Mild hyperthermia induced by gold nanorods acts as a dual-edge blade in the fate of SH-SY5Y cells via autophagy

Unraveling unwanted side effects of nanotechnology-based therapies like photothermal therapy (PTT) is vital in translational nanomedicine. Herein, we monitored the relationship between autophagic response at the transcriptional level by using a PCR array and tumor formation ability by colony formation assay in the human neuroblastoma cell line, SH-SY5Y, 48 h after being exposed to two different mild hyperthermia (43 and 48 °C) induced by PTT. In this regard, the promotion of apoptosis and autophagy were evaluated using immunofluorescence imaging and flow cytometry analyses. Protein levels of Ki-67, P62, and LC3 were measured using ELISA. Our results showed that of 86 genes associated with autophagy, the expression of 54 genes was changed in response to PTT. Also, we showed that chaperone-mediated autophagy (CMA) and macroautophagy are stimulated in PTT. Importantly, the results of this study also showed significant changes in genes related to the crosstalk between autophagy, dormancy, and metastatic activity of treated cells. Our findings illustrated that PTT enhances the aggressiveness of cancer cells at 43 °C, in contrast to 48 °C by the regulation of autophagy-dependent manner.

The photothermal conversion efficiency of BSA-AuNRs. To test whether BSA-AuNRs can introduce efficient photothermal effects, the temperature of BSA-AuNRs aqueous solution was recorded upon NIR irradiation using a digital thermometer while PBS-free nanoparticles were used as a control group (Fig. 3A). Upon excitation by incident radiation of appropriate wavelength, BSA-AuNRs generated sharp local heating by the photothermal conversion of the absorbed light energy, rendering these particular particles as extremely efficient "nano-heaters". Irradiating 4.5 ppm (uptaken nanoparticles) aqueous solution of BSA-AuNRs with a continuous wave laser (808 nm) for 8 min with 1.4 W/cm 2 power of laser resulted in a rise in temperature from 27 to 50 °C (ΔT = 23 °C); however, when same solutions were irradiated with power densities 2 W/cm 2 , the temperature of the solution increased from 27 to 80 °C (ΔT = 53 °C) (Fig. 3A). Monitoring temperature change (ΔT) exhibited a rapid rise (ΔT > 7 °C) within the first minute of irradiation, followed by a saturation trend (ΔT < 1 °C) after 4 min.

BSA-AuNRs plus PTT decreased viability of human SH-SY5Y cells. The safety and biocompatibil-
ity of BSA-AuNRs were investigated on SH-SY5Y cells using MTT assay and resulted compared to the CTAB-AuNRs. Figure 3 showed cell viabilities of SH-SY5Y treated with different concentrations of AuNRs (5,10,15,20,25, and 30 ppm) decorated with CTAB and BSA after 24 h. The precise concentration of AuNRs in the CTAB-AuNRs and BSA-AuNRs complexes was calculated by AAS. The incubation of cells with CTAB-AuNRs promoted significant cytotoxicity at all concentrations (5-30 ppm) (Fig. 3B). At the same time, the replacement of CTAB with BSA increased the survival rate and closed to the near-to-control levels. Of note, the SH-SY5Y cell viability increased from 8 in the CTAB-AuNR-treated group (20 ppm) to 95.33% in the BSA-AuNRs (20 ppm) group (P < 0.0001). MTT assay showed that 90% of cells were viable even at the highest concentration of BSA-AuNRs (30 ppm) after 24 h (Fig. 3B, P < 0.0001).
