injected into bilateral hind legs of female BALB/c mice. Tumor growth was measured with a caliper every other day, and tumor volume was calculated using the formula: tumor volume = length × width2/2.
Macrophage sorting and adoptive transfer
Donor 4T1 tumor-bearing mice were euthanized, and their tumors were digested to obtain single-cell suspensions according to a previously reported protocol (7). The cells were incubated with PE-labeled rat anti-mouse F4/80 antibody (clone BM8; Sungene), and F4/80+ macrophages were sorted by flow cytometry (Becton Dickinson). For adoptive transfer, 2 × 105 F4/80+ macrophages were injected into tumors of the right hind legs of mice, whereas PBS (vehicle control) was injected into the left hind leg tumors in the bilateral 4T1 tumor mouse model. After 24 hours, mice were injected with GO(HPPH)-PEG, Evans Blue dye (Alfa Aesar), IRDye800-labeled human serum albumin (Dye-HSA), or 125I-HSA for further studies.
In vivo optical imaging of GO(HPPH)-PEG
4T1 tumor-bearing mice (n = 5/group) were administered GO(HPPH)-PEG (30 nmol of HPPH equivalent) by intravenous injection. In vivo NIRF imaging was performed at 6, 24, and 48 hours postinjection (p.i.) using a Maestro In-Vivo Imaging System (CRI) with excitation and emission wavelengths of 675 and 720 nm, respectively. Tumor uptake was measured using Meastro 2.4 software (CRI) and presented as fluorescence intensity of total radiant efficiency (photons/s/cm2). The fluorescence intensity of tumors was normalized by the total injected dose of GO(HPPH)-PEG.
In vivo PDT of GO(HPPH)-PEG
4T1 tumor-bearing mice with adoptive macrophage transfer were randomly divided into three groups (n = 4/group) 24 hours after intratumoral injection of F4/80+ macrophages into the right hind leg tumor of each mouse. Mice in the three groups received (i) intravenous injection of PBS; (ii) intravenous injection of GO(HPPH)-PEG; or (iii) intravenous injection of GO(HPPH)-PEG, followed by laser irradiation (670 nm) of tumors on both sides at 24 hours p.i. at 70 J/cm2. At the end of the PDT study (day 12), mice were euthanized and tumors were harvested, weighed, and photographed. The tumor tissue was then dissected and embedded in optimal cutting temperature medium, and 5-μm frozen sections were cut for immunofluorescence staining of Ki67 and terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL).
In vivo permeability assay
4T1 tumor-bearing mice (n = 4) with adoptive macrophage transfer were intravenously injected with 1 mg Evans Blue at 24 hours after intratumoral injection of F4/80+ macrophages into the right hind leg tumor. At 2 hours p.i. of Evans Blue, mice were perfused with PBS and the muscle and tumors on both the left and right sides were harvested, weighed, and photographed. To quantify Evans Blue staining, tissue samples were incubated overnight with 1 mL of 2,2 N-methylformamide overnight in a 37°C water bath with shaking. After centrifugation, the supernatant was collected, and the absorbance at 600 nm was measured with a spectrophotometer (Thermo Fisher Scientific).
In vivo NIRF and SPECT/CT imaging of HSA
The HSA NIRF imaging probe was prepared according to a previously described protocol (30, 31). Briefly, HSA was mixed with IRDye800-N-hydroxysuccinimide (NHS) (LI-COR, Inc.) at a molar ratio of 1: 10 in sodium bicarbonate solution (pH 8.4). After a 2-hour reaction, Dye-HSA was purified using a PD-10 desalting column (GE Healthcare). 125I-HSA was synthesized as described previously (31), and its purity after radiolabeling and subsequent purification with a PD-10 column was >98%, as determined by instant thin-layer chromatography.
In the adoptive macrophage transfer 4T1 tumor mouse model, Dye-HSA (2 nmol of IRDye800 equivalence) or 125I-HSA (25.9 MBq) was administered via intravenous injection 24 hours after injection of F4/80+ macrophages into the right tumor. NIRF imaging and SPECT/CT were performed at 2 hours p.i. using the Maestro In Vivo Imaging System (CRI) and a NanoScan SPECT/CT Imaging System (Mediso), respectively.
Microdistribution of GO(HPPH)-PEG
4T1 tumor-bearing mice were intravenously injected with GO(HPPH)-PEG (30 nmol of HPPH equivalent) and euthanized 24 hours later, followed by harvesting of tumor tissue, which was cut into frozen sections. Integrin β6, F4/80, and CD31 expression was detected by IHC to identify 4T1 tumor cells, infiltrating macrophages, and tumor vasculature, respectively. Briefly, after blocking with 10% FBS in PBS, sections were stained with anti-mouse F4/80 (Abcam), anti-mouse integrin β6 (R&D Systems), and anti-mouse CD31 (BD Biosciences) primary antibodies and then fluorophore-conjugated secondary antibodies, followed by visualization under a confocal microscope (Leica).
