Cationic Polyethyleneimine (PEI)–Gold Nanocomposites Modulate Macrophage Activation and Reprogram Mouse Breast Triple-Negative MET-1 Tumor Immunological Microenvironment

Nanomedicines based on inorganic nanoparticles have grown in the last decades due to the nanosystems’ versatility in the coating, tuneability, and physical and chemical properties. Nonetheless, concerns have been raised regarding the immunotropic profile of nanoparticles and how metallic nanoparticles affect the immune system. Cationic polymer nanoparticles are widely used for cell transfection and proved to exert an adjuvant immunomodulatory effect that improves the efficiency of conventional vaccines against infection or cancer. Likewise, gold nanoparticles (AuNPs) also exhibit diverse effects on immune response depending on size or coatings. Photothermal or photodynamic therapy, radiosensitization, and drug or gene delivery systems take advantage of the unique properties of AuNPs to deeply modify the tumoral ecosystem. However, the collective effects that AuNPs combined with cationic polymers might exert on their own in the tumor immunological microenvironment remain elusive. The purpose of this study was to analyze the triple-negative breast tumor immunological microenvironment upon intratumoral injection of polyethyleneimine (PEI)–AuNP nanocomposites (named AuPEI) and elucidate how it might affect future immunotherapeutic approaches based on this nanosystem. AuPEI nanocomposites were synthesized through a one-pot synthesis method with PEI as both a reducing and capping agent, resulting in fractal assemblies of about 10 nm AuNPs. AuPEI induced an inflammatory profile in vitro in the mouse macrophage-like cells RAW264.7 as determined by the secretion of TNF-α and CCL5 while the immunosuppressor IL-10 was not increased. However, in vivo in the mouse breast MET-1 tumor model, AuPEI nanocomposites shifted the immunological tumor microenvironment toward an M2 phenotype with an immunosuppressive profile as determined by the infiltration of PD-1-positive lymphocytes. This dichotomy in AuPEI nanocomposites in vitro and in vivo might be attributed to the highly complex tumor microenvironment and highlights the importance of testing the immunogenicity of nanomaterials in vitro and more importantly in vivo in relevant immunocompetent mouse tumor models to better elucidate any adverse or unexpected effect.


Introduction
Tumors exhibit a complex stroma where several cellular and matrix components play an important role in sustaining tumor cell growth and survival. Immune cells such as T, B, macrophages, and dendritic cells interact with each other and with tumor cells in a paracrine manner through soluble factors such as cytokines and chemokines [1], by communicating through extracellular vesicles [2], or by direct contact cell-to-cell through receptor-ligand cognates [3]. Other cells such as fibroblasts contribute to the intricate tumor microenvironment (TME) by supporting not only tumor cell growth and immune response apoptosis. Indeed, the induction of immunogenic cell death can prime the tumor site for immune response as tumor cell apoptosis can release large amounts of tumor-associated antigens accompanied by damage-associated molecular patterns (DAMP), such as heatshock proteins, that can initiate local immune response activation [38,39]. These effects have prompted innovative therapeutic approaches where PTT/PDT are combined with immunotherapeutic approaches, e.g., immune checkpoint inhibitors, to boost an efficient anti-tumor immune response otherwise inefficient without the PTT/PDT priming [40][41][42]. However, gold nanoparticles can exhibit intrinsic immunomodulatory properties independent of any other features [43], which entices the study of the immunomodulatory features of all gold nanoparticles beyond the intended therapeutic purposes.
Despite the thorough reports on how gold nanoparticles [44], PEI [45], or coating iron oxide nanoparticles [23] affect immune response at several levels, there is still a gap on how the PEI-coated AuNP (hereafter named AuPEI) nanocomposites influence the tumor immunological microenvironment. We, therefore, sought to assess the immunomodulatory ability of AuPEI nanocomposites in a murine triple-negative breast tumor model based on mouse mammary cancer cell line MET-1, known for its well-infiltrated microenvironment. Contrarily to what we expected, we found that AuPEI nanocomposites shifted the immunological tumor microenvironment toward an apparent anti-inflammatory phenotype. Altogether, this finding reveals that nanoparticles can behave differently depending on the microenvironment.

