Fluorinated PLGA-PEG-Mannose Nanoparticles for Tumor-Associated Macrophage Detection by Optical Imaging and MRI

Tumor-associated macrophages (TAMs) promote cancer growth and metastasis, but their role in tumor development needs to be fully understood due to the dynamic changes of tumor microenvironment (TME). Here, we report an approach to visualize TAMs by optical imaging and by Fluorine-19 (19F) magnetic resonance imaging (MRI) that is largely applied to track immune cells in vivo. TAMs are targeted with PLGA-PEG-mannose nanoparticles (NPs) encapsulating perfluoro-15-crown-5-ether (PFCE) as MRI contrast agent. These particles are preferentially recognized and phagocytized by TAMs that overexpress the mannose receptor (MRC1/CD206). The PLGA-PEG-mannose NPs are not toxic and they were up-taken by macrophages as confirmed by in vitro confocal microscopy. At 48 h after intravenous injection of PLGA-PEG-mannose NPs, 4T1 xenograft mice were imaged and fluorine-19 nuclear magnetic resonance confirmed nanoparticle retention at the tumor site. Because of the lack of 19F background in the body, observed 19F signals are robust and exhibit an excellent degree of specificity. In vivo imaging of TAMs in the TME by 19F MRI opens the possibility for detection of cancer at earlier stage and for prompt therapeutic interventions in solid tumors.


INTRODUCTION
Inflammation is one the major effect of cancer and it plays a pivotal role in cancer progression and metastasis (1). In healthy conditions, macrophages (Mφ) exert pro-inflammatory and cytotoxic effect leading the immune response against tumor development (2). In solid tumors, tumor associated-macrophages (TAMs) are generally skewed away from the classical activation toward an alternative tumor promoting phenotype and becoming the major constituent of tumor malignancy (3,4). Thus, presence of TAMs in the tumor microenvironment (TME) is correlated with increased tumor metastasis, angiogenesis, and tumor aggressiveness (5). In recent studies, histological sample of necrotic breast cancers have shown high tumor-associated macrophage infiltration correlating with unfortunate prognosis (6). Indeed, TAMs can efficiently enter the necrotic core of the breast cancer and still functioning in hypoxic-necrotic areas. In this regard, the ability to label and observe TAMs non-invasively and over the time can tremendously help to understand the temporal and spatial localization of this population in the TME (7).
Recently, the magnetic resonance imaging (MRI) technique has been used to image inflammation and to track immune cells in vivo with no need of radiation (8). In particular, perfluorocarbons (PFCs) are emerging as promising contrast agents for MRI cell tracking (9,10). This is because, fluorinebased contrast agents are found only in traces in biological tissue meaning that the fluorine background is minimal and that the signal from exogenous fluorine is highly specific in vivo (10). Amongst PFCs, perfluoro-15-crown-5-ether (PFCE) is one of the most attractive MRI contrast agents because it is FDA approved in a form of emulsion and therefore it is not toxic (11). However, most of PFCs are not miscible with hydrophilic or hydrophobic solvents due to the strong carbonfluorine covalent bond and strong electron withdrawing effects of fluorine. Thus, PFCs are typically prepared as lipid-based nanoemulsions with low toxicity and longer circulation time (12). However, nano-drops of PFCs show limited stability in vivo due to the low affinity amongst the PFCs, the continuous phase and the surfactant (7). In general, the physical structure of nanoemulsion may also restrict the combination with other functional molecules such as drugs, fluorescent tracker or surface ligand for specific targeting. To this purpose, biodegradable organicbased nanocarriers like liposomes, dendrimers, micelles, and polymeric NPs act as protector and provide a good stability of the payload (13). In this context, different strategies can be used for tumor targeting and tumor imaging (14). For instance, "Passively targeted" nanoparticles (NPs) exploit solely the enhanced permeability and retention (EPR) effect and allow to target cancer systemically. However, the circulation of passively targeted NPs is often prevented by main physiological barriers: the extravasation of the tumor vasculature especially for high-EPR tumors that reduces nanocarrier accumulation; the NPs clearance by mononuclear phagocytic system (MPS), sinusoidal cells of the liver and Kupffer cells (15). On the contrary, "actively targeted" nanoparticles can help to overcome such barriers and to deliver greater amount of payload to the desired compartment thanks to the functionalization of the polymeric surface. Amongst nanocarriers, poly-lactic-coglycolic acid (PLGA) an FDA approved copolymer, is one of the most exploited system in pre-clinical research owed to its biodegradability, biosafety, biocompatibility, versatility in formulation and functionalization and long shelf-life (16,17).
Herein, we have focused on 19 F-based PLGA nanoparticles (NPs) to detect TAMs accumulation in humanized mice bearing breast cancer as tumor model (4T1 cells). To this purpose, PLGA NPs have been designed to encapsulate PFCE contrast agent and preserving its magnetic properties (18,19). In addition, the polymeric shell of PLGA has been functionalized with polyethylene glycol (PEG) chains that enhance the plasmatic half-life of PLGA NPs and prevents the rapid opsonization by the mononuclear phagocyte system (MPS) for in vivo purposes. To actively target tumor-associated macrophages, the surface of PLGA-PEG nanoparticles has been also decorated with mannosamine ligand that is preferentially recognized and internalized by TAMs overexpressing mannose receptors (CD206) (20). In addition, fluorescein isothiocyanate (FITC) has been linked to the polymeric shell of the PLGA-PEG NPs allowing further in vitro and ex vivo validations. All in all, intravenously injected mannose-decorated 19− F based-PLGA-PEG NPs aim to enhance targeting of recruited tumor-associated macrophages in a humanized mouse model of breast cancer by 19 F-MRI.

