We firstly detected differential gene expression when cells fed with the dual PEG functional black porous silica nanoparticles (BPSi NPs) (offered by our collaborating lab [11]). Based on the Cluster heat map of differential gene expression (Figure S1), we selected more than 2 fold differential gene expressions for further investigation. Both Go (Fig. 2a) and KEGG Pathway (Figure S3) classifications were introduced to analyze the differential genes. From the Go (Fig. 2b) and KEGG Pathway (Figure S2) enrichment bubble maps, the metabolic and lysosome associated genes including phagolysosome assembly, phagocytosis and xenobiotic metabolic process, etc. were selected for further analysis. Notably, TFEB-CLEAR [12] associated genes expression were significantly activated. The RT-PCR results (Fig. 2c) also verified the gene sequencing results, the genes on TFEB-CLEAR pathway significantly increased, such as CTSD, CTSF, TFEB, MFN1, LAMP2 AND TPP1, etc.. Due to the fact that the major function of TFEB gene is to induce the bio-synthesis of lysosome and promote the occurrence of autophagy [13]. Western blot analysis was also employed to testify whether the autophagy happened when cells fed with BPSi NPs. From the Western blot results shown in Fig. 2d and 2e, TFEB, LC3B II/I and p62 proteins all significantly upregulated. As the upregulation of TFEB and LC3B II/I proteins indicating the activation of autophagy [14], we suspected that the BPSi endocytosis promote the occurrence of autophagy. However, the p62 protein is supposed to be downregulated during the autophagy process, due to the carrier protein nature which brought the endosomes to lysosomes and finally degraded. In our study, the significantly upregulation of p62 indicated the termination of degradation during the endo-lysosome fusion process [15], which probably caused by pH increase in endo-lysosome vesicles. Thus, the BPSi endocytosis may firstly induced the occurrence of autophagy, then inhibited the autophagy process by increasing the endo/lysosome pH values, owing to its amide alkalinity.
In order to testify the pH increase characteristics in endo/lysosomes by BPSi endocytosis, two commercial pH fluorescent probes were employed in our study, pHrodo™ Red Transferrin Conjugate (Thermo Fisher #P35376) and RatioWorks™ PDMPO.
pHrodo™ Red as a commercial intracellular pH indicator, usually presents weakly fluorescent at neutral pH but increasingly fluorescent as the pH drops. It was supposed to quantify cellular cytosolic pH in the range of 9 − 4 with a pKa of ~ 6.5 with excitation/emission of 560/585 nm. We could obtain a qualitative analysis conclusion from 6 cell line determinations that the endocytosis of BPSi NPs has the capability of increasing the pH values in endo/lysosomes, due to the weaken red fluorescent signals (Figure S4 and S5). However, after repeating the experiments several times following the product operation protocols, we hardly quantitative analyzed the exactly pH value decreased among different cell line before or after feeding with BPSi NPs due to no correlation between intensity and the pH values established.
PDMPO was then employed as a better solution for indicating the pH value changes after BPSi endocytosis, which introduce ratio imaging technics in pH quantitative measurement. PDMPO [2-(4-pyridyl)-5-((4-(2-dimethylaminoethy- laminocarbamoyl) methoxy) phenyl) oxazole] is characterized as acidotropic dual-excitation and dual-emission pH probe. It emits intense green fluorescence at lower pH and gives intense blue fluorescence at higher pH. This unique pH-dependent fluorescence makes PDMPO an ideal pH probe for acidic organelles with pKa = 4.47. PDMPO selectively labels acidic organelles (such as lysosomes) of live cells and the two distinct emission peaks can be used to monitor the pH fluctuations of live cells in ratio measurements. However, we still failed in measuring the pH values in 6 cell lines before and after BPSi feeding. As the results shown in Figure S6, no significant differences were observed in all 6 cell lines before and after BPSi feeding. Though a correlation has been established between the Blue/Green Ratio and the pH values (Figure S7), nonlinear correlation from pH 4–5 make the PDMPO method failed in quantitative analysis of endo/lysosomes before and after BPSi feedings.
From the data of two commercial pH indicators above, we firstly demonstrated our suspect that PEG decorated nanoparticles with amide on the PEG chain could make the endo/lysosome pH increase due to alkaline nature of amide. However, without the quantitative analysis of precise pH changes (0.1 pH range), we still cannot establish the correlations between the autophagy status and the endo/lysosome pH values, thus failed in autophagy prediction.
Based on our previous study on self-decomposable SiO2 nanoparticles, the MB dye grown in the center of the nanoparticles presented pH dependent release profiles. So, in this study, great efforts have been made to adjust the synthesize parameters, to make the MB release present linear correlation with the pH value changes. By adjusting MB and TEOS amount, we obtained 10 series self-decomposable nanoparticles with different sizes, MB loading efficiencies and drug release profiles, etc..