AuPEI Nanocomposite Synthesis
We followed a previously published protocol for polyethyleneimine (PEI)-AuNP nanocomposite (hereafter named AuPEI nanocomposite) synthesis [47]. Briefly, the onepot synthesis of AuPEI was carried out by the mixture of aqueous (0.5 mg/mL, Au 3+ ) HAuCl 4 ·3H 2 O (G4022, Sigma-Aldrich, St. Louis, MO, USA) solution with an aqueous solution of 25 kDa PEI (40872-7, Sigma-Aldrich) at different concentrations ([PEI]: [Au] ratio of 2.5 corresponds to AuPEI-2.5, and ratio of 5 corresponds to . Visual color inspection and absorbance profiling monitored the reaction evolution in the 350-800 nm wavelength range with a Multimode Plate Reader (EnSpire, Perkin-Elmer, Waltham, MA, USA). After 48 h, the resulted product was washed three times with distilled H 2 O (ultracentrifugation, 10,000× g) and dialyzed against distilled H 2 O using a 50 kDa cut-off membrane to remove polymer excess.
Samples were analyzed for hydrodynamic diameter distribution by dynamic light scattering (DLS). DLS measurements were carried out at 25 • C on a Zeta Sizer Nano ZS (Malvern Instruments, Malvern, UK) equipped with a 5.0 mW He-Ne laser operating at 632.8 nm and an Avalanche photodiode detector.

Cryo-TEM Imaging
AuPEI were characterized by cryogenic transmission electron microscopy (cryo-TEM) as published elsewhere [48]. A 4 µL droplet of particle aqueous solution was deposited on a Quantifoil ® (Quantifoil Micro Tools GmbH, Großlöbichau, Germany) holey carbon grid. The excess of liquid on the grid was absorbed with a filter paper, and the grid was quench-frozen quickly in liquid ethane to form a thin vitreous ice film. Once placed in a Gatan 626 cryo-holder cooled with liquid nitrogen, the samples were transferred to the microscope and observed at low temperature (−180 • C). Cryo-TEM images were recorded with a 2 k × 2 k Gatan Ultrascan 1000 CCD camera (Gatan, Pleasanton, CA, USA), using a LaB 6 JEOL JEM2100 (JEOL, Tokyo, Japan) cryo-microscope operating at 200kV. Images were taken with the JEOL low-dose system (Minimum Dose System, MDS) to protect the thin ice film from any irradiation before imaging and reduce the irradiation during the image capture (IMPMC, Sorbonne Université, 4 place Jussieu, Paris 75005, France).

Cell TEM Imaging
After incubation with AuPEI for 24 h, cells were harvested, washed 3× with PBS 1×, and resuspended in fixing buffer (0.1 M sodium cacodylate, 2.5% glutaraldehyde). Cell suspensions were mixed gently for 1 h at 4 • C. After being washed twice, cells were resuspended in 0.1 M sodium cacodylate buffer until the sample preparation. Samples were then contrasted with oolong tea extract (OTE) 0.5% in cacodylate buffer, post-fixed with 1% osmium tetroxide containing 1.5% potassium cyanoferrate, gradually dehydrated in ethanol (30% to 100%), and substituted gradually in a mix of ethanol-Epon and embedded in Epon (Delta Microscopie-Labège, France). Thin sections (70 nm) were collected onto 200 mesh copper grids and counterstained with lead citrate. Grids were examined with Hitachi HT7700 electron microscope operated at 80 kV (Elexience-Verrieres Le Buisson, France), and images were acquired with a charge-coupled device camera (AMT) at MIMA2 MET-GABI, INRAE, Agroparistech, 78352 Jouy-en-Josas, France.

In Vivo Experiment
All animal experimentations were carried out in agreement with the institute Ethical Committee from Université Paris Cité. MET-1 allograft was established by subcutaneous injection of 1.5 × 10 6 cells/50 µL PBS in the flank of C57BL/6 mice (6-8 weeks old). Tumor volume was monitored with a caliper and calculated as follows: V = (xy 2 )/2, where x is the longest and y the shortest of the perpendicular diameters.
Approximately 10 days after cell injection, MET-1 tumor-bearing mice (tumor size ≥ 5 mm, volume~250 mm 3 ) were randomized into three groups as follows: (i) control Intratumoral infiltrating lymphocyte (TIL), tumor-associated macrophage (TAM), and dendritic cell (TADC) cytometric analysis: Immune cells in tumors were stained as described previously [49]. Each tumor was individually analyzed. Briefly, tumors were mechanically dissociated and digested for 45 min at 37 • C in RPMI 1640 with 37.5 µg/mL Liberase TM (Roche, Basel, Switzerland) and 8000 U/mL DNase I from bovine pancreas (Merck Millipore, Burlington, MA, USA). After being filtered (70 µm cell strainer) and centrifuged, the red blood cells were lysed with ACK buffer on the remaining pellet and subsequently filtered through a 40 µm cell strainer. The cell suspension was then rinsed in PBS and stained in 96-well round-bottom plates with a LIVE/DEAD Fixable Blue Dead Cell Stain Kit (Invitrogen) for 20 min at 4 • C. Cells were washed and stained with antibodies (Ab) against surface proteins at a 10 µg/mL concentration for 20 min at 4 • C. After surface staining, cells were fixed with BD Fixation and Permeabilization Solution for 20 min at 4 • C. After washing in PBS, cells were resuspended in PBS 2% FBS and analyzed with a BDFortessa flow cytometer (BD Bioscience). Data were analyzed by FlowJo software.
The anti-mouse antibodies used were the following ( Table 1):