Chemotherapy
4T1 tumor-bearing mice were treated with one of four commonly used chemotherapy drugs: cyclophosphamide (CTX), docetaxel, doxorubicin, or 5-fluorouracil (5-FU). Mice (n = 10/group) were randomly divided into treatment and control groups. For each drug, we used the routinely used doses, routes, and frequencies according to the literature. The treatment regimens were as follows: cyclophosphamide (150 mg/kg in PBS; a single intraperitoneal dose; ref. 7), docetaxel (10 mg/kg in 13% ethanol; every other day for a total of three intraperitoneal injections; ref. 32), doxorubicin (5 mg/kg in PBS; a single intravenous dose; ref. 33), or 5-FU (25 mg/kg in PBS; every other day for a total of three intraperitoneal injections; ref. 34). Control groups were administered vehicle. On day 9, 5 mice from each group were euthanized, and their tumors were harvested to obtain single-cell suspensions for flow cytometry analysis.
Flow cytometry analysis
Tumors harvested from mice were digested to single-cell suspensions and stained with PE-labeled rat anti-mouse F4/80 antibody (clone BM8; Sungene). Cells were then analyzed using a flow cytometer (Becton Dickinson).
In vivo NIRF imaging of GO(HPPH)-PEG before and after cyclophosphamide treatment
4T1 tumor-bearing mice (n = 5/group) were treated with cyclophosphamide (150 mg/kg) via intraperitoneal injection on day 1 and intravenously injected with GO(HPPH)-PEG (30 nmol of HPPH equivalent) on days 0 and 9. Optical imaging was performed at 24 hours p.i. (days 1 and 10) using the Maestro In-Vivo Imaging System as described above.
Cyclophosphamide treatment combined with GO(HPPH)-PEG PDT
For all PDT studies, 4T1 tumor-bearing mice were intravenously injected with GO(HPPH)-PEG (30 nmol of HPPH equivalent) and irradiated at 24 hours p.i. with a 670-nm laser at 70 J/cm2. The mice were randomly divided into five groups (n = 5/group): (i) control (PBS); (ii) PDT on day 1; (iii) intraperitoneal injection of cyclophosphamide (150 mg/kg) on day 1; (iv) intraperitoneal injection of cyclophosphamide (150 mg/kg) on day 1 and PDT on day 1; and (v) intraperitoneal injection of cyclophosphamide (150 mg/kg) on day 1 and PDT on day 10. Tumor growth and body weight were monitored every other day. Mice were euthanized and their tumors harvested for analysis of Ki67 expression on day 22.
In vivo NIRF imaging and SPECT/CT imaging of macrophage infiltration
The F4/80-specific NIRF probe Dye-αF4/80-Fab and radiotracer 99mTc-αF4/80-Fab were prepared by directly labeling the Fab fragment of a rat anti-mouse F4/80 mAb (αF4/80; Bio X Cell) with IRDye800-NHS or via conjugation with the bifunctional chelator HYNIC-NHS, followed by labeling with 99mTc as described in Supplementary Materials and Methods. For in vivo NIRF imaging, 4T1 tumor-bearing mice were administered Dye-αF4/80-Fab (2 nmol of IRDye800 equivalence) by intravenous injection and then imaged at 6 hours p.i. using the Maestro In-Vivo Imaging System. For SPECT/CT imaging, 4T1 tumor-bearing mice were intravenously injected with 22.2 MBq 99mTc-αF4/80-Fab, and SPECT/CT was performed at 6 hours p.i. using the NanoScan SPECT/CT Imaging System (Mediso).
Immunofluorescence staining
For Ki67 staining, tumor sections were incubated with rabbit anti-mouse Ki67 antibody (Millipore) for 2 hours at room temperature and then visualized by incubation of fluorescently conjugated secondary antibody (Earthox) under a confocal microscope (Leica). For TUNEL assay, tumor sections were conducted using In Situ Cell Death Detection Kit (Roche). After staining, 10 random views of each tumor section were selected for the quantification. Tumor cell proliferation index and apoptosis were described as the percentage of Ki67- and TUNEL-positive nucleus in proportion of the total number of nucleus, respectively.
Statistical analysis
Quantitative data were are presented as mean ± SD. Statistical analysis was done using a one-way ANOVA and an unpaired Student t test. P < 0.05 were considered statistically significant.
