Chitosan-Coated Gold Nanoparticles Induce Low Cytotoxicity and Low ROS Production in Primary Leucocytes, Independent of Their Proliferative Status

(1) Background: Chitosan-coated gold nanoparticles (CH-AuNPs) have important theranostic applications in biomedical sciences, including cancer research. However, although cell cytotoxicity has been studied in cancerous cells, little is known about their effect in proliferating primary leukocytes. Here, we assessed the effect of CH-AuNPs and the implication of ROS on non-cancerous endothelial and fibroblast cell lines and in proliferative lymphoid cells. (2) Methods: The Turkevich method was used to synthetize gold nanoparticles. We tested cell viability, cell death, ROS production, and cell cycle in primary lymphoid cells, compared with non-cancer and cancer cell lines. Concanavalin A (ConA) or lipopolysaccharide (LPS) were used to induce proliferation on lymphoid cells. (3) Results: CH-AuNPs presented high cytotoxicity and ROS production against cancer cells compared to non-cancer cells; they also induced a different pattern of ROS production in peripheral blood mononuclear cells (PBMCs). No significant cell-death difference was found in PBMCs, splenic mononuclear cells, and bone marrow cells (BMC) with or without a proliferative stimuli. (4) Conclusions: Taken together, our results highlight the selectivity of CH-AuNPs to cancer cells, discarding a consistent cytotoxicity upon proliferative cells including endothelial, fibroblast, and lymphoid cells, and suggest their application in cancer treatment without affecting immune cells.


Introduction
Currently, cancer is still the main cause of death for patients worldwide, with increasing incidence [1]. Cancer cells are characterized by uncontrolled division and proliferation, and by their ability to invade other tissues [2]. It is currently accepted that the proliferative signaling pathways in cancer cells harbor one or more driving alterations that provide them a survival edge [3,4]. Therefore, their cell-death resistance and the continuous replicative state of cancer cells limits the success of current treatments [3][4][5]. Additionally, most cancer treatments promote immunosuppression, as they are highly cytotoxic to proliferating non-cancer cells, which is the case of immune-system cells.
The application of nanotechnology in medicine seeks to innovate with new techniques and materials for diagnosis, treatment, and prevention therapies for different diseases [6,7]. Given their nanometric size, nanoparticles (NPs) are considered a possible treatment for cancer, as they can accumulate in tumor tissues (potential improvement of the therapeutic effect), show a reduced systemic toxicity [8], and their surface has the capacity to be functionalized, which can lead to a targeted therapy [9,10]. Gold nanoparticles (AuNPs) have mouse (male, 6-8 weeks). BMCs and splenic mononuclear cells were maintained at 5 × 10 6 cells/mL at 37 • C in 5% CO 2 atmosphere, using RPMI 1640 medium (Life Technologies, Grand Island, NY, USA) supplemented with 1 µg/mL amphotericin B, 1 µg/mL penicillin and 2.5 × 10 −3 µg/mL streptomycin, and 10% FBS (Life Technologies, Grand Island, NY, USA).

Nanoparticle Synthesis
The Turkevich method, previously described in [19], was used to synthetize the CH-AuNPs and SC-AuNPs. For the CH-AuNP synthesis, we prepared an acid solution of chitosan (CH, 2% w/w in acetic acid 0.4 M) by dissolving CH (medium molecular weight, 300,000 g/mol, with 75-85% of deacetylation) in 2 mM hydrochloroauric acid solution (HAuCl 4 ), then we homogenized the solution on a magnetic plate at room temperature for 15 min at 80-90 rpm until it changed to the color of red wine [18]. For the SC-AuNP synthesis, sodium citrate and HAuCl 4 were purchased from Sigma-Aldrich, sodium citrate was dissolved in distilled water to obtain a 1% solution, the 1 mM HAuCl 4 solution was mixed with sodium citrate and placed in a water bath for 15 min at 100 • C ± 2 • C, to a ratio of 1:1 (HAuCl 4 /sodium citrate) volume/volume, until it changed to the color of red wine. Finally, the synthesis was allowed to settle at room temperature, and was stored for later use. The CH-AuNPs and SC-AuNPs were diluted 1:1 in RPMI 1640 medium (GIBCO ® by Life Technologies). The concentrations were determined based on precursor salt (HACl 4 ) concentration (µM) involved in AuNPs synthesis.