Statistical Analysis
Data were analyzed and depicted with GraphPad Prism and FlowJo software. Data are presented as mean ± SEM, and the number of samples and independent experiments are stated in each figure caption for precision. Data were analyzed with a one-tailed Mann-Whitney test under the assumption of non-Gaussian distribution (nonparametric test), with a 95% confidence, and * p < 0.05.

AuPEI Nanocomposite Synthesis and Characterization
To obtain the AuPEI nanocomposites, we followed a one-pot synthesis method [47] where 0.5 mg/mL Au 3+ solution (prepared from HAuCl 4 ·3H 2 O salt) was mixed thoroughly with polyethyleneimine (PEI) for a final [PEI]: [Au] ratio of 2.5 and 5 ( Figure 1A). The assynthesized nanocomposites were directly dispersed in milliQ water. This one-pot synthesis is based on the weak reducing potential of amine residues present in the polymer. AuPEI nanocomposites produced at [PEI]:[Au] ratios of 2.5 (AuPEI-2.5) and 5 (AuPEI-5) showed a gold nanoparticle core diameter of 11.3 and 9.7 nm, respectively, as measured by cryogenic transmission electron microscopy (cryo-TEM) ( Figure 1A), and arrangement into clusters as described in detail in [50]. One-year-old AuPEI exhibited the canonical localized surface plasmon resonance (LSPR) band around 525 nm corresponding to the transversal mode; nonetheless, AuPEI-2.5 also displayed a secondary longitudinal LSPR band, suggesting a plasmon coupling phenomenon most likely due to the closeness of nanoparticles in  Figure 1B). Importantly, we did not detect significant cell toxicity at lower AuPEI concentrations (<30 µg/mL) ( Figure 1C), both in the mouse RAW264.7 macrophage cell line and in MET-1 tumor cells.
gold nanoparticle core diameter of 11.3 and 9.7 nm, respectively, as measured by cryogenic transmission electron microscopy (cryo-TEM) ( Figure 1A), and arrangement into clusters as described in detail in [50]. One-year-old AuPEI exhibited the canonical localized surface plasmon resonance (LSPR) band around 525 nm corresponding to the transversal mode; nonetheless, AuPEI-2.5 also displayed a secondary longitudinal LSPR band, suggesting a plasmon coupling phenomenon most likely due to the closeness of nanoparticles in the fractal-like clusters ( Figure 1B). Importantly, we did not detect significant cell toxicity at lower AuPEI concentrations (<30 µg/mL) ( Figure 1C), both in the mouse RAW264.7 macrophage cell line and in MET-1 tumor cells.

AuPEI Nanocomposites Activate Murine Macrophage Cell Line, RAW264.7
Based on previously published reports on PEI immunomodulatory intrinsic ability [23,26], we sought to assess whether AuPEI nanocomposites affect macrophage activation