Results
Adoptive macrophage transfer increases tumor accumulation and PDT effect of GO(HPPH)-PEG
We synthesized GO(HPPH)-PEG according to a previously published protocol (28). HPPH was loaded onto the surface of GO-PEG to form the GO(HPPH)-PEG complex via π–π stacking (Supplementary Fig. S1A) as a photodynamic agent for in vivo cancer treatment. Single-layered carbon nanosheets of uniform size and thickness were observed by atomic force microscopy (Supplementary Fig. S1B). Characteristic HPPH peaks were present in the fluorescence spectra of GO(HPPH)-PEG, confirming the successful loading of HPPH onto the nanographene (Supplementary Fig. S1C). GO(HPPH)-PEG exhibited excellent stability in a range of physiologic solutions, including water, PBS, FBS, and cell culture medium (Supplementary Fig. S1D). In addition, HPPH was firmly absorbed in the GO surface, and no significant free HPPH was released from GO(HPPH)-PEG up to 96 hours after incubating in different solutions (Supplementary Fig. S2). The PDT effect of GO(HPPH)-PEG was confirmed by the in vitro singlet oxygen sensor green assay (Supplementary Fig. S3A and S3B).
In vivo NIRF imaging studies showed a significantly higher tumor uptake of GO(HPPH)-PEG as compared with that of free HPPH at all time points examined (P < 0.01; Supplementary Fig. S4A and S4B), confirming improved tumor accumulation by loading HPPH onto the surface of GO-PEG. Because the highest tumor uptake of GO(HPPH)-PEG was observed at 24 hours p.i. (Supplementary Fig. S4B), we performed the PDT studies at 24 hours p.i. of GO(HPPH)-PEG in subsequent PDT studies. In vivo PDT results showed a significant effect on tumor suppression by GO(HPPH)-PEG, followed by irradiation with 70 and 90 J/cm2 as compared with control groups and PDT with 20 J/cm2 (Supplementary Fig. S5A). Notably, GO(HPPH)-PEG–based PDT with 70 and 90 J/cm2 showed comparable antitumor efficiency, suggesting that 70 J/cm2 resulted in a favorable PDT effect and correlating with the administered dose of GO(HPPH)-PEG (30 nmol HPPH equivalent). A significantly higher antitumor efficacy was observed in the GO(HPPH)-PEG group as compared with that observed using free HPPH (Supplementary Fig. S5B), further confirming the advantages of GO-PEG coating for improving HPPH tumor targeting and PDT effects.
TAMs contribute to the delivery of nanoparticles into tumors (25, 26). To investigate how TAMs impact tumor uptake of GO(HPPH)-PEG, F4/80+ TAMs were sorted from donor 4T1 tumor-bearing mice and injected into the right tumors of bilateral 4T1 tumor-bearing mice (Fig. 1A and B). By labeling the sorted F4/80+ TAMs with either IRDye800 or DiI, their behaviors within the tumor after intratumoral injection were able to be visualized by either in vivo NIRF imaging or ex vivo tissue examination. The results demonstrated tumor retention of the injected macrophages for up to 72 hours (Supplementary Fig. S6), as well as their even dispersal and distribution within the tumor tissue after 24 hours (Supplementary Fig. S7).
Enhanced tumor uptake and PDT effect of GO(HPPH)-PEG by adoptive macrophage transfer. A, Schematic illustration of macrophage sorting and adoptive transfer and PDT. B, Representative dot plots of F4/80+ macrophages sorted by flow cytometry. C, Representative NIRF images of GO(HPPH)-PEG in mice bearing 4T1 tumors injected with PBS or macrophages (MΘ) at 6, 24, and 48 hours p.i. D, Quantitative analysis of GO(HPPH)-PEG uptake in tumors with or without macrophage transfer. Data are expressed as mean ± SD, n = 5. E, Growth curves of 4T1 tumors in mice with (MΘ/) or without (PBS/) adoptive macrophage transfer after the indicated treatments, PBS, GO(HPPH)-PEG alone, or GO(HPPH)-PEG plus light irradiation. F and G, Photographs (F) and measured weight (G) of tumors harvested from 4T1 tumor-bearing mice after the indicated treatments. Data are expressed as mean ± SD, n = 4. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
After F4/80+ TAM intratumoral injection, mice were administered GO(HPPH)-PEG and observed by NIRF imaging at 6, 24, and 48 hours p.i. GO(HPPH)-PEG uptake was markedly higher in the right (macrophage-transferred) tumor than in the left tumor at all time points examined (P < 0.05; Fig. 1C and D). Ex vivo NIRF imaging of tumors and organs at 24 hours p.i. of GO(HPPH)-PEG revealed that HPPH fluorescence was significantly higher in tumors injected with macrophages as compared with those injected with PBS (P < 0.01; Supplementary Fig. S8A and S8B). These results demonstrate that adoptive macrophage transfer increased tumor accumulation of GO(HPPH)-PEG.