Proton Nuclear Magnetic Resonance Spectroscopy ( 1 H-NMR)
Chemical structures and number-average molecular weight (Mn) of synthesized polymers were characterized by proton nuclear magnetic resonance spectroscopy (1H NMR, Varian Mercury plus 400, Crawley, UK) using CDCl3 or D2O as solvents. Chemical shifts were referred to the solvent peak (δ = 7.26 ppm for CDCl3, δ = 4.79 ppm for D2O).

Gel Permeation Chromatography
Gel permeation chromatography (GPC) was employed to determine the weight average molecular weight (Mw), number average molecular weight (Mn) and the polydispersity index (PDI) of copolymers. GPC measurements were carried out by using a TSK gel G4000HHR column (Tosoh Bioscience, Tokyo, Japan), 7.8 mm ID × 30.0 cm L, pore size 5 µm. Polystyrenes of defined molecular weights ranging from 580 to 377,400 Da were used as calibration standards. The eluent was tetrahydrofuran (THF), the elution rate was 1.0 ml/min and the column temperature was 35 • C. The samples were dissolved in THF at a concentration of 5 mg/ml.

Fourier Transform Infrared Spectroscopy
Infrared spectra of the synthesized polymer were recorded by a Fourier transform infrared spectrophotometer (FT-IR, PerkinElmer, USA) at the wavelength range of 4,000-500 cm −1 . All the spectra were recorded against a background of an air spectrum.
Synthesis and Characterization of Polymers: PLGA-PEG-COOH, PLGA-PEG-FITC, and PLGA-PEG-Mannosamine Carboxyl terminal groups of PLGA were activated and converted to PLGA-NHS for the subsequent conjugation with polyethylene glycol (PEG). Briefly, 2 g PLGA 503H polymer was dissolved in 10 ml anhydrous dichloromethane (DCM) followed by adding an excess of N-hydroxy succinimide (NHS, 46.0 mg, 0.4 mmol) and N,N'-dicyclohexylcarbodiimide (DCC, 82.5 mg, 0.4 mmol). The reaction was stirred at room temperature overnight, under the N2 atmosphere. To purify, PLGA-NHS was precipitated in diethyl ether and washed by cold mixture of diethyl ether and methanol three times to remove the residual NHS.
Chemical Synthesis Scheme of PLGA-NHS PLGA-PEG copolymers with carboxyl terminal groups and amino terminal groups were synthesized by conjugated amino groups of NH 2 -PEG-COOH and NH2-PEG-NH 2 correspondingly to the N-hydroxysuccinimide esters of resulting PLGA-NHS. In details, PLGA-NHS (500 mg, 0.015 mmol) was dissolved in 5 ml anhydrous DCM. Then NH 2 -PEG-COOH or NH2-PEG-NH 2 (44.1 mg, 0.015 mmol) was added in the DCM solution with trimethylamine (TEA, 13.3 µl, 0.09 mmol) as catalyst. The reaction was processed at room temperature overnight, under N 2 atmosphere. PLGA-PEG copolymer was precipitated with a cold mixture of diethyl ether and methanol and washed three times by the same solvents, then dried by desiccator under vacuum. number-average molecular weight (M n ), the molecular weights (M w ) and the polydispersity index (PDI) were characterized by gel permeation chromatography (GPC, TSK gel G4000HHR column (Tosoh Bioscience, Tokyo, Japan). Mn and chemical structures were determined by proton nuclear magnetic resonance spectroscopy ( 1 H-NMR, arian Mercury plus 400, Crawley, UK).