AuPEI Nanocomposites Activate Murine Macrophage Cell Line, RAW264.7
Based on previously published reports on PEI immunomodulatory intrinsic ability [23,26], we sought to assess whether AuPEI nanocomposites affect macrophage activation status in vitro. After 24 h of treatment, AuPEI nanocomposites (5 µg/mL) were arranged in endosome-confined clusters in RAW264.7 cells (Figure 2A). Concomitantly, murine macrophages RAW264.7 secreted a high level of pro-inflammatory cytokine TNFα ( Figure 2B) and chemotactic chemokine CCL5, known for its pivotal role in recruiting leukocytes to inflamed tissues ( Figure 2C). However, we did not detect pro-inflammatory IL-6 or immune suppressive IL-10 cytokines in the supernatant after 24 h of treatment ( Figure 2D,E). Noteworthy, murine macrophages exposed to AuPEI-5 and AuPEI-2.5 ex-hibited a significant increase in metalloproteinase inhibitor TIMP-2 ( Figure 2G), together with a slight decrease in metalloproteinase MMP2 secretion ( Figure 2F) suggesting that the macrophage migratory and invasion behavior could be affected in a relevant tissue environment [51,52]. SerpinE-1, another metalloproteinase inhibitor, either slightly decreased (AuPEI-2.5) or was unchanged (AuPEI-5) ( Figure 2G). status in vitro. After 24 h of treatment, AuPEI nanocomposites (5 µg/mL) were arranged in endosome-confined clusters in RAW264.7 cells (Figure 2A). Concomitantly, murine macrophages RAW264.7 secreted a high level of pro-inflammatory cytokine TNFα ( Figure 2B) and chemotactic chemokine CCL5, known for its pivotal role in recruiting leukocytes to inflamed tissues ( Figure 2C). However, we did not detect pro-inflammatory IL-6 or immune suppressive IL-10 cytokines in the supernatant after 24 h of treatment ( Figure 2D,E). Noteworthy, murine macrophages exposed to AuPEI-5 and AuPEI-2.5 exhibited a significant increase in metalloproteinase inhibitor TIMP-2 ( Figure 2G), together with a slight decrease in metalloproteinase MMP2 secretion ( Figure 2F) suggesting that the macrophage migratory and invasion behavior could be affected in a relevant tissue environment [51,52]. SerpinE-1, another metalloproteinase inhibitor, either slightly decreased (AuPEI-2.5) or was unchanged (AuPEI-5) ( Figure 2G). Overall, our results indicate that AuPEI exposure for 24 h has significant effects on RAW macrophage secretion with important upregulation of inflammatory cytokines such as TNFα. Metalloproteinase inhibitor TIMP-2 was also affected in a significant manner. However, the use of macrophage-like cell lines, independently of a relevant tissue-specific or disease-specific environment, is oversimplified to elucidate the multifactorial effect of nanocomposites.
We thus switched to a more specific and well-known immune-infiltrated tumoral environment in which particles were injected intratumorally. Intratumoral injection is particularly relevant for nanoparticles such as AuNPs that can be activated at distance for photothermal therapy or radiosensitization in clinical practice. They can be injected transcutaneously or by endoscopy in order to shrink the tumor before resection as an adjuvant treatment. Intratumoral injection is the best option to maximize the tumor concentration of nanoparticles, Overall, our results indicate that AuPEI exposure for 24 h has significant effects on RAW macrophage secretion with important upregulation of inflammatory cytokines such as TNFα. Metalloproteinase inhibitor TIMP-2 was also affected in a significant manner. However, the use of macrophage-like cell lines, independently of a relevant tissue-specific or disease-specific environment, is oversimplified to elucidate the multifactorial effect of nanocomposites.
We thus switched to a more specific and well-known immune-infiltrated tumoral environment in which particles were injected intratumorally. Intratumoral injection is particularly relevant for nanoparticles such as AuNPs that can be activated at distance for photothermal therapy or radiosensitization in clinical practice. They can be injected transcutaneously or by endoscopy in order to shrink the tumor before resection as an adjuvant treatment. Intratumoral injection is the best option to maximize the tumor concentration of nanoparticles, minimize their release in the circulation and translocation to healthy tissue and diminish side effects. Moreover, nanoparticles can eventually exert immunomodulatory effects on their own directly in the context of the tumoral microenvironment which can prime the TME before irradiation. For all these reasons, we aimed to characterize the effects of AuPEI directly in the TME of the MET-1 syngeneic mice model of triple-negative breast cancer.