These results suggested that macrophages could enhance the PDT effect of GO(HPPH)-PEG. We used 4T1 bilateral tumor-bearing mice for in vivo PDT studies to investigate this possibility. GO(HPPH)-PEG PDT inhibited tumor growth as compared with tumors without PDT (P < 0.05; Fig. 1E). Furthermore, macrophage transfer followed by GO(HPPH)-PEG PDT more potently inhibited tumor growth as compared with that observed in tumors without macrophage transfer (P < 0.01; Fig. 1E) and according to confirmation by the lower weight and smaller size of the tumors harvested from mice (P < 0.01; Fig. 1F and G).
To further evaluate the impact of macrophage transfer on the PDT effect of GO(HPPH)-PEG, tumor cell proliferation and apoptosis after PDT were assessed by histologic analysis of Ki67 immunofluorescence staining and the TUNEL assay, respectively. There were fewer proliferative cells in tumors with GO(HPPH)-PEG PDT than in those without PDT (P < 0.01; Fig. 2A and B). In addition, tumors with macrophage transfer followed by GO(HPPH)-PEG PDT showed significantly fewer proliferative cells as compared with tumors treated with GO(HPPH)-PEG PDT without macrophage transfer (P < 0.001; Fig. 2A and B). Furthermore, tumors injected with macrophages followed by GO(HPPH)-PEG PDT showed the highest number of apoptotic cells (P < 0.01; Fig. 2C and D). Therefore, macrophages within tumors increased tumor uptake of GO(HPPH)-PEG, subsequently enhancing the effect of PDT on tumor growth inhibition.
Ex vivo histologic analysis of proliferating and apoptotic cells in 4T1 tumor tissues harvested from mice (left and right tumors of each mouse underwent transfer of PBS and macrophages, respectively) after the indicated treatments, PBS, GO(HPPH)-PEG alone, or GO(HPPH)-PEG plus light irradiation. A and B, Representative images of Ki67 staining (A) and quantitative analysis of Ki67-positive (B) cells after the indicated treatments. C and D, Representative images (C) and quantitative analysis (D) of TUNEL-positive cells after the indicated treatments. All error bars are expressed as ± SD, n = 10. **, P < 0.01; ***, P < 0.001.
Mechanisms of enhanced tumor uptake and PDT effect of GO(HPPH)-PEG by adoptive macrophage transfer
Given that GO(HPPH)-PEG accumulated in tumors based on EPR effect (29), we hypothesized that enhanced tumor uptake of GO(HPPH)-PEG after adoptive macrophage transfer resulted, at least in part, from an improvement in the EPR effect. We used the Evans Blue assay (35) to evaluate tumor microvascular permeability 24 hours after macrophage transfer. Tumors injected with macrophages had a deeper color than those injected with PBS (Fig. 3A), and the optical density at 600 nm of Evans Blue extracted from macrophage-injected tumors was significantly higher than that from PBS-treated tumors (P < 0.01; Fig. 3B). To confirm these findings, HSA (65 kDa) (36, 37) was labeled with either IRDye800 dye or radioisotope and observed by NIRF imaging and SPECT/CT. The uptake of Dye-HSA was higher in tumors undergoing macrophage transfer than in those treated with PBS (P < 0.01; Fig. 3C and D). Because NIRF imaging has inherent limitations in deep-tissue penetration, we validated these results by SPECT/CT imaging using the radiolabeled HSA (125I-HSA). The accumulation of 125I-HSA was markedly higher in macrophage-transferred tumors than in PBS-treated tumors (Fig. 3E). These results indicated that macrophage transfer increased tumor microvascular permeability, which might have been attributed to inflammation induced by macrophages or cytokines, such as VEGF released by macrophages in the tumor microenvironment (27), and thereby enhancing tumor accumulation of GO(HPPH)-PEG.
Adoptive macrophage transfer increased tumor vasculature permeability, and GO(HPPH)-PEG can be phagocytosed by tumor-infiltrated macrophages. A, Photographs of muscles and tumors (with or without macrophage transfer) after Evans blue permeability assay. B, Quantitative analysis of the amount of Evans blue in muscles and tumors. C and D, Representative NIRF images (C) and quantitative analysis (D) of tumor uptake of IRDye800-HSA at 2 hours p.i. in mice bearing 4T1 tumors with or without macrophage transfer. E, Representative SPECT/CT images of 125I-HSA at 2 hours p.i. in mice bearing 4T1 tumors with or without adoptive macrophage transfer. F and G, Immunofluorescence analysis of integrin β6 (to detect 4T1 tumor cells), F4/80 (to detect TAMs), and CD31 (to detect tumor vasculature) in 4T1 tumor tissues harvested from mice 24 hours after injection of GO(HPPH)-PEG. H and I, Representative dot plots (H) and quantitative analysis (I) of F4/80+ macrophages in tumors with or without adoptive macrophage transfer, as determined by flow cytometry. Tumors in the NIRF and SPECT images are indicated by circles. All error bars expressed as ± SD, n = 4. **, P < 0.01; ***, P < 0.001.