Chemical synthesis scheme of PLGA-PEG-COOH Chemical Synthesis Scheme of PLGA-PEG-NH 2 D-mannosamine was covalently conjugated to the acid terminal groups of PLGA-PEG-COOH copolymer to yield PLGA-PEG-mannosamine copolymer. Briefly, the synthesized PLGA-PEG-COOH copolymers (200 mg, 0.006 mmol) were dissolved in 2.5 ml D-mannosamine solution in Dimethyl formamide (DMF) at a concentration of 0.025M. Then, 4dimethylaminopyridine (DMAP, 7.3 mg, 0.06 mmol) and DCC (123.8 mg, 0.6 mmol) were added stepwise. The reaction mixture was stirred at room temperature overnight under nitrogen atmosphere. PLGA-PEG-mannosamine was precipitated in a cold mixture of diethyl ether and methanol, dried by desiccator under vacuum. Mn, Mw, and PDI were characterized by GPC, and chemical structures were determined by 1 H-NMR spectroscopy.

Chemical Synthesis Scheme of PLGA-PEG-Mannosamine
Fluorescein isothiocyanate (FITC) was conjugated to PLGA-PEG-NH 2 to yield fluorescently labeled PLGA-PEG nanoparticles. FITC (4.21 mg, 0.011 mmol) and PLGA-PEG-NH 2 (100 mg, 0.0027 mmol) were dissolved in 2.5 ml anhydrous dimethylsulfoxide (DMSO) at room temperature overnight. To purify, the reaction mixture was dialyzed against DMSO and water sequentially (Mw cutoff = 12-24 kDa), then isolated by lyophilization as a yellow powder. The FITC conjugation was characterized by measurement of fluorescence absorption at an excitation wavelength of 490 nm and an emission wavelength of 530 nm using Spectramax (iD3, Molecular Devices, USA). The FITC conjugation yield was calculated according to Formula 1.

Nanoparticle Formulation
PFCE encapsulated PLGA-PEG NPs were formulated by PFCE/O/W double emulsion solvent evaporation method using PLGA-PEG, PLGA-PEG-mannosamine, or PLGA-PEG-FITC to obtain NPs with different surface ligands (7). PFCE loaded PLGA or PLGA-PEG NPs formulation was described in Figure 1B. Briefly, the first emulsion was prepared by dropwise adding PFCE (890 µl) into 3 ml DCM containing 90 mg polymer along with homogenization (Ultra-Turrax T25, IKA-WERKE, Germany) at 3,000 rpm for 20 min at RT. Subsequently, the first emulsion was dropped into 18 ml of 1% w/v PVA water solution and homogenized (Ultra-Turrax R T25 digital, IKA, Staufen, Germany) in an ice bath at a speed of 13,500 rpm for 20 min. Then the emulsion was gently stirred at RT overnight for solvent evaporation and NPs solidification. NPs were isolated by centrifugation (High speed micro-centrifuge, D3024R, Scilogex, Rocky Hill, CT, USA) at 10,000 g for 30 min at 4 • C, and washed three times by water to remove PVA. Afterwards, NPs were lyophilized (Freeze dryer, FreeZone, Labconco, Kansas City, MO, USA) by using 7% w/v sucrose as lyoprotectant and stored at −20 • C. Empty NPs were prepared by suspended the first PFCE/O emulsion step, with only the second O/W emulsion evaporation in the same conditions as described.