Local Tumor Modulation of Myeloid Population Induced by AuPEI Nanocomposites in MET-1 Breast Cancer Model in Immunocompetent Mise
Myeloid cells respond to environmental cues within tissues such as damaged cells, activated lymphocytes, xenobiotics, and nanoparticles to differentiate into distinct functional phenotypes [53,54]. Patrolling tissues through vasculature for any damage or danger highly depends on monocyte subsets. Of them, inflammatory monocytes play an important role as these cells express toll-like receptors (TLRs) and scavenger receptors that mediate the recognition of pathogen-associated molecular patterns and damage-associated molecular patterns. As a consequence, inflammatory monocytes infiltrate tissues where they can differentiate into macrophages and release effector molecules such as cytokines, myeloperoxidase, and superoxide, thereby initiating inflammation [55,56].
M1 macrophage phenotype produces high levels of pro-inflammatory cytokines in order to mediate resistance to pathogens, microbicidal properties, and promotion of Th1 responses. In the context of cancer, M1 macrophages can initiate and sustain inflammation, which can activate the immune response against cancer cells. In contrast, M2 macrophages generally involved in tissue remodeling and phagocytic activity can regulate immunity and participate in tumor promotion by immunosuppression. Reprogramming of protumoral M2 macrophages into inflammatory M1 subtypes has been proposed as part of anti-tumor therapy mediated by iron oxide nanoparticles [57].
To analyze the impact that AuPEI nanocomposites might exert on the TME, we injected intratumorally AuPEI nanocomposites (20 µg/40 µL, Au) in a mouse mammary tumor MET-1 (~10 days after tumor cell injection,~250 mm 3 ), a known immunogenic tumor type. A different infiltrating immune cell subpopulation was then analyzed by cytometry 24 h after injection for each tumor/mouse individually according to a gating strategy published elsewhere [58] (Figure 3A,B). As in the in vitro study, two different AuPEI nanocomposites were assessed, i.e., AuPEI-2.5 and AuPEI-5, in comparison to the vehicle injection. In the gating strategy represented in Figure 3B, first, the singlet and live cell populations were selected, and second, we focused our analysis on the myeloid subset CD11b + . From the CD11b + population, we can distinguish four subsets: (1) inflammatory monocytes defined as CD11b + MHC (Major histocompatibility complex II)II − Ly6C hi ; (2) immature macrophages defined as CD11b + MHCII hi Ly6C hi ; (3) M1-like tumor-associated macrophages defined as CD11b + MHCII hi Ly6C − ; and (4) M2-like tumor-associated macrophages defined as CD11b + MHCII − Ly6C − . These subsets describe the evolution of infiltrating monocytes and macrophages once inside the tumor microenvironment.
Among all myeloid cells, dendritic cells (DCs) are the most potent at inducing an efficient adaptive immune response as these cells process and present foreign antigens, e.g., tumor-associated antigens, to lymphocytes (B and T cells) in the context of MHC. Therefore, the presence of DCs in a TME is imperative for a proper anti-tumor immune response [59]. Dendritic cells normally represent a small fraction (1-5%) of the myeloid infiltrate in non-lymphoid tissues [60], and, although CD11c has been recently associated with a fraction of macrophages [61], CD11c remains still a bona fide DC marker. We thus analyzed the effect of AuPEI nanocomposites on tumor-infiltrating dendritic cells (TIDCs) based on the CD11c + CD45 + phenotype, which, according to the gating strategy, merely represents 4.44% of the whole cell population in untreated tumors. In untreated tumors, the CD11c + CD45 + CD11b hi subset merely represents 0.79% of the whole cell infiltrate, indicating that the incidence of CD11b in the prior analysis of CD11b + monocyte/macrophage gating is negligible. However, among the TIDC population, the CD11b hi subset is often associated with poor tumor progression prognosis [62,63] or tolerogenic scenarios [64] and can have a great influence on the behavior of TIDCs ( Figure 4A). Noteworthy, AuPEI nanocomposites lowered the proportion of both CD11b −/low -and CD11b hi -TIDC subsets ( Figure 4B,C). Moreover, AuPEI-5 nanocomposites appeared to decrease the proportion of MHCII hiexpressing cells within the CD11b − and CD11b + TIDC subsets ( Figure 4D).  Among all myeloid cells, dendritic cells (DCs) are the most potent at inducing an efficient adaptive immune response as these cells process and present foreign antigens, e.g., tumorassociated antigens, to lymphocytes (B and T cells) in the context of MHC. Therefore, the presence of DCs in a TME is imperative for a proper anti-tumor immune response [59]. Dendritic cells normally represent a small fraction (1-5%) of the myeloid infiltrate in non-lymphoid tis- analysis of CD11b + monocyte/macrophage gating is negligible. However, among the TIDC population, the CD11b hi subset is often associated with poor tumor progression prognosis [62,63] or tolerogenic scenarios [64] and can have a great influence on the behavior of TIDCs ( Figure 4A). Noteworthy, AuPEI nanocomposites lowered the proportion of both CD11b −/lowand CD11b hi -TIDC subsets ( Figure 4B,C). Moreover, AuPEI-5 nanocomposites appeared to decrease the proportion of MHCII hi -expressing cells within the CD11b − and CD11b + TIDC subsets ( Figure 4D).