Some nanoparticles are taken up by macrophages via phagocytosis (23, 24). Recent studies showed that macrophages can serve as a drug reservoir that improves drug delivery to tumors (26). To determine whether macrophages could increase tumor accumulation of GO(HPPH)-PEG by phagocytosis, we examined GO(HPPH)-PEG localization in tumor tissues 24 hours after intravenous injection. Tumor tissue samples were labeled with antibodies against integrin β6, F4/80, and CD31 to identify 4T1 tumor cells, macrophages, and tumor vasculature, respectively. 4T1 tumor cells are positive for integrin β6 (Supplementary Fig. S9; ref. 30), which is not expressed in normal tissues (38). We observed that GO(HPPH)-PEG colocalized with F4/80 but not with integrin β6 (Fig. 3F), indicating that it was taken up by macrophages in tumor tissues, but not by tumor cells. GO(HPPH)-PEG also accumulated around tumor vasculature, as evidenced by the colocalization of HPPH fluorescence with CD31 (Fig. 3G). Flow cytometry analysis revealed that F4/80+ macrophage infiltration increased in tumors injected with macrophages as compared with those treated with PBS (P < 0.001; Fig. 3H and I). Therefore, macrophages in tumors increased GO(HPPH)-PEG phagocytosis and accumulation within tumors.
Chemotherapy enhances tumor uptake and the PDT effect of GO(HPPH)-PEG
It was reported that cyclophosphamide treatment increased macrophage infiltration into tumors (33). Therefore, we investigated whether cyclophosphamide pretreatment could enhance tumor uptake and the PDT effect of GO(HPPH)-PEG in 4T1 tumor-bearing mice administered a single dose of cyclophosphamide on day 1 immediately after baseline NIRF imaging of GO(HPPH)-PEG, followed by imaging again on day 10. Baseline and post-cyclophosphamide flow cytometry detection of the F4/80+ macrophages was also performed on days 0 and 9, respectively (Fig. 4A). Tumor uptake of GO(HPPH)-PEG post-cyclophosphamide treatment was increased relative to the baseline (P < 0.05; Fig. 4B and C) and was also significantly higher than that in tumors treated with vehicle control (P < 0.05; Supplementary Fig. S10A—S10C). We also confirmed that cyclophosphamide treatment increased the percentage of F4/80+ macrophages after cyclophosphamide treatment relative to baseline percentages by flow cytometry (P < 0.01; Fig. 4D and E) and compared with vehicle-treated tumors by immunofluorescence staining (Supplementary Fig. S11). These results suggested that the combination of cyclophosphamide treatment and GO(HPPH)-PEG PDT would enhance antitumor effects by increasing tumor uptake of GO(HPPH)-PEG.
Enhanced tumor uptake and PDT effect of GO(HPPH)-PEG in 4T1 tumor-bearing mice after treatment with cyclophosphamide (CTX). A, Schematic illustration of cyclophosphamide treatment, flow cytometry analysis, and NIRF imaging of GO(HPPH)-PEG. B and C, Representative NIRF images (B) and quantitative analysis of tumor uptake (C) of GO(HPPH)-PEG before and after cyclophosphamide treatment at 24 hours p.i. (n = 5/group). Tumors are indicated with circles. D and E, Representative dot plots (D) and quantitative analysis (E) of F4/80+ macrophages in 4T1 tumors from mice before and after cyclophosphamide treatment, as determined by flow cytometry (n = 5). F, Schematic illustration of cyclophosphamide and GO(HPPH)-PEG PDT combination therapy. G, Tumor growth curves after the indicated treatments (n = 5/group). H and I, Immunofluorescence detection (H) and quantitative analysis (I) of Ki67-positive cells in tumor tissues harvested from mice after the indicated treatments (n = 10). J, Changes in body weight of 4T1 tumor-bearing mice after the indicated treatments (n = 5/group). All error bars expressed as ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