Nanoparticle Characterization
Ultraviolet-visible spectroscopy was used to determine the surface plasmon resonance using a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific, Bartlesville, OK, USA). Dynamic light scattering (DLS) in a Zetasizer ZS90-Nano (Malvern Instruments, Malvern, United Kingdom) was implemented to determine zeta potential (ZP). Mean particle diameter was measured by dynamic light scattering (DLS) using a Nanosizer NS90 (Siemens, Malvern, PA, USA). For analysis, samples were diluted in distilled water (1:1000). Transmission electron microscopy (TEM) in a field-emission gun (FEI TITAN G2 80-300) operated at 300 kV was employed to confirm the size of the AuNPs.

Cell-Viability Assay
Relative cell viability was determined by MTT; this assay consisted of measuring and quantifying spectrophotometric means of yellow tetrazolium (3-(4,5-dimethylthiazilyl-1)-2,5-diphenyl tetrazolium bromide) (Milliporesigma) reduction by metabolic activity of the cells to purple formazan. In 96-well microtiter plates (Corning), 5 × 10 3 cells per well were seeded. Cells were treated with CH-AuNPs, chitosan, HACl 4 , or SC-AuNPs, at different concentrations (25,50,75, 100, and 125 µM) for 24 h. Based on the µM concentration of the precursor salt (HACl 4 ), the concentrations of CH-AuNPs and SC-AuNPs (µM) were used for the synthesis of AuNPs. After treatment, PBMCs were centrifuged at 400× g for 10 min and carefully decanted, then MTT solution (2 mg/mL in phosphate-buffered saline (PBS)) was added to each well and incubated for two hours at 37 • C. Finally, MTT solution in the medium was aspirated, and cells were dissolved with dimethyl sulfoxide (DMSO) (Milliporesigma, Eugene, OR, USA) to solubilize the formazan crystals formed in the viable cells. The optical density was measured at 570 nm using a microplate reader (Synergy2, Biotek, Winooski, VT, USA).

Reactive Oxygen Species (ROS) Analysis
ROS production levels were measured using two staining methods, dihydroethidium (DHE; Invitrogen, St Louis, MO, USA) for O 2 − quantification and dichlorodihydrofluorescein diacetate (DCFDA; Invitrogen, St Louis, MO, USA) to quantify H 2 O 2 levels by flow cytometry. In brief, 5 × 10 4 cells per well were seeded in 24-well plates (Corning Inc. Costar ® , Corning, NY, USA) and treated with CC 50 of CH-AuNPs for 24 h. After treatment, cells were washed and stained with DHE (1 µM) or DCFDA (0.25 µM) and incubated for 30 min at 37 • C. The analysis was done by flow cytometry using FlowJo Software (Tree Star Inc., Ashland, OR, USA).

Cell Cycle Analysis
Cell cycle analysis was evaluated trough intracellular DNA quantification, using propidium iodide (PI) staining by flow cytometry. In 6-well dishes, 5 × 10 5 cells were seeded and treated with CC 25 , CC 50 , and CC 75 of CH-AuNPs for 24 h. Later, we washed and fixed with 70% ethanol overnight. After fixation, cells were washed again and incubated with PI (10 µg/mL; Milliporesigma) and simultaneous RNase (Sigma-Aldrich, USA) for 30 min at 37 • C. DNA degradation and cell DNA contents for the cell cycle were measured by flow cytometry and analyzed in FlowJo Software (Tree Star Inc., Ashland, OR, USA). A SubG1 population analysis was used for DNA degradation quantification.

Statistical Analysis
The data were analyzed using GraphPad Prism (GraphPad Software, San Diego, CA, USA). The results given in this study represent the mean of at least three independent experiments done in triplicate (mean ± SD). Statistical analysis was done using a paired Student's t-test. The statistical significance was defined as p < 0.05.