Local Tumor Modulation of Lymphoid Population Induced by AuPEI Nanocomposites
Since we observed a profound impact on MET-1 tumor-infiltrating myeloid cell populations upon AuPEI nanocomposite treatment in vivo, we sought to determine whether this phenomenon concurs with a change in tumor-infiltrating lymphocyte (TIL) subsets ( Figure 5A). T cell subsets are perfectly distinguished by the co-expression of the receptortype protein tyrosine phosphatase, CD45 [65], and the β chain of the T-cell receptor, TCRβ [66]. Both markers are among the most abundant proteins within the T-cell plasma membrane and are required for T-cell receptor signaling. Gating on live cells, we indeed noticed a trend in TIL (TCRβ + CD45 + ) percentage increase in all cases; TIL percentages varied from 6.17 ± 7.8% up to 11.33 ± 0.20% (AuPEI-2.5) and 13.11 ± 4.06% (AuPEI-5) ( Figure 5B). Nonetheless, this increase in TIL was not homogeneous as the ratio of CD4/CD8 also increased from~1 (untreated) up to 1.26 (AuPEI-2.5) and 1.34 (AuPEI-5).
Unlike the T-cell compartment, PD-1 has a diverse behavior between NK and NKT cells. That is, the PD-1 + NK cell subset decreased in proportion to PD-1 − NK cells ( Figure 6E upper and Figure 6F) while in the NKT subpopulation occurred the opposite balance shift ( Figure 6E lower and Figure 6G). We can then summarize the collective effect of AuPEI nanocomposites on reprogramming murine mammary tumor MET-1 microenvironment for infiltrating immune cells as shown in Scheme 1. As such, AuPEI seem to induce a heavy infiltration of T cells that exhibit a rather exhausted phenotype (PD-1 + ) and increment in the proportion of the anti-inflammatory M2-like macrophages.

Discussion
The use of inorganic nanoparticles for cancer treatment has grown enormously in the last decades due to the versatility of these nanosystems. Nonetheless, critical concern is raised as their utilization extended into in vivo models, i.e., whether the nanoparticles interact with the immune system and the nature of this interaction. In the context of cancer, this putative immunotropic behavior of nanoparticles is pivotal to comprehend better whether nanoparticle application changes tumor immunological microenvironment in favor of or to the detriment of an efficient anti-tumor immune response. It is indeed such interaction and its nature that could potentially drive the decision making on the best combination with current and future immunotherapeutics in order to improve anticancer response. We found that PEI-coated gold nanoparticles (AuPEI) triggered TNFα and CCL5 release by

