AuNPs Characterization and Cytotoxicity
In this study, macrophage polarization along with response to the toxic doses of AuNPs was investigated by SERS. AuNPs were used for two different purposes. First, a low concentration of AuNPs (2.5 nM) without any cytotoxicity effect was used to enhance intracellular Raman scattering to observe molecular changes during LPS-induced macrophage polarization. On the other hand, slightly increasing concentrations of AuNPs (5–8 nM) were used to observe the response of non-polarized and polarized macrophages response to the toxic doses of AuNPs. In these cases, AuNPs not only enhanced the intracellular Raman scattering but also caused cytotoxicity and changed molecular dynamics.
AuNPs were synthesized by citrate reducing method and characterized by UV/Vis and DLS, and the results are shown in Fig. 1a, 1b, and Table S2. The SPR maximum of colloidal suspension of AuNPs is observed at 530 nm and shifted to 539 nm while hydrodynamic size increased from 52 nm to 72 nm and zeta potential of -26.8 mV increased to -18.5 mV in the cell culture media due to protein corona formation.
After the characterization of synthesized AuNPs, their interaction with macrophages was investigated through cellular uptake, cell viability, ROS formation, and nitrite levels. The uptake of AuNPs by macrophages was investigated using flow cytometry in a dose-dependent manner. As shown in Fig. 1c, starting from the lowest dose, a significant increase in the side scatter shift (SSC) was observed which resulted by the uptake of AuNPs.
The cytotoxicity of AuNPs was investigated on RAW 264.7 based on cell viability, ROS formation, and nitrite secretion from macrophages, and results are shown in Fig. 1d-f. As seen, the AuNPs caused a significant reduction on cell viability of macrophages with increasing concentrations except for the 2.5 nM AuNPs which is used as a control in SERS experiments due to the non-toxic behavior. Along with cytotoxicity, AuNPs also caused ROS formation and nitrite secretion from macrophages. However, ROS formation and nitrite secretion were found to be different from the cell viability assay. Although after 2.5 nM AuNPs, all concentrations caused a decrease in the cell viability, an increase in the ROS formation was observed starting with 6 nM, and an increase in the nitrite secretion was observed only after 7 nM. Moreover, 2.5 nM to 6 nM AuNPs caused also a significant reduction of nitrite secretion. This behavior can be explained by the inhibitory activity of AuNPs with nontoxic concentrations while induction of toxicity response with high concentrations. It is known that AuNPs could reduce nitrite levels by interfering with different important steps of inflammatory response such as inhibition of IL-1β-induced activation and block of MAPKs and Akt phosphorylation [6]. On the other hand, with high concentrations, cell death can be observed as a result of ROS formation and nitrite release [19].
To understand the AuNP response of non-polarized macrophages by SERS, macrophages were treated with increasing concentrations of AuNPs starting from nontoxic doses. As seen in Fig. 1g, slight changes were observed with the increasing concentrations of AuNPs. The peak intensity changes originating from proteins (501, 618, 636, 1272–1352 cm− 1), amino acids (653, 1030, 755 cm− 1), and nucleic acids (678 cm− 1) were observed, which could be due to the death of the cells leading cellular disruption, protein degradation and denaturation, and DNA fragmentation [20]. However, dramatic changes due to cell death with high concentrations could not be observed as shown also in the canonical function plots and leave-one-out classification results in Figure S1a and Table S3. This was probably caused by the removal of dead cells during the elimination of cell culture media and washing steps before the SERS measurement.
Investigation of Macrophage Polarization in the Presence of AuNPs
For the investigation of the macrophage polarization by SERS, macrophages were polarized by LPS in the presence of AuNPs. The concentration of AuNPs was kept constant at a nontoxic concentration (2.5 nM) and the concentration of LPS increased gradually. When SSC shift after polarization of macrophages along with the AuNPs treatment was investigated (Fig. 2a), a significant SSC shift was observed in both polarized cells without any treatment and polarized cells treated with AuNPs. The shift after polarization was observed due to the higher granulation rate of the cell after LPS uptake due to the increase in vacuoles within the activated cells [21].
Polarization of macrophages was tracked by investigation of ROS formation and nitrite secretion as shown in Fig. 2b, c. As seen, with or without AuNP treatment, the presence of LPS caused ROS formation due to the macrophage polarization process. It is known that M1 macrophages produce nitrite and ROS when they are in contact with a pathogen. They use NADPH and NADPH oxidase through the pentose phosphate pathway for the production of ROS and when macrophages are stimulated by LPS, it promotes recognition by TLRs and interaction of LPS with receptors leads to the production of ROS and gene alterations [22].
When the LPS treatments are compared in the absence and presence of AuNPs, the presence of AuNPs causes a reduction in ROS production. This should be the result of the inhibitory effect of AuNPs as seen in nitrite formation results (Fig. 1f). Thus, with the inhibition of polarization by AuNPs, lower ROS is achieved when compared to the LPS treatment in the absence of AuNPs.