We then investigated whether cyclophosphamide combined with GO(HPPH)-PEG PDT exhibited synergistic inhibitory effects on tumor growth. 4T1 tumor-bearing mice were treated with PBS alone, GO(HPPH)-PEG PDT (on day 1), cyclophosphamide alone, cyclophosphamide plus GO(HPPH)-PEG PDT (PDT on day 1, denoted as cyclophosphamide +PDT-1), or cyclophosphamide plus GO(HPPH)-PEG PDT (PDT on day 10, denoted as CTX + PDT-2; Fig. 4F). Compared with the PBS control, PDT or cyclophosphamide alone partially suppressed tumor growth, whereas CTX + PDT-1 showed a similar antitumor effect as compared with cyclophosphamide alone, which was likely because cyclophosphamide had not yet induced macrophage infiltration on day 1. In contrast, tumor growth was almost completely inhibited in the CTX + PDT-2 group, in which tumor sizes were smaller than those in the CTX + PDT-1 group from day 14 up to the end of the study (day 22; Fig. 4G). The tumor cell proliferation index (%Ki67-positive cells) in the CTX + PDT-2 group (1.79% ± 1.37 %) was significantly lower than that in the groups treated with PDT alone (14.70% ± 2.49%), cyclophosphamide alone (9.23% ± 1.97%), and CTX + PDT-1 (7.42% ± 3.12%; P < 0.001; Fig. 4H and I). Moreover, the body weight of tumor-bearing mice decreased by <20% as a result of combination therapy (Fig. 4J), indicating that this treatment was well tolerated. These results demonstrated that cyclophosphamide pretreatment in combination with GO(HPPH)-PEG PDT effectively inhibited tumor growth by promoting macrophage recruitment and infiltration into tumors, with consequent accumulation of GO(HPPH)-PEG.
Noninvasive imaging of macrophage infiltration into tumors following treatment with chemotherapeutic drugs
We then investigated whether other chemotherapeutic drugs could induce macrophage infiltration into tumors and whether GO(HPPH)-PEG PDT can be used as a universal strategy for treatment of tumors following chemotherapy by testing the effects of the cancer chemotherapy drugs docetaxel, doxorubicin, and 5-FU on macrophage infiltration in 4T1 tumor-bearing mice. Flow cytometry analysis revealed an increase in the percentage of F4/80+ macrophages in tumors after docetaxel, doxorubicin, and 5-FU treatment (Fig. 5A–F). These results demonstrated that chemotherapeutic drug treatment enhanced macrophage infiltration into tumors, highlighting their potential for use in combination with GO(HPPH)-PEG PDT to increase antitumor efficacy.
Noninvasive visualization of macrophage infiltration into tumors after chemotherapy by NIRF and SPECT/CT imaging. A–F, Representative dot plots (A–C) and quantitative analysis (D–F) of F4/80+ macrophages by flow cytometry in tumors treated with vehicle (control), docetaxel (DTX), doxorubicin (DOX), or 5-FU. G and H, Representative NIRF images (G) and quantitative analysis of tumor uptake (H) of Dye-αF4/80-Fab at 6 hours p.i. in 4T1 tumor-bearing mice after treatment with PBS (control), cyclophosphamide, docetaxel, doxorubicin, or 5-FU. I, Representative SPECT/CT images of 99mTc-αF4/80-Fab at 6 hours p.i. in mice before (baseline) and after treatment with vehicle (control), cyclophosphamide, or 5-FU. Tumors are indicated by circles. All error bars expressed as ± SD, n = 5. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
We also investigated whether infiltration into tumors after chemotherapy could be noninvasively detected by NIRF imaging and SPECT/CT. We synthesized macrophage-targeting probes including the NIRF and radiolabeled probes Dye-αF4/80-Fab and 99mTc-αF4/80-Fab, respectively (Supplementary Fig. S12). The high purity of Dye-αF4/80-Fab was confirmed by SDS-PAGE followed by NIRF imaging (Supplementary Fig. S13). The macrophage-targeting specificity of Dye-αF4/80-Fab was confirmed by blocking experiments using F4/80+ RAW 264.7 cells (Supplementary Fig. S14A and S14B) and by in vivo NIRF imaging (Supplementary Fig. S15A and S15B). The in vitro macrophage-binding specificity of 99mTc-αF4/80-Fab was confirmed by a cell-binding assay using F4/80+ RAW 264.7 cells (Supplementary Fig. S16).
These two probes were then used for in vivo visualization of macrophage infiltration into tumors following chemotherapy. Tumor uptake of Dye-αF4/80-Fab was significantly increased 9 days after treatment with cyclophosphamide, docetaxel, doxorubicin, or 5-FU relative to control tumors at 6 hours after probe injection (P < 0.05; Fig. 5G and H). To validate these observations, we performed small-animal SPECT/CT using 99mTc-αF4/80-Fab. Tumor uptake of 99mTc-αF4/80-Fab increased following cyclophosphamide and 5-FU treatment as compared with baseline levels and vehicle control (Fig. 5I). Therefore, chemotherapy drugs increased macrophage infiltration into tumors, which was noninvasively visualized by NIRF imaging and SPECT/CT using F4/80-targeting probes. These findings suggested that GO(HPPH)-PEG PDT in combination with chemotherapy is a promising treatment strategy, with effects capable of being monitored noninvasively with Dye-αF4/80-Fab and 99mTc-αF4/80-Fab.