Next, the viability of SH-SY5Y cells was investigated after treatment with BSA-AuNRs and PTT. To this end, cells were treated with 30 ppm BSA-AuNRs and irradiated with an 808-nm NIR laser, after 4 h incubation and 3 times washing with cold PBS. We noted that the temperature of groups increased up to 43 and 48 °C eight minutes after exposure to laser intensities 0.3 and 0.9 W, respectively. Interestingly, data revealed the reduction of survival rate 85.6 to 24.6% compared to the control group as the temperature of the sample increased from Assessment of autophagic response by flow cytometry assay. The autophagic response was also monitored using flow cytometry analysis. Data showed that the percent of apoptotic cells was not statistically significant in cells treated with BSA-AuNRs compared to the non-treated group (Fig. 4B). By increasing the temperature from 43 to 48 °C (Fig. 5A), the number of apoptotic cells expressing LC3 was also increased. Despite an increase in the number of LC3-positive cells in groups exposed to the combination of BSA-AuNRs and PTT, these changes were not significant. Also, our findings show that in addition to the increase of autophagic cell percentage, the amounts of LC3 increase in response to the temperature elevation (Fig. 5B,C). Besides, we confirmed these findings qualitatively (Fig. 5D). These data showed that the treatment of human cancer cells with BSA-AuNRs plus PTT could alter the number of cells entering apoptosis.
Exposure of AuNRs to PTT changed the expression of autophagy machinery effectors in a temperature-dependent manner. To evaluate the effect of temperature on certain autophagy responses in human SH-SY5Y cells, the expressions of autophagy-related genes were monitored in two temperatures (43 °C and 48 °C) after treatment with the combination of AuNRs and PTT by RT 2 Profiler PCR Arrays-Human Autophagy 27 . Our findings showed that in a lower and higher level of mild temperature hyperthermia (43 °C The number of necrotic cells after PTT induction in the treated groups (P < 0.0001). One-Way ANOVA and Tukey post hoc analysis (n = 3). **P < 0.01; ***P < 0.001; ****P < 0.0001.   www.nature.com/scientificreports/ and 48 °C) significantly increased the expressions of several autophagy-related indicators in the SH-SY5Y cell line compared to the control group (Fig. 6D). The results of the transcriptomic analysis showed a significant change in the expression of effectors playing a critical role in different components of autophagy machinery like chaperone-mediated autophagy (CMA) and macroautophagy (Fig. 6D, Supp. Data).

Scientific Reports
Macroautophagy machinery components. The autophagy regulating machinery is composed of different genes that control three distinctive processes, including macroautophagy, microautophagy, and CMA 28 . Through the invagination of the lysosomal membrane, microautophagy controls the internalization of cytosolic recycled proteins and organelles 29,30 . Of 86 genes associated with autophagy, the expression of 54 genes was changed in response to temperature fluctuation induced by PTT protocol (Fig. 6D, Supp. Data). According to our data, the expression of most genes was up-regulated at 43 °C and 48 °C. It seems that the intensity of these changes was different between 43 and 48 °C groups (Fig. 5D). The association of these effectors with other signaling pathways seems to dictate specific cell behavior.
Vacuole formation, induction, and nucleation. Vacuole formation, vacuole targeting, ubiquitination, and autophagosome-lysosome linkage are the most important components of macroautophagy [31][32][33] . The autophagyrelated genes (ATG) have an indisputable role in controlling the autophagosome formation and macroautophagy processes 34 . As shown in Fig. 6 Ubiquitination and phagophore expansion. Two ubiquitin-like systems, including LC3 conjugation and Atg12-Atg5-Atg16L, are crucial partners for the elongation of the pre-autophagosomal structure (PAS) system 35 (Fig. 6D, Supp. Data). These two systems work in two different manners. By an enzymatic function of Atg7 and Atg10, the protein Atg12 is attached to the Atg5 factor. After that, the Atg12-Atg5 complex binds non-covalently to the Atg16L protein, and this complex serves as an e3-like ubiquitin ligase for autophagosomal membrane formation and development 34 . Secondly, Atg7 and Atg3 cooperatively connect another ubiquitin-like protein (LC3) with phosphatidylethanolamine (PE) to increase the intracellular concentration of LC3II. LC3-II is touted as a final mediator which participates in the elongation phase of the autophagosome 36 . In the SH-SY5Y cells treated with mild hyperthermia, the expression level of LC3-I and MAP1LC3B was up-regulated at 43 °C and 48 °C. However, the intensity of expression varied in both groups. This different but aligned pattern in the expression of LC3 confirmed the fact that autophagy could be stimulated at different temperatures with a certain activity. These features lead to different consequences and behavior in cancer cells 37 . Another important gene that belongs to autophagy machinery, HDAC6 (Histone deacetylase 6), is crucial for ubiquitin-based control of autophagy and facilitating the lysosome-autophagosome attachment 38 . The combination of AuNRs and PTT did not alter the expression of HDAC6 at 43 °C and 48 °C groups compared to the control (Fig. 6D, Supp. Data).