Based on our previous study on self-decomposable nanoparticles[16–19], we kept the same concentration of Ammonium hydroxide in 75% ethanol, but adjusted the MB and TEOS concentrations. 2 series of TEOS amount have been set as 100 µL and 80 µL, in order to obtain different shell thickness and pore size. 10 series of MB amount have been set to obtain different size of center-hollow structure and MB loading efficacies. The MB and TEOS amounts added in the protocols were as described in the Table 1 below,
Table 1
MB and TEOS amounts added in the reaction solution
Groups | MB amount (mg) | TEOS amount (µL) |
NPs 1.0/100 | 1.0 | 100 |
NPs 1.5/100 | 1.5 | 100 |
NPs 3.0/100 | 3.0 | 100 |
NPs 4.0/100 | 4.0 | 100 |
NPs 6.0/100 | 6.0 | 100 |
NPs 1.5/80 | 1.5 | 80 |
NPs 2.0/80 | 2.0 | 80 |
NPs 2.5/80 | 2.5 | 80 |
NPs 5.0/80 | 5.0 | 80 |
NPs 7.5/80 | 7.5 | 80 |
As shown in Fig. 3a and 3b, the nanoparticles size increased with the increase of the MB amount, in both TEOS concentrations (100 µL and 80 µL). At the same MB concentration, the particle size increased with the increase of TEOS amount. Moreover, with the increase of the TEOS amount, the shell thickness growth up, which has been proved by the element mapping (shown as Fig. 3c). The morphology studies predicted that with the increase of MB amount, the loading efficiency will grow up, leading to the faster release profile, while with the increase of TEOS amount, the release will be slow down. And we need to find out the appropriate MB and TEOS concentration, with which we could obtain the optimized nanoparticle systems, that we may be able to make the MB release profile linear correlated with the pH changes.
The MB loading efficiency was determined by UV-Vis spectrum. The standard curve (Figure S8) of MB was firstly drawn using series concentrations of MB solution (From 6.25 µg/mL to 46.88 µg/mL), with the equation as y = 67.63x + 0.10919, R2 = 0.9987. As calculated with the equation above, we obtain MB loading efficiency of 10 self-decomposable nanoparticles with specific parameters, detailed data shown in Figure S9.
Before study the MB release profiles in different pH solutions, the release profiles in pure water have been studied. As shown in Figure S10 and Figure S11, all the nanoparticles with TEOS amount of 80 µL presented increased MB release along with the duration increase, which was reflected by the UV-Vis absorption. Moreover, with the MB encapsulated amount increase, the growth trend of MB release become more significant. Also, the release velocity growth faster. However, as the TEOS amount increase to 100 µL, the particle surface became more dense and the release become slower when the MB amount below 3.0 mg, almost no increase trend could be observe in the MB release in water during 14 days release. As long as the MB amount increase to above 4.0 mg, obvious increase trend of MB release could be observed. One thing to be noticed is that the nanoparticles parameter of both NPs 7.5/80 and NPs 6.0/100 presented solid growth as the time prolong, almost showed linear increase trend during the first 7 days and then reached the platform.
Then we focused on the MB release behavior in different pH buffers to figure out whether self-decomposable nanoparticles with specific parameter could have the linear pH dependent MB release.
Firstly, we carried out the MB release experiments at pH 4.0 buffer solution. From Figure S12, we could easily reach the conclusion that with the same TEOS amount of 100 µL, the MB release velocity presented positive correlation with the MB encapsulated amount. Similar trend was observed in the 5 nanoparticle systems of TEOS at 80 µL (Figure S13). The center concentrated MB diffuse into the surrounding solution via diffusion due to concentration difference.
The bigger concentration gradient makes the faster MB release. Compared with the MB release in pure water, we found that the acidic environment speeded up the release of MB (Figure S12 and S13 compared with Figure S10 and S11), indicating that the MB release is not only driven by diffusion, however, in acidic solutions, electrostatic repulsion is also an important driven force due to the positive charge nature of MB. We then calculated the release percentage of each nanoparticle parameter according to the MB loading efficiency, MB standard curve and the dilution ratio at measurements. The release percentage reflected the release speed of MB in pH 4.0 acidic solution, and the results (Fig. 4) showed that only the release percentage of NPs 7.5/80 presented linear release in pH 4.0 solution. Other nanoparticle systems with specific MB and TEOS parameters though showed similar release trends, the release percentage did not linear growth. One exception is NPs 6/100, the MB release reached the platform in only 72 hrs, thus, it was hard to tell whether the MB release could growth linear before that duration at this stage.
Meanwhile, we tested the MB release profiles in near-neutral and alkali buffers (pH 6.86 and pH 9.18). The results in both Figure S14, S15 and Figure S16 and S17 demonstrated that the MB release slowed down with the solution pH increase to 6.86, moreover, with the central MB concentration increased, the MB release percentage decreased. At pH 9.18, all nanoparticles with 10 specific parameters presented very slow MB release (Figure S18 and S19),no matter in UV-Vis absorption or the release percentage, the trend was similar with the one in pH 6.86 buffer, but with even lower release percentage. So, it was clear that the self-decomposable nanoparticles only presented MB release linear growth in acidic solutions. We thought back to the endo/lysosomes pH, from 4–5, which is exactly the pH range of MB linear growth as a function of time in specific MB/TEOS parameter. Thus, we get more confident that self-decomposable nanoparticle system maybe an accurate measuring tool for quantitative determining the endo/lysosome average pH, then provide evidence on the exactly pH value of autophagy status.