Gold Nanoparticles
SC-AuNPs and CH-AuNPs showed a typical surface plasmon resonance of around 520 nm ( Figure 1A). SC-AuNPs revealed a zeta potential (ZP) of −10 mV, and CH-AuNPs exhibited a positive ZP of +36.7 mV ( Figure 1B). The average size was tested by dynamic light scattering (DLS), and SC-AuNPs revealed a size of 3-10 nm, with a mean value of 3.5 nm, while CH-AuNPs showed an average size of 3-10 nm, with a mean value of 3.75 nm ( Figure 1C); in both cases, the polydispersity was 0.3. We corroborated the size and shape of the CH-AuNPs by transmission electron microscopy (TEM), and the average size was 7.78 nm ( Figure 1D), which corresponded to the average size detected by DLS.

CH-AuNPs Induce Low Affections in Endothelial, Fibroblast, and Peripheral Blood Mononuclear Cells, and High Selective Cytotoxicity in Cancer Cell Lines
We tested cell viability after treatment with CH-AuNPs, synthesis precursors (Chitosan and HAuCl 4 ), and SC-AuNPs (used as AuNP control) by MTT assay in peripheral blood mononuclear cells (PBMCs), human umbilical vascular endothelial cells (HUVECs), murine embryonic fibroblast cells (NIH3T3s), and the non-small-cell lung cancer cell line A549. Cell viability of PBMCs ( Figure 2A) was only slightly decreased (less than 30%) at 100 µM of CH-AuNP treatment, similar to HUVECs ( Figure 2B) and NIH3T3s ( Figure 2C). Furthermore, neither chitosan, HAuCl 4 alone, nor SC-AuNPs were cytotoxic. In contrast, A549 cells showed a concentration-dependent loss of cell viability, with a mean inhibitory concentration (IC 50 ) of 75 µM and a complete inhibitory concentration (IC 100 ) of 125 µM of CH-AuNP treatment ( Figure 2D). To correlate the loss of cell viability with CH-AuNP cytotoxicity and discard a metabolic alteration, we assessed cell death by analyzing phosphatidylserine exposure and membrane permeability. Cell-death analysis of noncancer cells (PBMCs, HUVECs, and NIH3T3s) versus cancer cells (human tumor cell line A549, human leukemic cell line Jurkat, and mouse lymphoblast cell line L5178Y-R) is visualized in Figure 2E. We used the CEM cell line as a positive control, as it was previously reported to be sensitive to CH-AuNPs [18]. Human PBMCs, HUVECs, and NIH3T3s did not present more than 20% of cell death at 125 µM of CH-AuNPs. However, A549 cells presented 25% of cell death (CC 25 ) at 25 µM, increased in a dose-dependent manner, with mean cytotoxic concentration (CC 50 ) at 75 µM, and CC 75 at 125 µM, after 24 h of CH-AuNP treatment. Compared to CEM cells, Jurkat and L5178Y-R cells were more resistant to CH-AuNP treatment, and showed a CC 50 at 50 µM and CC 100 at 100 µM of CH-AuNPs. This confirmed that the CH-AuNPs were selective cell-death inductors only in cancer cells, and not their synthesis precursors, and revealed cancer cells' susceptibility. Taken together, these results indicated that CH-AuNPs did not significatively affect the integrity of non-malignant cells, regardless of the species (murine or human).