Vacuole targeting and autophagosome-lysosome linkage. The next step after autophagosome formation is tethering and merging the autophagosome with lysosomal vacuole and, subsequently, the release of autophagic vesicle contents into the lysosomal inner space 39 . In the following steps, the contents are degraded by lipases and hydrolases in the lysosomes. Finally, the products of degradation processes are released into the cytosol by efflux transporters 40 . The transcriptome profile of some associated proteins involved in the fusion process has been evaluated in this study. Lamp1, the Lysosomal-associated membrane protein 1 gene, was significantly up-regulated at 43 °C (3.64-fold) and 48 °C (3.08-fold), respectively. In contrast, Atg4b was significantly downregulated 43 °C (− 3.28-fold) and 48 °C (− 5.08-fold). Atg4c, another autophagosome-lysosome linkage-related gene, was up-regulated (3.31-fold) at 43 °C, while the treatment of cells at 48 °C did not alter transcription level compared to the control (Fig. 6D, Supp. Data). The expression of GABARAP, the Gamma-aminobutyric acid (GABA) A receptor-associated protein-like, was not changed at both temperatures. Concerning these findings, it seems that the autophagosome-lysosome linkage step of autophagy was stimulated more effectively at 43 °C rather than 48 °C.
Chaperone-mediated autophagy. CMA is a specific pathway that cells used directly for lysosomal degradation of assembled proteins without any lysosomal membrane changing following intra/extracellular stresses 35 . CMA signaling can be controlled by some of the heat shock proteins like 71 kDa protein (Hsc70), also known as HSPA8, an important member of the heat shock protein 70 family (Hsp70). HSPA8, in combination with HSP90AA1 (cytosolic class A member 1 co-chaperones) forms chaperone/substrate complex. This complex uses a pentapeptide motif (KFFRQ) targeted misfolded cytosolic proteins 41

The activation of autophagy by PTT alters the colorogenic capacity of human SH-SY5Y cells.
In this study, we used the colony formation assay to investigate the PTT's effect at two different temperatures, 43 °C, and 48 °C, on human SH-SY5Y cells. Our results showed that incubation of cells at 43 °C had protective effects on the cancer cells, while the promotion of autophagy at 48 °C led to cytotoxicity in SH-SY5Y cells (Fig. 7A). To show the protective effect of autophagy in SH-SY5Y cells after being exposed to www.nature.com/scientificreports/ PTT, we used an autophagy inhibitor, 10 µM HCQ. We noted that the incubation of cells with HCQ sensitized them to the detrimental effect of PTT, which was blunted by the activation of autophagy. Data confirmed that the number of colonies was significantly decreased in the presence of HCQ after being exposed to 43 °C. This finding suggests that autophagy stimulation by PTT at 43 °C leads to increased SH-SY5Y cell survival, indicating the protective effect of autophagy in cancer cells exposed to mild hyperthermia. The inhibition of the final phase of autophagy by HCQ contributes to the accumulation of toxic metabolites and unfolded/misfolded proteins, inducing apoptosis 42 . In contrast, the exposure of cancer cells to higher temperatures, such as 48 °C destroyed many cells quickly and inhibit colorogenic capacity (Fig. 7B). These findings are consistent with the results obtained from the study of genes involved in autophagy. The changes in the expression profile of genes in response to environmental stresses confirmed, and its possible role in the alteration of cancer cell properties have been established in the last decade 43 .