Size and Zeta Potential
Zeta potential, polydispersity index (PDI) and size of the nanoparticles were characterized by dynamic light scattering (DLS) at fixed at fixed 90 • scattering angle at 25 • C by Malvern Zetasizer 2000 (Malvern, UK). Suspensions were diluted in distilled water. Measurements were performed in triplicate at room temperature.

Scanning Electron Microscopy
Nanoparticle morphology was determined by a field emissionscanning electron microscope (FE -SEM Zeiss Sigma 300, Zeiss, Germany). SEM sample stage was prepared by placing a doublesided adhesive carbon tape on an aluminum stub. One small drop of 1 mg/ml nanoparticle sample suspended in ultrapure water was placed on the sample stage and then dried at 37 • C overnight. Subsequently, the dried sample was sputtered under vacuum with a chromium layer of approximately 100 Å thickness (Quorum Q150T ES, Quorum Technologies, UK) prior to analysis.

Determination of PFCE Encapsulation
Efficacy by 19    Magnetic resonance spectrometry was used to measure the 19F content per cell. The 19 F MRS spectrum was recorded using a EchoSCI sequence (TR/TE = 1,250/15 ms, NEX = 128, FOV = 6 cm, slice thickness = 2.5 cm). MRS processing was performed in SAGE 7.6.2 (GE Healthcare, Little Chalfont, UK) on the MR 901 Discovery system. For processing of the data, time domain signals were apodized with a 10 Hz line broadening function, after which the signal was zerofilled to 4,096 points. Subsequently the time domain signal was Fourier transformed and the resulting spectrum was properly phased to show an absorption mode resonance line. 19 F in the sample was quantified by reference to a standard curve, which was obtained by measuring a dilution series of PFCE with known 19 F content.

Ex vivo Determination of PFCE Encapsulation Efficacy by 19 F NMR Spectroscopy
A 400 MHz Bruker Avance II NMR spectrometer (Bruker Biospin, Rheinstetten, Germany) was used to perform 19F NMR with a 5 mm broadband probe, which can operate at 376.5 MHz for ex vivo experiments. Excised organs (liver, lungs, and spleens and 4T1 subcutaneous tumors) were harvested and flash frozen by liquid nitrogen. The prepared sample (0.4 ml, mixed with D2O) was transferred to a 5 mm NMR tube (Wilmad, Vineland, NJ, USA). As a reference compound, 5-fluorocytosine (0.1 ml, 5 mM 19F concentration) was added to determine the chemical shift and 19F concentration for each sample. The pH value of the sample was confirmed to be around 7 when preparing the sample using Bromothymol blue indicator and the temperature was maintained at 37 C during the experiment. The acquisition parameters were as follows: frequency = 376.5 MHz, spectral width = 350 ppm, relaxation delay = 5 s, data points = 64 k. After phase and baseline correction of the acquired 19 F NMR spectra using the Topspin software (Bruker Biospin, Rheinstetten, Germany), the 19 F NMR signals were quantified relative to the 5-fluorocytosine signals (reference) by peak integration. The total 19 F content of the excised organs (4T1 tumor, liver, spleen and lungs) was determined and the results were normalized to the tissue weight generating a signal expressed as a number of fluorine atoms per gram of tissue.

Statistical Analysis
The data are presented as mean ± standard deviation (SD). In vitro and in vivo tests were performed using Graphpad 7 software and One-way ANOVA and t-test analysis of variance were used to analyze the differences between the groups. Significance was attributed when P < 0.001 ( * ) for in vitro tests and P < 0.05 ( * ) for in vivo tests.