CH-AuNPs Induce Different ROS Profiles in Cancer and Non-Cancer Cells
CH-AuNPs increase ROS production in cancer cells, essential in the cell-death mechanism [17,18], and could be a substantial feature in selective cytotoxicity to cancer cells. To reveal the CH-AuNPs' effect on non-cancer cells, we tested intracellular ROS levels using two different dyes, DCFDA (which has affinity principally to H 2 O 2 ) [21] and DHE (which has affinity to O 2 − ) [22] in the HUVEC cell line, PBMCs, and lymphoid cells derived from mouse bone marrow (BM), and compared them to cancer cell lines (CEM and A549). CH-AuNPs increased DCFDA fluorescence in HUVECs (23%) compared to untreated cells (5.5%), whereas in PBMCs, we observed 5.5% fluorescence in CH-AuNPs-treated cells and 3.3% in the control; in BM cells, CH-AuNP treatment increased the fluorescence from 21% in the control to 26% in the treated cells ( Figure 3A). This was in contrast to the CEM cell line, in which CH-AuNPs enhanced fluorescence from 3.5% to 35%; and in A549, from 11% to 43.3% ( Figure 3A). On the other hand, the DHE analysis showed that CH-AuNPs increased the fluorescence from 4.3% to 15% in HUVECs, and from 9% to 27% in PBMCs ( Figure 3B). In BM cells, DHE fluorescence was not significatively modified after CH-AuNP treatment (26%) when compared to the control (22%). In cancer cells, the fluorescence potentiated to 37% and 49% in CEM and A549 cells, respectively ( Figure 3B). These results revealed that CH-AuNPs increased different ROS production depending on the cell type,

CH-AuNPs Induce Different ROS Profiles in Cancer and Non-Cancer Cells
CH-AuNPs increase ROS production in cancer cells, essential in the cell-death mechanism [17,18], and could be a substantial feature in selective cytotoxicity to cancer cells. To reveal the CH-AuNPs' effect on non-cancer cells, we tested intracellular ROS levels from 11% to 43.3% ( Figure 3A). On the other hand, the DHE analysis showed that CH-AuNPs increased the fluorescence from 4.3% to 15% in HUVECs, and from 9% to 27% in PBMCs ( Figure 3B). In BM cells, DHE fluorescence was not significatively modified after CH-AuNP treatment (26%) when compared to the control (22%). In cancer cells, the fluorescence potentiated to 37% and 49% in CEM and A549 cells, respectively ( Figure 3B). These results revealed that CH-AuNPs increased different ROS production depending on the cell type, being O2 − in PBMCs and H2O2 and O2 − in endothelial cells.

ROS Production Promoted by CH-AuNPs Is Crucial to Low Cytotoxicity in Non-Cancer Cells
Intracellular ROS play an important role in numerous physiological (including cellcycle progression, proliferation, and cell death) and pathological processes (cancer progression) [23,24]. While ROS promote cell death in cancer cells, [17,18] their implication in non-cancer cells is still unknown. Thus, the role of ROS in non-cancerous cells' cytotoxicity was assessed. In Figure 4A, we show that NAC prevented DCFDA fluorescence induced by CH-AuNPs on HUVECs (2.3%) and BM cells (17.5%), but not significatively in PBMCs

CH-AuNPs Induce Low O2 − Production and Low Cytotoxicity in Proliferative PBMCs
Since most chemotherapies affect proliferating healthy cells, we assessed the effect of CH-AuNPs on proliferating PBMCs. PBMCs were stimulated with the mitogen concanavalin A (ConA) and then treated with CH-AuNPs for 24 h. ConA is a mitogenic lectin (polyclonal activator) that activates lymphocytes, including memory-type cells, irrespec-  Next, we assessed the role of ROS in cell death. In Figure 4D, we depict representative dot plots where NAC inhibited the low cell death induced by CH-AuNPs from 25% to 13% in HUVECs, from 25% to 4% in PBMCs, and from 23% to 16% in BM cells. This revealed that the low levels of ROS induced by CH-AuNPs in non-cancer cells were crucial for low cytotoxicity, and that ROS detected by DHE (O 2 − ) might be implicated in cell death.