Discussion
The use of inorganic nanoparticles for cancer treatment has grown enormously in the last decades due to the versatility of these nanosystems. Nonetheless, critical concern is raised as their utilization extended into in vivo models, i.e., whether the nanoparticles interact with the immune system and the nature of this interaction. In the context of cancer, this putative immunotropic behavior of nanoparticles is pivotal to comprehend better whether nanoparticle application changes tumor immunological microenvironment in favor of or to the detriment of an efficient anti-tumor immune response. It is indeed such interaction and its nature that could potentially drive the decision making on the best combination with current and future immunotherapeutics in order to improve anticancer response. We found that PEI-coated gold nanoparticles (AuPEI) triggered TNFα and CCL5 release by murine macrophages RAW264.7 in vitro. TNFα is an intriguing cytokine formerly thought to be pro-inflammatory, hence anti-tumoral cytokine. Nonetheless, TNFα has also been associated with tumor progression lately [74], and this dual activity seems to depend on its spatiotemporal availability [75] and pleiotropic effects on the tumor microenvironment, leading even to impairment in cytotoxic CD8 + T cells within the tumor niche [76]. Likewise, CCL5, which is mainly secreted by T cells, has been linked to a pro-tumorigenic effect as its interaction with the cognate receptor CCR5 triggers tumor cell proliferation and progression into a more aggressive phenotype [77,78]. We, therefore, interrogated a highly immunogenic tumor microenvironment, i.e., MET-1, for the immunological phenotype upon AuPEI intratumoral injection to better comprehend the global effect in vivo.
Our results support that AuPEI induces a putative pro-tumor microenvironment by changing the balance within both myeloid and lymphoid populations. Tumor development (growth, progression, and metastasis) highly depends on the phenotype of the tumorassociated macrophages (TAMs) present in the tumor environment [79]. The accepted paradigm refers to M1 as anti-tumorigenic and M2 as pro-tumorigenic and both types of the landscape tumor microenvironment as their balance varies. As such, TAMs are considered primordial for aggressive tumor behavior [80]. We found that within the TAM population in MET-1 tumors, as dissected according to procedures published elsewhere [58,81], AuPEI increased the M2-like subpopulation, more likely due to the differentiation of immature macrophages rather than the M1-like TAM reconversion, as the proportion of M1-like TAM does not change significantly. Henceforth, a precise balance favoring a pro-tumorigenic environment (M2/M1) agreed with the increase in TNFα and CCL5 in RAW264.7 cells. However, while AuPEI favors the balanced M2/M1, they appear to diminish the immunosuppressive tumor-infiltrating DCs (CD11b + TIDCs) [62,82], although the proportion of MHC-II + TIDCs also decreased. The tumor-infiltrating CD11b + CD11c + myeloid cells, however, have also been associated with a totally opposite function in a murine mammary carcinoma model. Infiltrating CD11b + CD11c + myeloid cells have proven to have great potential in mediating cell death of mammary carcinoma of HER-2/neu transgenic mice. These effects appeared to depend on inducible nitric oxide synthase [83]. Therefore, the decrease in the proportion of CD11b + CD11c + myeloid cells might as well facilitate MET-1 tumor cell survival and progression independently of the increase in the tumor immunosuppressive microenvironment. Further studies should be performed on the functions of these different myeloid cells as per their cytokine/chemokine profile and suppressor capacity in vitro.
In our MET-1-based in vivo model, it appears that there is an increment in tumorinfiltrating lymphocytes (untreated, 6.17%, vs. AuPEI-2.5-treated, 11.33%, and AuPEI-5treated, 13,11%). It is now generally accepted that a tumor contains a great population of infiltrating lymphocytes with broad specificity for self or mutated tumoral antigens but inactive against tumor cells. More important, sometimes such lymphocytes exhibit markers for clonal expansion. Therefore, a huge effort has been focused on reactivating such lymphocytic populations. Because of the relatively short time elapsed from the AuPEI intratumoral injection and the increase in TILs, it is unlikely that such lymphocytic number explosion arises from a clonal expansion. It rather indicates an influx from circulating lymphocytes. Noteworthy, the time elapsed from the nanoparticle intratumoral injection and the analysis of tumor-infiltrating immune cells can affect the final results as it is rather a dynamic process. For instance, Nicolás-Boluda A et al. found that gold-coated iron oxide nanoflowers induced an increase in the CD4/CD8 ratio after 24 h, like in our system. However, this ratio significantly decreased after 12 days concomitantly with a drop in the CD4 T cell population in MET-1 tumors [84]. Thus, a long-term analysis should be performed in the future with the AuPEI.
Although CD4 + and CD8 + T-cell proportion was greatly augmented within the MET-1 tumors, most of these cells express the checkpoint molecule PD-1, likely attenuating Tcell activation [67]. From our experiments, it is not clear whether the T cells infiltrating the tumor already exhibited an exhausted phenotype (PD-1 expression) or gained such phenotype once within the tumor. Nonetheless, the high expression of PD-1 might be therapeutically useful in an anti-PD-1 therapeutic approach. That is, we can treat tumors with AuPEI, which leads to an increment in PD-1 + T cells, combined with an anti-PD-1 or -PD-1L antibody which, in turn, can reactivate these T cells to react against tumor cells. Thus, AuPEI might function as an attractor for T cells into the tumor where a checkpoint inhibitor could reactivate them.