The induction of autophagy can result in three types of consequences, including dormancy, invasiveness, and death 44,45 . Along with this statement, we measured the expression level of Ki-67 protein as a marker of cell division using ELISA (Fig. 6A). Based on our data, Ki-67 was not changed in cells exposed to 43 °C compared to the control. In contrast, applying 48 °C inhibited significantly the synthesis of Ki-67 and LC3 compared to non-treated control cells and 43 °C group (Fig. 6A,B). Our results showed the up-regulation of important genes involved in dormancy, including CTSB, P53, IFN-B, IFN-Y, CDKN1B (P27), CDKN1A (P21), and PTEN at 43 °C group compared to the 48 °C group (Supp. Data, Fig. 6C,D). Considering the potency of 43 °C to induce the synthesis of nuclear factor Ki-67 and stimulate colorogenic capacity and autophagic response in the SH-SY5Y cells (Fig. 6A), we could hypothesize that the incubation of cancer cells in mild temperatures could inhibit active dynamic growth but simultaneously promote autophagic response which is a compensatory response to resist again insulting stimuli. The activation of certain genes in autophagy signaling could trigger other resistance mechanisms such as dormancy in the host cells (Supp. Data). In support of this notion, the blockade of autophagic response could decrease the cancer cells' survival rate compared to control matched groups. However, it should not be forgotten that the stimulation of autophagic responses to a certain extent can be protective. The over-stimulation of autophagy effectors could alter the expression and activity of other parallel signaling pathways, such as apoptosis, as seen in the groups exposed to 48 °C. Data from the flow-cytometry analysis further supported these features (Fig. 5).
Interestingly, the expression of autophagy machinery genes related to the invasion was increased at 48 °C. Despite the loss of a large number of cells at 48 °C, cells can form colonies in the presence of HCQ (Fig. 7B). It seems that the distinct effect of temperature on the modulation of autophagy-related genes could be responsible for certain cancer cells' behavior. Consistent with our data, it has been shown that the tumor formation process that arises from the Ras pathway activation requires high levels of autophagy for survival and cell proliferation 42 .
Concerning the induction of some effectors in autophagy signaling such as MAPK14, LAMP1, eIF4E, eIF4GI, ATG16L1, CDKN1A, CDKN1B, CTSB, and GABARAPL1 and their contribution to metastasis, we conclude that some genes have a regulatory role over other counterparts. For instance, EIF4E and eIF4G translation initiation factors are involved in tumorigenesis and patient survival (Tables 1, 3). The important role of these factors in cell migration, amplification of EMT promoters, and metastasis has been demonstrated previously (Table 2). Like ARHI, some genes play a crucial role in controlling the outcome of autophagy 13,46 . It has been shown that the re-expression of ARHI can lead to autophagic cell death even in the presence of HCQ 47 .

Discussion
According to our findings, the induction of autophagy using AuNRs seems to be completely dependent on the generation of sudden intracellular temperature. The induction of autophagy at two temperatures (43 °C and 48 °C) resulted in completely different consequences. Based on our findings, the BSA-AuNR complex with NIR to heat conversion efficiency and good biocompatibility to 30 ppm could be pretty well candidates for photothermal therapy. Also, an increase of intracellular temperature up to 43 °C caused clean cell death (apoptosis) compared to higher temperatures (48 °C) (necrotic). As we mentioned earlier, the type of cell death has a profound effect on the fate of other cells located in the tumor. Interestingly, LC3 expressing apoptotic cells enhanced by increasing the temperature from 43 to 48 °C which shows autophagic cell death (Fig. 5D, 6B). These data showed that the treatment of human cancer cells with BSA-AuNRs plus PTT could alter the number of cells entering apoptosis. Based on our best knowledge this is the first report about temperature-dependent autophagic cell death by PTT.