CH-AuNPs Induce Low O 2 − Production and Low Cytotoxicity in Proliferative PBMCs
Since most chemotherapies affect proliferating healthy cells, we assessed the effect of CH-AuNPs on proliferating PBMCs. PBMCs were stimulated with the mitogen concanavalin A (ConA) and then treated with CH-AuNPs for 24 h. ConA is a mitogenic lectin (polyclonal activator) that activates lymphocytes, including memory-type cells, irrespective of their antigenic specificity [25]. First, we confirmed proliferative PBMC status. Human PBMCs were labeled with CFSE before treatment with ConA, and after 96 h of mitogenic stimulation, several peaks with lower CFSE intensity were detected in the CFSE profiles, indicating that multiple rounds of cell division occurred during this time frame ( Figure 5A). Once we confirmed the PBMCs' proliferative status, the next step evaluated cell viability. Figure 5B shows the relative cell viability analysis, revealing that CH-AuNPs did not decrease cell viability, even at 125 µM, similar to their synthesis precursors chitosan and HAuCl 4 ( Figure 5B). To discard the cytotoxicity of CH-AuNPs in proliferating PBMCs, we assessed cell death. CH-AuNPs did not increase fluorescence for annexin V and PI more than 20% at 125 µM of CH-AuNPs (CC 100 in cancer cells) ( Figure 5C). Cell-cycle alterations induced by CH-AuNPs were evaluated in PBMCs stimulated with ConA. ConA increases the percentage of cells in phase S and G2, when compared to control PBMCs without ConA ( Figure 5D). In addition, we did not observe cell-cycle modifications in PBMCs stimulated with ConA after CH-AuNP treatment when compared to untreated PBMCs stimulated with ConA ( Figure 5D). Thus, CH-AuNPs did not affect the cell integrity or cell-cycle progression of proliferative PBMCs.
The HUVEC line and PBMCs converged on O 2 − production, which was low when compared to cancer cell lines, indicating that ROS played a crucial role in cell death. Thus, we tested H 2 O 2 and O 2 − in proliferative PBMCs. We did not observe differences in fluorescence to DCFDA in untreated cells (3.5%) and treated cells (5%), indicating that treatment did not enhance H 2 O 2 production ( Figure 5E). However, in the DHE analysis, we observed that cells treated with CH-AuNPs had enhanced fluorescence in comparison to the control, from 13% to 27% ( Figure 5F), confirming O 2 − production. Additionally, NAC inhibited O 2 − production induced by CH-AuNP treatment. Finally, to determine the role of O 2 − in cell death, we assessed phosphatidyl serine exposure with annexin V by flow cytometry in the presence of NAC. In Figure 5G, we show the detection of low fluorescence induced by CH-AuNPs (22%), and this fluorescence diminished in presence of NAC (6.5%). This indicated that low O 2 − produced by CH-AuNPs in PBMCs stimulated with ConA were involved in the low cytotoxicity, and suggested that ROS are implicated in other metabolic processes.

CH-AuNPs Do Not Modify Cell Viability in Primary Lymphoid Cells during Proliferative Stimulus
Proliferative cells are the principal target of chemotherapy, including cancer cells and non-cancer cells derived from the mouth, digestive system, hair follicles, and immune system. This is why one of the principal adverse effects of chemotherapy is the high cytotoxicity in immune-system cells. To determine the cytotoxicity of CH-AuNPs in immune system-derived cells, we tested cell death on splenic mononuclear cells, BMCs, and PBMCs, with or without the presence of two proliferative stimuli, lipopolysaccharide (LPS) and concanavalin A (ConA). We used etoposide, a widely used chemotherapeutic drug, and SC-AuNPs as controls. In Figure 6A, we can observe that CH-AuNPs and SC-AuNPs did not induce significant cell death in BM cells, which did not increase significantly during proliferative stimuli with LPS or ConA. In contrast, the cell death induced by etoposide in BM cells significantly increased under both proliferative stimuli ( Figure 6A). In the splenic mononuclear cells analyses ( Figure 6B), we can observe a similar pattern, in which CH-AuNPs and SC-AuNPs did not significatively decrease cell viability even in presence of proliferative stimuli. On the other hand, the cell death induced by etoposide significantly increased under LPS treatment. Finally, in the PBMC analysis ( Figure 6C), the results showed that proliferative stimuli did not increase the cell death induced by CH-AuNPs or SC-AuNPs, contrary to etoposide, which was highly cytotoxic to PBMCs with or without the proliferative stimuli. Additionally, in the presence of NAC, the low cell death induced by CH-AuNPs decreased in BM cells ( Figure 6A), splenic mononuclear cells ( Figure 6B), and PBMCs ( Figure 6C), indicating that ROS played a crucial role in cell death, even under proliferative stimuli of lymphoid cells.