For the nitrite secretion, similar results with ROS formation were observed. The presence of LPS in the absence or presence of AuNPs caused significant nitrite release as expected due to being a marker of polarization into M1 type [23]. However, when LPS polarization was achieved with AuNPs treatment, a reduction in the nitrite secretion was observed. This is again due to the inhibitory effect of the AuNPs as explained above in the section of AuNPs cytotoxicity.
Polarization mechanisms of macrophages into M1 type were tracked by SERS. To observe polarization, macrophages were treated with LPS and AuNPs simultaneously and the SERS spectra are shown in Fig. 2d. As seen, when LPS and AuNPs were used together, the peak intensities attributed to from proteins (501, 618, 636, 653, 755, 838, 1030, 1180, 1218, 1272–1352 cm− 1) [24], and phospholipid (1130 cm− 1) [20] changed significantly which is caused by the M1 type polarization mechanisms induced by increasing concentrations of LPS treatment.
When macrophages were treated with LPS, LPS internalized by cells with endocytosis as well as AuNPs [25]. Thus, when cells were exposed to LPS and AuNPs simultaneously, changes in the surrounding of AuNPs will be reflected in the intracellular SERS spectra. For the uptake of LPS and AuNPs, two scenarios can happen. It is known that AuNPs can interact with LPS [26] and in this case, they can be uptaken together and starting from the first interaction of LPS with membrane components, all changes in the surrounding of AuNPs will be reflected in the SERS spectra. Furthermore, LPS can be uptaken also by endocytosis pathways such as phagocytosis, micropinocytosis, and clathrin-mediated endocytosis. Also in this case, all the surroundings of AuNPs will be the same with LPS and LPS-caused changes will be reflected in the intracellular SERS spectra.
For the LPS uptake, LPS first binds to a serum LPS binding protein (LBP), and LBP transfers the LPS monomer to the membrane-bound CD14, and then myeloid differentiation protein 2 (MD-2)/TLR4 complex initiating the LPS response [27]. After binding of LPS to MD-2, dimerization of MD-2/TLR4 complex is induced, and this induction initiates the myeloid differentiation factor (MyD88)-dependent and -independent pathways. With the activation of the MD-2/TLR4 complex, the Toll-interleukin-1 receptor (TIR) domain of TLR4 refolds and recruits 4 adaptor molecules TIRAP, MyD88, TRAM, and TRIF. With the recruitment of TIRAP and MyD88, MyD88 dependent pathway is activated through IκB kinase (IKK) which provides the activation of the nuclear factor (NF)-κB and the expression of pro-inflammatory cytokines such as TNF, IL-1β, IL-6, IL-8, IL-12, IL-23 [28]. With the recruitment of TRAM and TRIF, the MyD88-independent pathway is activated, and this activation results in the activation of interferon regulatory factor (IRF3) which is used for the expression of type I IFNs including IFN-β [29]. During all these dynamics, all the events will be happening close to the AuNPs due to the interaction with LPS, and all these changes will be reflected in the intracellular spectra as peak intensity changes in the protein peaks at 501, 618, 636, 653, 755, 838, 1030, 1180, 1218, 1272–1352 cm− 1 due to the same microenvironment of AuNPs during internalization.
As the second scenario, LPS and AuNPs can be uptaken separately, but even if uptake starts separately, the formed vesicles will be merged and continue to mature as early endosomes, late endosomes, and endolysosomes. During all these steps, many macromolecules such as proteins, cytoskeleton, and lipids play critical roles in organizing the development of the endocytosis process. After merging the vesicles that contain AuNPs and LPS, all the dynamics of the endocytosis will be reflected as well in the SERS spectra [17, 18]. Thus, the variations in the intensities of the peaks attributed to the proteins and phospholipids shown above are kept responsible from the dynamics of the endocytosis pathway.
To evaluate the variation of the observed intracellular spectra, a multivariate analysis was conducted. For this purpose, first principal component analysis was performed and then obtained principal scores were used to conduct linear discriminant analysis (LDA). Canonical function plots were extracted from LDA analysis results which provide separation of groups by showing the group. As shown in Figure S1a and S1b, better separation was achieved with SERS spectra of polarized macrophages when compared with non-polarized macrophages treated with AuNPs in both cases. This is an expected result because more significant changes in the intracellular SERS spectra were also achieved after the polarization of macrophages. This can be caused due to the inclusion of the macrophage polarization process. In the case of non-polarized cells, only the amount of AuNPs was changed to observe the cytotoxic response to the AuNPs. However, in the case of polarization, all the mechanisms required for the polarization including the signaling pathway with endocytosis of LPS and AuNPs were reflected in the SERS spectra.
Investigation of M1-Polarized Macrophages Response to the AuNPs
After polarization of the macrophages, their ability to phagocyte AuNPs was also investigated by using high concentrations of AuNPs to cause cytotoxicity. When uptake of AuNPs by polarized macrophages was investigated as shown in Fig. 3a, a significant SSC shift was observed in the presence of AuNPs after 2.5 nM when compared with polarized cells without any AuNPs treatment. This observation could be the result of an induced phagocytosis ability of activated macrophages. After polarization, macrophages can internalize more AuNPs than the non-activated macrophages as observed in the previous studies [21, 30]. Thus, when the macrophages are activated, and then treated with AuNPs, this could cause a higher uptake rate of AuNPs.