For the detailed address of this behavior, we showed that of 86 genes associated with autophagy, the expression of 54 genes was changed in response to temperature variation. Up to now, it has been confirmed that changes in the expression profile of autophagy-related genes contributed to the fate of cancer cells like cell death type, metastasis ability, colony formation, invasiveness, and dormancy which are summarized in (Tables 1, 2, 3). Also, Lu et al. showed that the expression intensity or level of autophagy-related genes affects the cell fate 1 . Interestingly, our results illustrated that the intensity and pattern of these changes were different between 43 and 48 °C (Tables 1 and 2, Supp. Data). In addition, the pattern of some expressed genes is in contrast to each other in two different temperatures like AMBRA1, ATG9B, LC3-I, MAP1LC3B, Atg5, Atg4c, and Atg16. As aforementioned, these findings added a notion that changes in the autophagy effectors contributing to autophagy levels which resulted in altering cell behavior and fate. Furthermore, we showed for the first time that the CMA pathway 2 , which is controlled by some of the heat shock proteins like HSPA8 is activated at 48 °C.
All together molecular, transcriptional, and colony formation experiments in this work demonstrated that the low level of mild hyperthermia (43 °C) has protective effects on the cancer cells, while the higher level of mild hyperthermia (48 °C) led to cytotoxicity in SH-SY5Y cells. Therefore, despite the induction of more apoptosis, our findings suggest that autophagy stimulation by PTT at 43 °C leads to increased autophagic survival and www.nature.com/scientificreports/ dormancy. The findings are consistent with the results obtained from the study of genes involved in autophagy in this study, which finally can threaten the patients' lives in the clinic. The most important finding in this study was the rethinking of the usage of autophagy inhibitors in the PTT therapeutic regimens. Colony formation assay findings as the outcome of PTT were shown that, despite the loss of a large number of cells at 48 °C, cells can form colonies in the presence of autophagy inhibitor (Fig. 7A-E). It means that using autophagy inhibitors in lower temperatures can be beneficial, while in higher temperatures cause to exacerbate the fate of resistant tumor cells. It seems that a higher level of autophagy (e.g., EIF4E and eIF4G) and released factors from neurotic cells mutually evoked the invasion and metastasis-related signaling pathways at 48 °C (Tables 1, 3). For instance, it has been shown that the tumor formation process that arises from the Ras pathway activation requires high levels of autophagy for survival and cell proliferation 3 .
Since the use of hyperthermia to eliminate cancers has entered into the clinical phase, the results of this study allied to the others 4 ring the alarms for all of us, which forcefully recommended using all necessary experiments to prevent unwanted complications. As shown in this study, after eradication of a large number of cancer cells at 48 °C, the remaining cells in the presence of an autophagy inhibitor will be able to form a tumor. This finding is in line with other scientists who have previously shown that even 100 cancer stem cells are enough to form a tumor 5 . So, we strongly suggest further investigations to find the relationship between autophagic gene machinery and PTT-induced autophagy. Characterization. UV-Vis spectra were recorded on a spectrophotometer (Cecil UV, UK). Transmission electron microscopy (TEM) analyses were performed by a Zeiss EM900 80 kV electron microscope. TEM samples were prepared by dropping a small quantity of dispersion onto formvar carbon 300 mesh thick grids. The size of AuNRs was determined by the dynamic light scattering (DLS) technique using Zetasizer Nano ZS90; Malvern Instruments, UK. The samples were irradiated with a laser model MDL-III-at a wavelength of 808 nm and the energy of 2.5 W (Changchun New Industries Optoelectronics Tech. Company). The dispersions' temperature was monitored using a digital thermometer with a thermocouple probe from Pyrometer Instrument Company. In the MTT assay, the absorbance (OD) of samples was measured at 570 nm using a microplate reader (BioTek, USA). Flow cytometry tests were performed on a flow cytometer (Partech space flow ® ), and data were analyzed by FlowJo software (version V.10).

Materials.
Gold nanorods (AuNRs) synthesis. The synthesis of AuNRs was done according to the previously described seedless growth method 26 . Briefly, 5 mL of 1.0 mM HAuCl 4 was added to 5 mL of 0.20 M CTAB. Then, 250 mL of 4.0 mM AgNO 3 was added, and the solution was gently shaken. By adding 8 mL of 37% HCl pH of the solution was adjusted to 1-1.15. Subsequently, 70 mL ascorbic acid (78.8 mM) was added to the solution then gently shaking until the solution was clear. After that, 15 mL of 0.01 M ice-cold NaBH 4 was injected to the unstirred growth solution and allowed to react for 6 h. The solution was centrifuged at 10,000 rpm for 15 min, and the supernatant was removed. The pellet was suspended in water and centrifuged at the same speed for an additional 15 min.