Discussion
We synthetized CH-AuNPs by a chemical method and obtained NPs with a surface plasmon resonance of 520 nm, a diameter of 3-10 nm, and a zeta potential (ZP) of +36.7 mV. These characteristics were similar to the ones previously reported for CH-AuNPs with cytotoxic properties in tumoral and leukemic cell lines [17,18]. CH-AuNPs did not decrease the cell viability of HUVECs, NIH3T3s and PBMCs more than 30% at the concentration at which 100% of cell-viability loss was observed in A549. SC-AuNPs and the synthesis precursors, chitosan and HAuCl 4 , did not exhibit cytotoxicity. The cell-death analysis confirmed that CH-AuNPs possessed potential cytotoxic activity against A549, Jurkat, L5178Y-R, and CEM cancer cell lines ( Figure 7A), and lower toxicity to non-cancer (HUVEC and NIH3T3) cell lines ( Figure 7B) and PBMCs ( Figure 7C). This was similar to our previous reports on tumoral (HeLa and MCF-7) [17] and leukemic (K562 and CEM) cell lines [18].
AuNPs did not increase their cytotoxicity in immune-system cells, even in presence of a proliferative stimulus (Figure 7), in contrast to the conventional chemotherapy etoposide. Our data showed similar results to other evidence, in which two chemotherapies, etoposide and campotothecin, demonstrated the ability to induce apoptosis in proliferative-peripheral lymphocytes [58]. In addition, another study showed that cisplatin and gemcitabine inhibited PBMC proliferation induced by PHA [59].  [17,18], and enhanced intracellular ROS production (O2 − , •OH and H2O2). ROS were inhibited using NAC, which avoided cell death. (B) In non-cancer cell lines, CH-AuNPs enhanced •OH and H2O2 production (inhibited by NAC), leading to cell death. (C) In lymphoid cells, with or without proliferative stimulus, CH-AuNPs enhanced a slight O2 − production, which was inhibited by NAC.