Synthesis and Characterization of Polymers
PLGA-PEG-NH-2 , PLGA-PEG-COOH, PLGA-PEG-FITC and PLGA-PEG-mannose copolymers were successfully synthesized with a yield of ∼60-85%. The synthesis of PLGA-PEG-mannose copolymer is shown in Figure 1A. Details of the polymer characteristics such as number-average molecular weight (M n ), the molecular weights (M w ) and the polydispersity index (PDI) are provided in Table 1. All the resulting copolymers presented PDI approximately of 1.4-1.7 with unimodal and symmetric peak in the GPC. This result confirmed the synthesized copolymer possessed narrow distribution of molecular weight.
The conjugation of mannosamine to PLGA-PEG copolymer was confirmed by by comparing 1 H-NMR spectra of PLGA-PEG-mannose to those of PLGA-PEG-COOH (Supplementary Figure 1A) and D-mannosamine (Supplementary Figure 1B). The peaks at 1.58 and 5.22 ppm were characteristics of methyl groups and methine groups of glycolic acid (GA) segments, while the peak at 4.82 ppm was attributed to methylene group of lactic acid (LA) segments, which were both contributed by PLGA chains (Supplementary Figure 1C). The peak observed at 3.64 ppm corresponded to methylene groups of PEG segment. The integration ratio between the characteristic peaks of PEG and PLGA chains reveals that PEG was chemically conjugated on PLGA with mole ratio around 1:1. The peaks of mannosamine overlapped with the peak of PEG (3.62 ppm), therefore only one small peak at 4.11 ppm was detected and attributed to mannosamine (21).
FT-IR analysis was carried out to further confirm the amide bond formed between PLGA and PEG segments, and also the chemical conjugation of mannosamine onto PLGA-PEG-COOH copolymer. FT-IR spectra of PLGA-PEG-COOH and PLGA-PEG-Mannosamine copolymers were displayed in Supplementary Figure 2. The absorption peaks at 1,630 and 1,510 cm −1 were assigned to the C=O and N-H bonds, respectively, which reveals the successful conjugation via amide linkage between PLGA and PEG segment of the PLGA-PEG copolymer (22). Compared with the spectrum of PLGA-PEG-COOH copolymer, the one corresponding to PLGA-PEGmannosamine presents new peaks at 3,264, 2,917, 2,851 cm −1 attributed to the stretching vibrations of the O-H and the C-H bonds of methylene and methyne groups, respectively, which were contributed by mannosamine. This result demonstrated that the mannosamine was efficiently conjugated onto the PLGA-PEG-COOH copolymer. The FITC conjugation molar ratio of PLGA-PEG-FITC copolymer was 85% measured by the fluorescence absorption according to standard curve build by pure FITC solution.

Formulation and Characterization of PLGA-PEG Nanoparticles
Formulation of PFCE loaded PLGA-PEG nanoparticles (NPs) is illustrated in Figure 1B. NPs showed a mean diameter in the range of ∼239 and ∼345 nm depending on nanoparticle formulation. PFCE encapsulated nanoparticles showed a slightly higher diameter of ∼50 nm compared with the empty ones. All the particles were also monodispersed presenting a low polydispersity index (PDI). Additionally, all the nanoparticles displayed negative zeta potentials due to the existence of terminal carboxyl groups in the PLGA polymer that is in the deprotonated form at physiological pH (23). Indeed, zeta potential values were between −31 and −17 mV ( Table 2). Mannosamine decorated PLGA-PEG NPs presented a less negative zeta potential compared with the other NPs without  [i] PLGA, poly(lactide-co-glycolide).
[iii] Polydispersity index. mannosamine due to the partial neutralization of the negative charges by mannosamine. Nanoparticles (NPs) were observed by scanning electron microscopy (SEM) where images confirmed that the PLGA-PFCE nanoparticles had a relatively uniform diameter of ∼ 200 nm with a spherical shape and smooth surface (Figure 1C).
In some cases, nanoparticles showed dented appearance supposedly due to the phase separation of PFCE from nanoparticles during the analysis (24). This finding is therefore an indirect effect of the presence of PFCE within nanoparticles.
As depicted in Figure 1D, two peaks were highlighted in the spectrum at δ = −75.65 ppm for fluorine atoms of TFA and at δ = −89.50 ppm for fluorine atoms of PFCE. The amount of PFCE encapsulated was calculated from the integration ratio between PFCE peak and TFA peak. PLGA-PEG and PLGA-PEG-mannose nanoparticles had a comparable fluorine encapsulation efficiency: 65.2% for PLGA-PEG, 67.1% for PLGA-PEG-mannose, and similar load content (6.45 µl/mg. for PLGA-PEG and 6.64 µl/mg for PLGA-PEG-mannose). Quantification and encapsulation efficiency are outlined in Table 3.