Surface modification of AuNRs with BSA (BSA-AuNRs). BSA was used to modify the surface of the AuNRs according to a method described by Tebbe et al. 48 . Briefly, the NP dispersions were diluted with deionized water and slowly added to the BSA solution under ultrasonication (BSA solution/NP dispersions, 1:1 v/v). The BSA solution contains BSA (10 mg/mL), and 0.02% citrate and pH were adjusted to 7. The NPs were sonicated for 30 min and then centrifuged at 6500 rpm for 6 min followed by replacement of supernatant with 10X-diluted BSA solution (1 mg/ml, pH = 12, 0.02% citrate). The final solution was stirred for at least 24 h. Eventually, the particles were centrifuged and washed with deionized water before use. Analyzing cellular uptake of AuNRs using atomic absorption spectroscopy (AAS). SH-SY5Y cells were seeded in 6-well plates with at an initial cell density of 2 × 10 5 cells/well. After reaching 70-80% confluence, the culture medium was replaced with fresh medium containing 30 ppm BSA-AuNRs and 2% FBS. The culture medium was discarded after a 4-h incubation period. Then, the cells collected using an enzymatic solution (Trypsin-EDTA) dissolved in Aqua Regia overnight and heated to about 140 °C to excluding hydrogen chloride and nitrogen oxides until the solution became colorless and clear 49 . After that, the digested solution incubated in an aqueous solution containing 2% nitric acid and 1% hydrogen chloride for loosing and detaching gold atoms in nanoparticle lattice; the total cellular gold content was determined by AAS.

AuNR-mediated PTT protocol in vitro.
In this study, SH-SY5Y cells were randomly allocated into three different groups as follows: Control, 30 ppm BSA-AuNRs (43 °C), and 30 ppm BSA-AuNRs (48 °C). In this regard, cells were maintained for 4 h, followed by replacement with BSA-AuNR free culture medium. Then, cells were irradiated by an 808-nm NIR laser at a power density of 0.3 and 0.9 W for 8 min (to reaching the 43 °C and 48 °C) in the absence and presence of 10 µM HCQ. During laser irradiation, the temperature of the culture medium was carefully monitored.
MTT assay. The localized cell-killing ability of PTT at the varied densities (0.3 and 0.9 W) was assessed by a typical MTT assay. After irradiation, cells were maintained for the next 24 h, and the survival rate was measured by using a standard MTT assay. To this end, the culture medium was discarded, replaced with 5 mg/ml MTT solution, and incubated at 37 °C for 3-4 h. Thereafter, the supernatants were removed, and 100 µl DMSO solution was added to each well to dissolve formazan crystals. Percent of cell viability was denoted as the relative absorbance of treated versus untreated viable cells. The following formula was used to calculate the inhibition of cell growth:

Flow cytometric analysis of apoptotic cells.
To quantify the effect of PTT at varied energy densities on SH-SY5Y cells, we performed flow cytometry analysis. To this end, cells were cultured in each well of 24-well plates and irradiated as above-mentioned. Twenty-four hours post-PTT, cells were harvested, permeabilized, and stained using the Annexin-V-BioXBio Staining kit according to the manufacturer's instructions. The percent of alive, apoptotic, and necrotic cells were analyzed by the Partech Cyflow space ® Flow cytometry system. The raw data were processed using FlowJo software (version V.10).
Monitoring autophagy status using flow cytometry and immunofluorescence staining. Flow cytometry analysis. To assess autophagy status after PTT, the SH-SY5Y cells were incubated with the FITC-Cell viability (%) = mean of the absolute value of treatment group/mean of the absolute value of control ×100%.