The precondition of using the specific self-decomposable nanoparticles as an endo/lysosome pH indicator is that the nanoparticles stay stable in the endo/lysosome during the whole measurement process. Secondly, the MB release in endo/lysosome should occur smoothly when the measurement carried out. The colocalization of the nanoparticle in the endo/lysosomes by cell TEM study and the MB release in 6 different cell lines were studied. From the cell TEM results, all of the 10 series nanoparticles stayed in the endo/lysosomes without escaping, after 24 hrs incubation with the HepG-2 cells (Fig. 5). We also investigated the intracellular location of the nanoparticles in other 5 cell lines, 4 nanoparticles were randomly selected to demonstrate the nanoparticles were trapped in the endo/lysosomes (Figure S20). The nanoparticles with all parameters showed central hollow structure in all other 5 cell lines after 24 hrs incubation, indicating the MB release. Moreover, under more precise observation, we noticed the MB release may be different due to different hollow sizes, pointing to the fact that 1. The endo/lysosomes pH in different cells are different, 2. The MB release from the nanoparticles is very sensitive to the endo/lysosomes pH, especially for the NPs 6/100 and NPs 7.5/80. From Figure S21 we can find that nanoparticles have already realized the endocytosis and stayed in the vesicles 2 hrs after nanoparticles feeding, then nanoparticles gradually accumulated in lysosomes.
We then evaluated the correlation between the pH values and the OD values in pH 4.0–4.8. From the results in Fig. 6 and Figure S22, for NPs 6/100 and NPs 7.5/80 nanoparticle systems, the MB release presented linear decrease as a function of pH in the pH range from 4.0 to 4.8.
We then converted the OD value to MB release percentage according to the MB loading efficiency and feeding amount. As shown in Figure S23, in the first 6 hrs the MB release percentage in NPs 6/100 and NPs 7.5/80 nanoparticle systems also presented as a function of pH values. We then calculated the Residual Sum of Squares and Pearson's related coefficient at 6 hrs and 12 hrs release duration, respectively, as the Residual Sum of Squares present negative correlation with closeness of linear fitting, while the closer the absolute value of Pearson's related coefficient to 1, the more linear it is. As shown in Table S1 and S2, the highest degree of linearity is the fitting of NPs 6.0/100 nanoparticle systems, followed by the one of NPs 7.5/80 at 6 hrs release.
Till then, we were so excited by the results that the method for precisely monitoring the pH values has been established, especially with the accuracy less than or equal to 0.1 pH value interval. That means, we have great possibilities to figure out the correlation between endo/lysosome pH values and the autophagy status, which is of great significance for better studying the autophagy mechanism and predicting the autophagy process. As we can see in figure S24, the MB release in HepG-2 cells have already reach the plateau after incubation for 4 hrs. Thus, we chose the 6 hrs as the observation time point.
We then carefully investigated the MB release of NPs 6.0/100 in 6 cell lines in the nanoparticle cell interaction duration of 6 hrs, including liver cancer HepG-2 cell line, colon cancer HCT8, HCT 15 and HCT 116 cell lines, lung cancer A549 cell line and myomelanocytic cancer B16 cell line.
As shown in Table 2, we clearly differentiate the endo/lysosomes in 6 cancer cell lines, with the accuracy at 0.01 pH values, which is impossible to be done with the commercial intraocular pH indicator kits.
Table 2
Endo/lysosomes pH values calculated by equation established in 6 different cell lines
Cell lines | OD values of MB release | pH values calculated by equation y=-0.07079x + 0.40574 | Errors |
HCT8 | 0.0656 | 4.806 | 0.077 |
HCT15 | 0.111 | 4.166 | 0.083 |
HCT116 | 0.081 | 4.593 | 0.065 |
HepG-2 | 0.072 | 4.708 | 0.088 |
A549 | 0.063 | 4.847 | 0.078 |
B16 | 0.062 | 4.855 | 0.092 |
Moreover, we re-evaluated pH values in endo/lysosomes of the HepG-2 cells before and after cultured with BPSi nanoparticles. We reached the conclusion that BPSi uptake significantly increase the endo/lysosome pH values, from 4.70 ± 0.09 to 5.59 ± 0.05, perfectly illustrating the reason for BPsi uptaken induced the autophagy initially then terminated the autophagy flux. The intracellular uptake of BPSi make the quantities of the endo/lysosomes increased, which was consistent with the results of gene sequencing, that autophagy related genes (TFEB-CLEAR) was activated. Meanwhile, the autophagy termination by the increased pH values in endo/lysosomes was also coincide with the results of p62 proteins upregulation in Western Blot study.