Discussion
TAMs play important roles in promoting tumor progression via angiogenesis, invasion, and metastasis and by suppressing anticancer immune responses. Several studies showed that chemotherapy can increase TAM recruitment to tumors, leading to tumor resistance and recurrence after chemotherapy (33, 39, 40). Although significant efforts have been made to deplete TAMs using various strategies, few studies have examined how TAMs can be exploited for tumor treatment. In this study, we investigated whether macrophage recruitment to tumors could affect the PDT effect of GO(HPPH)-PEG. We observed that adoptive transfer of F4/80+ macrophages into tumors increased tumor accumulation and the PDT effect of GO(HPPH)-PEG (Fig. 6A). Interestingly, significantly increased tumor F4/80+ macrophage infiltration can also be induced by treatment with several commonly used chemotherapeutic drugs. Therefore, the tumor-suppressive effects of chemotherapy could be further enhanced by GO(HPPH)-PEG PDT (Fig. 6B). In addition, NIRF imaging and SPECT using macrophage-specific imaging probes allowed sensitive visualization of postchemotherapy macrophage infiltration into tumors (Fig. 6C). Our results demonstrated that chemotherapy combined with GO(HPPH)-PEG PDT is a promising strategy for the treatment of tumors, especially those resistant to chemotherapy. Furthermore, molecular imaging of tumor-infiltrating macrophages using NIRF and SPECT could guide the rationale design of combination therapy.
Summary of major findings of this study. A, Adoptive transfer of F4/80+ macrophage into tumors enhanced the tumor uptake and PDT effect of GO(HPPH)-PEG. B, Chemotherapy induced macrophage recruitment to tumors and led to improved tumor uptake and PDT effect of GO(HPPH)-PEG. C, Tumor macrophage infiltration induced by chemotherapy was noninvasively visualized by F4/80-targeted molecular imaging.
TAMs have been shown to accumulate drugs that are loaded onto nanoparticles and then to release them into neighboring tumor cells (26). We demonstrated that enhanced tumor accumulation of GO(HPPH)-PEG after adaptive macrophage transfer resulted from phagocytosis of GO(HPPH)-PEG and enhanced tumor perfusion by macrophages. Because tumor uptake of nanoparticles was closely associated with the degree of TAM infiltration, we speculated that noninvasively imaging TAMs would allow us to predict the amount of nanoparticles taken up by tumors. Indeed, we successfully detected tumor-infiltrated macrophages in a mouse model in vivo using murine macrophage-specific NIRF and SPECT/CT imaging. Note that we only tested the imaging results in one tumor model with similar macrophage levels and vasculature status. Anatelli and colleagues (41) previously developed a maleylated albumin–photosensitizer conjugate to specifically target macrophages via the scavenger receptor, which is overexpressed in TAMs. The accumulated levels of the photosensitizer in the tumors correlated well with tumor TAM levels in three tumor models exhibiting different degrees of macrophage infiltration (41). Therefore, our future studies should investigate whether the tumor uptake values of TAM-specific probes in various tumors with different macrophage levels are linearly correlated with the accumulated levels of GO(HPPH)-PEG as determined by NIRF imaging and also correlated with TAM recruitment levels as measured by flow cytometry.
Several approaches have been described to increase EPR to improve the delivery of macromolecules to tumors, including normalizing tumor vasculature (42), modulating the tumor matrix (18), and coapplication of PDT, hyperthermia, radiotherapy, and sonoporation (43), which in most cases requires additional treatment agents. In this study, we demonstrated that chemotherapy-induced macrophage infiltration can serve as an in situ factor to enhance the EPR effect of tumors, thereby leading to increased tumor accumulation of GO(HPPH)-PEG and consequently, increased PDT efficacy upon light irradiation. We found that macrophage infiltration was increased in the presence of several commonly used chemotherapy drugs, including cyclophosphamide, docetaxel, doxorubicin, and 5-Fu, which enhanced tumor delivery of GO(HPPH)-PEG. A variety of previous studies also demonstrated that tumor exposure to chemotherapeutic drugs promotes TAM infiltration of tumors (20, 44). This phenomenon results from multiple mechanisms, including chemotherapy results in tumor hypoxia, which increases hypoxia-inducible factor-1α expression and subsequently promotes high expression of C-X-C motif chemokine ligand (CXCL)12. CXCL12, in turn, recruits circulating bone marrow–derived monocytes to differentiate into TAMs in tumors by recognizing its receptor (CXCR4) expressed on bone marrow–derived monocytes. Another mechanism includes chemotherapy-induced increases in colony-stimulating factor-1 and C-C motif chemokine ligand-2 expression (44, 45). The rapid recruitment of TAMs in tumors after chemotherapy promotes tumor progression and limits the efficiency of chemotherapy.