When ROS formation was evaluated after the AuNPs treatment, as shown in Fig. 3b, it was found that up to 6 nM AuNPs, the formation of ROS was higher in the absence of AuNPs. After 6 nM, ROS formation increased with AuNPs treatment. These results can be associated with the cytotoxicity results of ANPs (Fig. 1d-f). At the lower AuNPs concentrations, the ROS production of polarized macrophages was lower due to ROS formation caused by macrophage activation. On the other hand, after 6 nM concentration of AuNPs, serious cytotoxicity was observed, and this resulted in higher ROS production than the polarization process by LPS.
After the investigation of ROS formation, nitrite secretion was also observed in polarized and non-polarized macrophages treated with AuNPs as shown in Fig. 3c. When AuNPs were used after polarization, it was seen that even though there was a significant nitrite level increase with only LPS treatment, the presence of AuNPs significantly increased the nitrite secretion after polarization. These results are also correlated with the cytotoxicity of AuNPs. Without any polarization, AuNPs caused a dose-dependent nitrite secretion from macrophages (Fig. 1f). After the polarization, the same pattern was observed however without any dose dependency. An incremental nitrite secretion was observed but it was significantly higher starting from 2.5 nM. The difference between non-polarized and polarized macrophages can be explained by the uptake rates of AuNPs. After 2.5 nM, the uptake of AuNPs was significantly higher in the polarized macrophages. Thus, due to the higher uptake rate, nitrite release becomes significantly higher even at lower concentrations.
The response of the polarized macrophages against AuNPs was also examined by SERS. Figure 3d shows the intracellular SERS spectra of polarized macrophages treated with 2.5 to 8 nM AuNPs. When the SERS spectra are compared, more dramatic changes were observed compared with non-polarized macrophages and simultaneous treatment of macrophages with LPS and AuNPs. When the cells were co-treated with LPS and AuNPs, LPS and AuNPs took the same route and found in the steps of macrophage polarization. On the other hand, when LPS was given first without AuNPs, macrophages were activated by LPS but then LPS including cell culture media was removed and different concentrations of AuNPs were added in the cell culture medium for the phagocytosis. This causes a response of polarized macrophages against AuNPs. On the SERS spectra, with the AuNPs concentration increase, dramatic intensity changes are observed with the peaks originating from not only proteins (618, 636, 653, 755, 838, 882, 1002, 1272, 1352 cm− 1) but also membrane structures including cholesterol (548 cm− 1), phospholipid (1130 cm− 1), nucleic acids (678 cm− 1).
When AuNPs are added into the cell culture of polarized macrophages, they are phagocytosed. After the phagocytic receptors recognize a target particle and aggregate to initiate signaling pathways, the actin cytoskeleton is regulated for the membrane protrusions and formation of pseudopodia. As the first step of actin remodeling, the membrane-associated cortical cytoskeleton is disrupted, and then F-actin polymerization is initiated with actin filaments nucleation. Then, pseudopodia is extended, and a phagocytic cup is formed for the internalization of the target. Lastly, actin depolymerization occurs for the formation of phagosomes. During these steps, AuNPs were covered by a remodeled plasma membrane and actin is used for the remodeling of the membrane [31]. Interaction with the receptors and their mobility, membrane remodeling, and actin involvement during the phagosome formation could lead to the intensity changes of the peaks originating from lipid structures such as phospholipids (1130 cm− 1) and cholesterol (548 cm− 1).
After the formation of pseudopods, phagocytic cup, and lastly phagosome forms, it matures and fuses with lysosome to form phagolysosomes. In the maturation process, several steps are involved. In the formation of early phagosomes, small GTPase Rab5 is used for the membrane fusion and recruited EEA1 to provide phagosome-early endosome fusion. After the fusion, maturation occurs and proteins of early phagocytosis are recycled back and proteins for the maturation are recruited to the site such as Rab7, Rab-interacting lysosomal protein (RILP). Then lumen is acidified, ILVs are formed, and phagosomes fuse with lysosome [32]. As seen, the maturation process of phagocytosis requires the recruitment of various proteins to the phagocytic site and then disassembling from the site and then again recruitment of new proteins for maturation. These types of alterations in the protein profile near or on the phagosomes are observed in the SERS spectra as the intensity changes of the peaks originating from proteins such as 618, 636, 653, 755, 838, 882, 1002, 1272, 1352 cm− 1.
When obtained spectral changes were examined by means of variation (Figure S1c and Table S3), it was found that the best separation between the control group and groups of AuNPs treatments was achieved by the response of polarized macrophages to the AuNPs. This is caused by more significant changes in the case of polarized macrophage response to the AuNPs. Furthermore, similar to the polarization results, good specificity, accuracy, and sensitivity rates were achieved in between 84.3–96.5%, 84.4–92.4%, and 50-84.8%respectively. Lower sensitivity results were achieved due to the heterogenous nature of cells and their unique responses to the external stimulus such as activation by LPS and interaction with AuNPs.