Cell Cytotoxicity and NPs Uptake Assay of Polarized Macrophages
Initial studies were performed to assess cell viability when M2-like Raw264.7 macrophages were treated with targeted or untargeted PLGA nanoparticles. Cytotoxicity assay was performed incubating macrophages with different concentration of nanoparticles ranged between 0 and 2.5 mg/ml and incubated for 24 h at 37 • C. Around 100% of cells were not affected by the treatment and all type of nanoparticles were well-tolerated (Figure 2A).
To test if fluorescent targeted or untargeted nanoparticles were preferentially up-taken by macrophages with M2-like phenotype, we first polarized Raw264.7 macrophages into M1-like macrophages, M2-like macrophages, and unpolarized macrophages (M0-like phenotype). Later, cells were incubated with targeted and untargeted PEGylation or not, FITC-PLGA nanoparticles (1 mg/ml) for 6 h. Figure 2B shows that targeted and PEGylated PLGA nanoparticles were preferentially up-taken by M2-like polarized macrophages. Particularly for the M0-like phenotype, not significant differences were highlighted by the uptake of the different nanoparticles. For the M1-like phenotype group, targeted-PEG-PLGA NPs particles had greater uptake (∼1.20-fold) than un-PEGylated-mannose NPs and also higher uptake than untargeted PEGylated particles (∼1.7-fold). For M2like phenotype group, targeted-PEG-PLGA NPs have roughly a 2.5-fold higher uptake compared to untargeted-pegylated nanoparticles and ∼eight-fold higher than un-PEGylatedmannose NPs. Statistical analysis was performed with one-way ANOVA, for triplicate samples and significance attributed when P < 0.001. Taken together, PLGA-PEG-mannose nanoparticles resulted in the highest uptake by M2-like macrophages compared to M0-like and M1-like macrophages, suggesting that both PEGylation and mannose ligand stimulate the cellular uptake.
To confirm cellular internalization of NPs, M0, M1 and M2like Raw 264.7 macrophages were treated with targeted and untargeted nanoparticles conjugated or not with PEG (1 mg/ml). Confocal images of labeled polarized macrophages are shown in Figure 2C, where PLGA-PEG nanoparticles conjugated with FITC were stained in green, cell membrane in magenta, and lysosomes in red.

In vivo Fluorine-19 Nuclear Magnetic Resonance
In order to calculate the amount of 19 F encapsulated in targeted and untargeted PLGA-PEG NPs, a curve of reference was built using different dilutions of pure 19 F ranged between 5 and 100 µl ( Figure 3A). In vitro quantification of 19 F spins encapsulated in 1 mg/ml of targeted and untargeted PLGA-PEG NPs revealed an adequate number of 19 F-atoms for further in vivo evaluations. In particular 7.01E + 19 19 F were detected for PLGA-PEG nanoparticles and 4.95E + 19 19 F were detected for PLGA-PEGmannose nanoparticles.
4T1-breast xenograft mice (n = 4 per group) received 1 mg/ml of targeted or untargeted-PLGA-PEG nanoparticles intravenously (200 µl of suspension in PBS). 19 F signals from the tumor site were quantified 48 h after NPs injection by MRS. A spectrum of 19 F signal was successfully measured from the tumor area of mice injected with targeted PLGA nanoparticle by 7T MRI as shown in Figure 3B (top). As for untargeted-PLGA-PEG nanoparticles, the signal-to-noise ratio measured from the tumor site was low and fluorine quantification was not possible. This suggests that the targeted-PLGA-PEG nanoparticles have a more efficient accumulation at the tumor site compared to the untargeted-PLGA-PEG nanoparticles. Notably, preliminary data from ex vivo 19 F NMR suggests that 19 F signal detected in the tumors after treatment with targeted-PLGA-PEG NPs was four-fold greater compared to untargeted-PLGA-PEG NPs (Supplementary Figure 3).