It is well known the anti-tumoral activity of NK cells as part of the innate lymphoid cells [85,86]. Moreover, NKT cells are known to bridge innate and adaptive immune responses. In particular, NKT cells play a crucial role in tumor immunity based on the large amount of IFN-γ these cells produce upon engagement of CD1d-presented lipids [87,88]. Curiously, it is described that in anti-PD-1-resistant tumor models, the activation of NKT cells overcomes such resistance by reinvigorating exhaustive CD8 + T cells [89]. Therefore, increasing the NKT population upon i.t. injection of AuPEI, from 27.2% to 73.2% (AuPEI-2.5) and 80.7% (AuPEI-5), can potentiate the anti-tumor immunity if an external NKTactivating factor is implemented. However, a slight rise in the PD-1-expressing NKT cell subpopulation might pose another hurdle for the proper anti-tumor immune response.
Since there is no evidence in the literature for the conversion of NK cells into NKT cells, and the fact that some NKT cells can express CD4 and CD8, we believe that the increment in the NKT cell subpopulation might arise from the high infiltrate of the lymphocytic population after AuPEI injection. Thus, reducing intratumoral NK cell and PD-1-positive NKT cell proportion might hinder the natural anti-tumoral immune response in the MET-1 tumor model (Scheme 1). However, the combination with an NKT-activating ligand could potentially overcome this phenomenon. Like T-cell subsets, the combination with an anti-PD-1 antibody might counteract the signaling cascade driven by PD-1 on NKT cells, reinforcing a putative anti-tumor activity. Altogether, the lymphoid population suffers a shift toward an immunosuppressive and, henceforth, pro-tumoral phenotype at least through the rise in PD-1 + T cells and a significant decrease in NK cell proportion.
It is not the first time that gold nanoparticles have shown immunosuppressive effects. For instance, gold nanoflowers synthesized with the extract of Acanthopanacis cortex hamper LPS-induced activation of RAW264.7 macrophages by reducing the expression of enzymes iNOS and COX2 in vitro [90]. Likewise, spherical silver and hexagonal gold nanoparticles inhibit NF-κB activation in LPS-activated RAW264.7 cells, thereby exerting an anti-inflammatory effect [91]. In addition, gold-organic compound complexes have been used for decades to treat inflammation, specifically arthritis rheumatoid [92]. Even spherical citrate-coated gold nanoparticles (5 nm) exhibit a complete disruption of the IL-1β pathway in THP1 cells, and this effect disappears as the gold nanoparticle diameter increases [93]. Furthermore, 21 nm gold nanoparticles seem to trigger a systemic antiinflammatory response in mice as measured by the level of TNFα mRNA and IL-6 mRNA in the abdominal fat [43]. More recently, alkyl-terminated gold nanoparticles have proven to prevent imiquimod-induced psoriasis by inhibiting the IL-17 pathway in mice [94]. However, most of these studies have been performed in the context of inflammatory disorders, and little is known about the collective effect that gold nanoparticles have on the tumor microenvironment so far. Since gold nanoparticles are being thoroughly studied as drug delivery systems and photodynamic and photothermal therapies in cancer, it is pivotal to understand how these nanoparticles prime the TME for possible immunotherapy approaches. Here, the main implications of our findings in how AuPEI can reshape the MET-1 tumor microenvironment are that it can potentially prime tumors for a checkpoint inhibitor (anti-PD-1 or anti-PD-L1 antibody) treatment, enhancing the therapeutic efficacy of the immunotherapy.
Furthermore, the manner in which AuPEI reshape the TME might also influence the efficacy of other therapeutic interventions such as PTD/PTT where an immunogenic cell death can be induced. We indeed demonstrated that AuPEI can trigger tumor hyperthermia upon laser irradiation [47]. Hereby, we demonstrated that the intratumoral injection of AuPEI alone is capable of shifting the tumor immunological microenvironment toward an anti-inflammatory response where PD-1 signaling may prevail. In this scenario, we foresee a triple therapeutic approach: AuPEI-induced PD-1-enriched immunological tumor microenvironment; assisted immunogenic cell death induced by photothermal therapy that potentiates the presentation of tumor-associated antigens to the local and systemic immune system; and the concomitant administration of an anti-PD-1 antibody to reactivate tumor-infiltrating lymphocytes. Altogether, not only can AuPEI attract heavy infiltration of lymphocytes but also can further stimulate the immune system locally and systemically. These AuPEI potentials entice further study in multitherapeutic approaches in the future.
Moreover, the nature of nanoparticle coating can also affect the immune response as demonstrated elsewhere. As such, 20 nm spherical gold nanoparticles produced by ultrasonic spray pyrolysis but with different stabilizers (sodium citrate, poly-ethylene glycol, or polyvinyl-pyrrolidone) exhibit differential immune-modulation. While citrate-coated and PEG-coated gold nanoparticles greatly inhibit the production of pro-inflammatory cytokines by human peripheral blood mononuclear cells, e.g., IFNγ, IL-1, IL-6, IL-8, and TNFα, the PVP-coated gold nanoparticles promote an inflammatory environment by triggering the production of T H 1 cytokines [95]. AuPEI, however, are coated with polyethyleneimine (PEI), a well-known pro-inflammatory polymer [27,45] which indicated the possibility that AuPEI might promote an inflammatory environment. In contrast, AuPEI induced a rather immunosuppressed TME. Such dichotomy might arise from the presence of gold in the nanosystem which might counteract any effect from PEI. This is in disagreement with other PEI-coated nanoparticles, such as iron oxide nanoparticles, where the collective effects are conveyed in the pro-inflammatory environment in vitro and in vivo [23], indicating that not only the coating matters but also the core when it comes to the immune system modulation.

Conclusions
We collectively profiled the immunological microenvironment of the murine mammary MET-1 tumor model upon intratumor injection of PEI-coated AuNPs. By analyzing both myeloid and lymphoid populations, we observed a clear bias toward the anti-inflammatory microenvironment supported by an increase in M2/M1 balance, an increase in PD-1 + T cells, and a reduction in NK cells. This study highlights the necessity of profiling all new nanoparticles intended for cancer therapy for the global immunological effect within the tumor niche to understand their intrinsic immunomodulatory behavior better.