Conclusions
Taken together, our results highlighted the selectivity of CH-AuNPs to cancer cells in a ROS-dependent manner (Figure 7), discarding a consistent cytotoxicity upon proliferative cells, including endothelial, fibroblast, and lymphoid cells, and suggested their application in cancer treatments without affecting immune cells. Differences were found when detecting ROS production, as we were unable to detect ROS production in PBMCs when using DCFDA, but we detected them using DHE, while in all cell lines, ROS were detected irrespective of the detection method. In addition, we did not observe significant cell death in lymphoid cells using proliferative stimuli that mimicked infection. This work opens the door to further research to determine the specific mechanisms for ROS production induced by CH-AuNPs in PBMCs, as well as in vivo experiments exploiting their selectivity to cancer cells irrespective of the proliferative status of lymphoid cells.  The dispersity of NPs is involved in cytotoxicity, and may be related to the increase of cellular endocytosis and ROS [26]. Additionally, the interaction between cationic AuNPs and negatively charged plasma membrane were shown to be determinant for the cytotoxicity [27,28], and this positive charge of the CH-AuNPs could be also determine the selectivity to cancer cells. CH-AuNPs and SC-AuNPs showed similar polydispersity (0.3 for both); however, CH-AuNPs had a positive charge (+36.7 mV) compared to a negative charge for SC-AuNPs (−10 mV). Other cationic AuNPs showed similar cytotoxicity in a cervical cancer cell line (HeLa) and in a normal human dermal fibroblast cell line (NHDF) [27]. Physicochemical properties such as surface, size, and dispersity of NPs also determine their biological impact. Several shapes of AuNPs, such as rods, stars, and spheres, showed unselective cytotoxicity in osteosarcoma (143B, MG63) cell lines and in human fetal osteoblast (hFOB 1.19) [29]. Flower-shaped and spherical AuNPs synthesized with different precursors decreased cell viability in human endothelial cells [30,31] (HUVECs), which showed an intracellular accumulation of AuNPs [32]. Other authors revealed the attenuation of cell growth in different mammalian cell lines treated with AuNPs, including the NIH3T3 cell line [32,33]. AuNPs-calreticulin did not importantly affect the cell viability of HaCaT, HUVECs, and NIH3T3 cells [34]. AuNPs-PMAM showed effects in PBMCs [35], in contrast to green AuNPs obtained from C. guianensis, which showed antitumor activity without affecting PBMCs [36]. Other reports showed that antigen-presenting cells (APCs) effectively internalized chitosan-coated FAPLGA and SC-FA-PLGA nanoparticles, causing low cytotoxic effects [37]. AuNPs obtained from Marsdenia tenacissima by green synthesis [38], sodium citrate AuNPs [39], Justicia adhatoda-AuNPs [40], AuNPs synthetized using marine bacteria Enterococcus sp [40], and AuNPs in combination with irradiation [27] inhibited cell proliferation in a concentration-dependent manner and decreased cell viability in the A549 cell line. Other AuNPs (green synthesis) using Illicium verum showed cytotoxicity in the A549 cell line [41], and 4 nm AuNP induced cytotoxicity in vitro in the L5178Y cell line [42]. These data highlight the biological effects of AuNPs depending on the shape, size, and synthesis used, and remarkably, cell lineage.
CH-AuNPs induced H 2 O 2 production in cancer cells and HUVECs, which was not observed in PBMCs or BM cells. In contrast, increased O 2 − production induced by CH-AuNPs was observed in cancer and HUVEC cell lines, and in PBMCs. DCFDA is a fluorescein-based nonspecific and indirect probe that measures H 2 O 2 and non-specifically detects other ROS, such as hydroxyl radicals (·OH), peroxynitrite (ONOO − ), and a heme protein [21]; however, none of these were detected in PBMCs. On the other hand, DHE is an intracellular ROS probe that is most commonly used for the detection of superoxide (O· 2 − ), although it also reacts with hydrogen peroxide (H 2 O 2 ) in the presence of peroxidases, and with oxidases and cytochromes [22,43,44]. The O 2 − produced by NAD(P)H oxidases, present in all cell types, participates in inflammation and may lead to toxic effects, and when produced at high levels, it may also modulate inflammation [45]. Here, we did not evaluate cytokine release, but it is possible that CH-AuNPs could induce a pro-inflammatory profile in PBMCs. Thus, further studies must be done to determine the effect of CH-AuNPs on the induction of other types of ROS, the mechanism leading to the specific redox modification in PBMCs, and their role in inflammation. Additionally, it would be important to determine if the differences in mitochondrial respiration and increased glucose consumption in cancer cells could lead to higher mitochondrial damage by CH-AuNPs, and thus ROS production in cancerous cells, rather than PBMCs, explaining these differences between normal and cancer cells.