DISCUSSION
In this study, we have assessed PLGA-PEG NPs decorated with mannose ligand for TAMs detection by 19 F-MRI. This approach  has exploited the use of different types of PLGA nanoparticles that are not toxic, stable and by definition more resistant to mechanical stress. In addition, polymeric nanoparticles of PLGA offer the advantage to be further functionalized with target ligands. PEGylation and mannosylation show an influence in circulation and cellular uptake of nanoparticles. Actually, first of all, PEGylation of PLGA nanoparticles protects them from complement activation (i.e., opsonization) with longer circulation in the blood stream, with the consequence of an improved opportunity for the drug to be released to the target site. Secondly, mannosylation can act as cellular membranedocking ligand allowing for nanoparticle internalization in mannose-expressing macrophages especially the M2 macrophages due to overexpression of mannose receptor.
Here, we decided to encapsulate the PFCE perfluorocarbon as contrast agent for 19 F-MRI. PLGA nanoparticles used as carrier ensure PFCE stability for long storage, allowing for lyophilize, solubilize in suspensions and freeze the particles. In our study, we could produce PLGA nanoparticles of narrow size the distribution and a size approximately between 330 and 390 nm irrespective to the nanoparticle formulation ( Table 2). SEM images for PLGA-PFCE NPs confirmed rounded and smoothed surface of nanoparticles. However, PLGA-PEG NPs did not provide resolved photos due to the interference of PVA surfactant with the analysis. Thus, images of NPs obtained by TEM would provide more accurate analysis for size and shape of PLGA-PEG NPs. NPs showed minimal toxicity in vitro when incubated with macrophages also for higher concentration like 2.5 mg (Figure 1A). However, it has been shown that vitality of cells is not affected if treated with 20 mg/million cells of PLGA nanoparticles (27). The 19 F-payload of particles is similar amongst the different groups as demonstrated by 19 F-MRS and 19 F-NMR analyses. All the PFCE-nanoparticles were also able to target TAMs and be internalized by them especially if they were PEGylated and mannosylated (Figures 2B,C).
Finally, PFCE used for in vivo studies is known to be nontoxic in biological systems. Finally, PFCE used for in vivo studies is known to be non-toxic in biological systems. However, 19 F-based cell tracking suffers from detection sensitivity and in general thousands of cells per voxel are required for detection of labeled cells (28). When injection of targeted-PEG-PLGA nanoparticles, fluorine signals were detected 48 h after injections. We observed higher liver retention of nanoparticles in vivo most probably due to the continuous uptake by liver like Kupffer cells, liver sinusoidal endothelial cells (LSECs), and hepatic stellate cells (HSCs) (29). This might be overcome by treating mice with clodronate liposomes before NPs injection and blocking non-specific uptake by Kupffer macrophages in the liver and increasing 19 F-signal due to the greater retention of targeted-PLGA NPs in the tumor site (30). Recently, it has also been demonstrated that an antifouling-polymer coatings may block non-specific uptake of nanoparticles by liver (31). For untargeted PLGA-nanoparticles, the 19 F signals in the tumor did not exceed the background noise arising from the surrounding organs. For targeted PLGA-nanoparticles in the tumor site, the 19 F signals in the tumor had a weak intensity when measured by MRI in vivo, but the concentration of the PFCE could be successfully quantified by 19 F MRS. Furthermore, ex vivo 19 F NMR data confirmed higher retention of 19 F signals at the tumor site after injection of targeted-PLGA-PEG NPs compared to injected untargeted-PLGA-PEG NPs (Supplementary Figure 3). We believe that higher magnetic field strengths or more sophisticated detectors might compensate with the sensitivity of the detection allowing to measure fewer amount of fluorine in the tumor. Altogether the results presented in the study prove the efficacy of delivering PLGA-PEG-mannose nanoparticles by TAMs reaching the tumor site in vivo. Future studies will focus on accumulation of functionalized PEGylatednanoparticles delivered by TAMs as a function of tumor growth stage and as a function of the trafficking and timing of TAMs in vivo.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

ETHICS STATEMENT
The animal study was reviewed and approved by Erasmus MC Animal Experiments Committee (Animal work protocol number 17-867-53).