Several preclinical and clinical studies also demonstrated the synergistic effects of combining PDT and chemotherapy for improving tumor responses via multiple mechanisms (4). Gallagher-Colombo and colleagues (46) showed that pretreating tumors with an EGFR inhibitor (erlotinib) significantly improves the antitumor efficacy of PDT, even in erlotinib-resistant tumors. It was also reported that pretreatment with PDT markedly increases intracellular concentrations of nanoliposomal irinotecan, resulting in elevated inhibition of tumor growth as compared with monotherapies (6). These results suggested the potential of combination strategies for PDT and conventional antitumor drugs. Using different mechanisms, we took advantage of chemotherapy-induced recruitment of TAMs within tumors to increase the tumor accumulation of GO(HPPH)-PEG, which in turn improved the antitumor efficiency of GO(HPPH)-PEG-based PDT. This strategy would be especially suitable for the treatment of tumors that have acquired resistance to chemotherapy.
Notably, PDT is typically used to treat superficial cancers due to the limited penetration of light through tissues, which has ranges in the depths allowing tumor destruction of a few millimeters to one centimeter (4). Therefore, the combined chemotherapy plus PDT strategy described in the current study might only be suitable for superficial cancers, such as skin carcinoma and breast cancer. Recent studies suggested that tumor cells destroyed by local PDT could serve as an in situ tumor vaccine that stimulates dendritic cell maturation and primes cytotoxic T cells to attack metastatic and other lesions in an effect potentiated by immune checkpoint inhibitors (31, 47). Therefore, PDT of GO(HPPH)-PEG after chemotherapy might be further improved by combination with checkpoint blockade immunotherapy (48).
The results presented here have implications for the development and clinical application of PDT based on other types of nanomaterials in combination with chemotherapy. Because tumor-infiltrated macrophages can be noninvasively detected by macrophage-specific NIRF and SPECT/CT imaging, selecting patients most likely to benefit from nanoparticle-based PDT combined with chemotherapy is guaranteed. In addition, this method might also be extended to other diseases that involve local macrophage recruitment, such as atherosclerosis and myocardial infarction. It should be noted that F4/80 is a marker of mouse macrophages, and we used F4/80-targeted imaging probes for in vivo imaging in this proof-of-concept mouse model study to demonstrate the role of molecular imaging of macrophages for predicting chemotherapy-induced macrophage recruitment into tumors. In humans, CD68, CD163, CD16, CD312, and CD115 are the major markers of macrophages (49); therefore, for future clinical translation, antibodies targeting human macrophage-specific biomarkers are needed to design and synthesize related molecular imaging probes for immuno-PET or immunoSPECT imaging (50).
In summary, we found that increased recruitment of TAMs to tumors enhanced the tumor uptake and PDT effect of GO(HPPH)-PEG by increasing tumor vascular permeability and GO(HPPH)-PEG phagocytosis. GO(HPPH)-PEG enhanced the efficacy of chemotherapy for inhibiting tumor growth, offering a potential solution for treating tumors resistant to chemotherapy. In addition, detection of TAM infiltration into tumors by noninvasive NIRF imaging or SPECT/CT could inform the design of combination regimens suited to each patient.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Authors' Contributions
Conception and design: Y. Zhao, Z. Liu
Development of methodology: Y. Zhao, C. Zhang, X. Yu, J. Lai, Z. Liu
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Zhao, C. Zhang, L. Gao, X. Yu, J. Lai, D. Lu, R. Bao, Y. Wang, B. Jia, F. Wang
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Zhao, C. Zhang, L. Gao, J. Lai, Z. Liu
Writing, review, and/or revision of the manuscript: Y. Zhao, C. Zhang, Z. Liu
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L. Gao, X. Yu, D. Lu, R. Bao, Y. Wang
Study supervision: F. Wang, Z. Liu
Grant Support
This work was supported, in part, by National Basic Research Program (973 Program) of China (2013CB733802 to Z. Liu) and the National Natural Science Foundation of China (81471712 and 81671747 to Z. Liu; 81630045 and 81420108019 to F. Wang).
Footnotes
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
- Received June 6, 2017.
- Revision received August 16, 2017.
- Accepted August 31, 2017.
- ©2017 American Association for Cancer Research.
References
Volume 77, Issue 21