The effect of anti-cancer agents on cell-cycle progression is important. Most, if not all, human cancer types show a deregulated control of G1 progression, a period in which cells decide whether to begin proliferation or stay quiescent [46]. In the cell-cycle analysis, we observed that CH-AuNPs did not induce cell-cycle alterations in HUVECs, similar to our observations in tumor (HeLa and MCF-7) [17] and leukemic (K562 and CEM) [18] cell lines. In PBMCs and BM cells, we did not observe a significant percentage of cells in the S and G2 phases, as these were primary cell cultures, in contrast to immortalized cell lines. We did not observe differences in either cell line during NAC treatment. Interestingly, NAC inhibited cell death induced by CH-AuNPs, which was lower in non-cancer cell lines. NAC is a precursor of L-cysteine and is a source of sulfhydryl groups in cells; it also interacts with ROS, making it a scavenger of free radicals such as •OH and H 2 O 2 [47]. Because CH-AuNPs directly enhanced ROS production, pretreatment with NAC inhibited interaction with free radicals. Autophagy is a key protective mechanism against mitochondrial damage and the consequent ROS-induced cellular accumulation [48]. Previously, we observed prosurvival autophagy in leukemic cells [18]. It is probable that alterations in ROS production could increase autophagy on lymphoid cells to avoid cell death.
However, we previously tested the effect of CH-AuNPs on PBMCs, and proliferation was not previously induced in the cell model. Thus, we assessed the effect of CH-AuNPs on PBMCs in the presence of the mitogen ConA to induce proliferation in PBMCs. We observed that ConA induced DNA synthesis and cell division in PBMCs, as previously reported [25,49]. We did not observe alterations in cell viability or integrity in PBMCs during the proliferative state. In addition, we observed that CH-AuNPs enhanced O· 2 − production, similar to PBMCs alone and cancer cell lines, even if cytotoxicity was selective only to cancer cells. SC-AuNPs and AuNPs-PMAM increased intracellular ROS in the HepG2 cell line and in PBMCs, which mediated cytotoxicity [35]. Some studies revealed that SiO 2 NPs induced oxidative stress and triggered a cytokine inflammatory response [50][51][52][53]. In addition, AuNPs capped with nucleic acid augmented PBMC proliferation in response to phytohemagglutinin, and increased release of IL-10 and IFN-γ in comparison to uncapped AuNPs [54] and IL-2 [55]. This suggested that ROS induced by CH-AuNPs could induce a proinflammatory response in PBMCs; these results reinforced that CH-AuNPs' cytotoxicity is selective only to cancer cells, and is independent of the proliferative status. Previous reports observed that glyco-thiol AuNPs showed more cytotoxicity to the A549 cell line in comparison to PBMCs, because their hydrophobic nature allowed them to cross the cancerous cell membrane more easily [54]. The A549 cell line revealed selective internalization of S15-APT QDs via classical clathrin-dependent, receptor-mediated endocytosis, in comparison to normal human bronchial epithelial cells (BEAS2B) [56]. SC-PLGA NPs were internalized more efficiently than PLGA, presumably because of receptor-mediated endocytosis; among PBMCs, APCs showed higher uptake of both NP preparations than lymphocytes [37,57]. This highlighted the different effects of CH-AuNPs on healthy and cancerous cells, which could also be due to molecular differences and the different receptors panel in cancer cells.
Finally, to test the cytotoxicity of CH-AuNPs in other immune-system cells in proliferation, we tested cell death in BMCs, splenic mononuclear cells, and PBMCs with or without the presence of two different proliferative stimuli that mimic infectious diseases (LPS and ConA) and treated with CH-AuNPs. Our results showed that CH-AuNPs and SC-AuNPs did not increase their cytotoxicity in immune-system cells, even in presence of a proliferative stimulus (Figure 7), in contrast to the conventional chemotherapy etoposide. Our data showed similar results to other evidence, in which two chemotherapies, etoposide and campotothecin, demonstrated the ability to induce apoptosis in proliferative-peripheral lymphocytes [58]. In addition, another study showed that cisplatin and gemcitabine inhibited PBMC proliferation induced by PHA [59].

Conclusions
Taken together, our results highlighted the selectivity of CH-AuNPs to cancer cells in a ROS-dependent manner (Figure 7), discarding a consistent cytotoxicity upon proliferative cells, including endothelial, fibroblast, and lymphoid cells, and suggested their application in cancer treatments without affecting immune cells. Differences were found when detecting ROS production, as we were unable to detect ROS production in PBMCs when using DCFDA, but we detected them using DHE, while in all cell lines, ROS were detected irrespective of the detection method. In addition, we did not observe significant cell death in lymphoid cells using proliferative stimuli that mimicked infection. This work opens the door to further research to determine the specific mechanisms for ROS production induced by CH-AuNPs in PBMCs, as well as in vivo experiments exploiting their selectivity to cancer cells irrespective of the proliferative status of lymphoid cells.