Designed heterogeneous palladium catalysts for reversible light-controlled bioorthogonal catalysis in living cells

As a powerful tool for chemical biology, bioorthogonal chemistry broadens the ways to explore the mystery of life. In this field, transition metal catalysts (TMCs) have received much attention because TMCs can rapidly catalyze chemical transformations that cannot be accomplished by bio-enzymes. However, fine controlling chemical reactions in living systems like bio-enzymes is still a great challenge. Herein, we construct a versatile light-controlled bioorthogonal catalyst by modifying macroporous silica-Pd0 with supramolecular complex of azobenzene (Azo) and β-cyclodextrin (CD). Its catalytic activity can be regulated by light-induced structural changes, mimicking allosteric regulation mechanism of bio-enzymes. The light-gated heterogeneous TMCs are important for in situ controlling bioorthogonal reactions and have been successfully used to synthesize a fluorescent probe for cell imaging and mitochondria-specific targeting agent by Suzuki–Miyaura cross-coupling reaction. Endowing the bioorthogonal catalyst with new functions is highly valuable for realizing more complex researches in biochemistry.

ray diffraction (XRD) analysis on a Rigaku-Dmax 2500 diffractometer by using CuKα radiation. X-ray photoelectron Spectroscopy (XPS) spectra were analyzed by Thermo Fisher Scientific ESCALAB 250Xi Spectrometer Electron Spectroscopy (America). ICP-MS measurements were performed on a TheroScientific Xseries Ⅱ inductively coupled plasma mass spectrometer. The flow cytometry data were obtained by BD LSRFortessa™ Cell Analyzer. Microwave reactions were carried out by Biotage Initiator+ Microwave System.

Synthesis of dendrimer-like macroporous silica nanoparticles (DMSN)
The DMSN were prepared according to the literature with little modification. 1 In brief, a mixture of cetyltrimethylammonium tosylate (CTATos, 1.92 g), triethanolamine (0.31 g) and water (100 mL) was stirred at 80 °C for 1 h, then 15.6 mL TEOS was added and the mixture was stirred at 80 °C for another 2 h. The synthesized DMSN were filtered, washed with water and ethanol and dried in oven at 60 °C for 12 h.
To modify the DMSN inner and outer surfaces with amino groups, the surfactant template CTATos in the pores of DMSN were removed first. The as-synthesized DMSN were dispersed in a solution of hydrochloric acid in ethanol (10 % V/V) and extracted at 78°C for 24 h. This process was repeated for three times. Afterwards, 500 mg of the template-remove DMSN were dispersed in 50 mL toluene by sonication and 200 µL (3-aminopropyl) triethoxysilane were added to the suspension prior to refluxing at 113°C under N2 for 12 h. Thus, DMSN were aminated and then collected after centrifugation, washing with ethanol and drying under vacuum.

Deposition of Pd nanoparticles in the pores of the DMSN (SP)
For synthesis of SP, [PdCl4] 2ions were absorbed onto the inner pore surfaces by coordinating and electrostatic interaction with amino groups. After in-suit reduction by NaBH4, [PdCl4] 2ions converted to Pd 0 leading to aggregation and the generation of Pd 0 nanoparticles. 2 Ultrafine and well-dispersed Pd 0 nanoparticles with diameter of 1-2 nm were formed and simultaneously attached to the inner surface. Typically, the aminated DMSN nanoparticles (100 mg) were dispersed in 10 mL distilled water by sonication for 30 min, followed by the addition of the H2PdCl4 (0.1 mL, 1M) diluted in 2 mL distilled water. After 20 min, a freshly prepared NaBH4 (36 mg in 4 mL cold water) was added into the above aqueous solution under vigorous stirring. After mixture, the resulting suspension was stirred for another 3 h. Finally, the suspension was centrifuged at 10000 rpm for 10 min to separate the SP. Then, SP was washed by water 3 times and dried under vacuum.

Preparation of azobenzene terminated SP (ASP)
The synthesis of ASP was according to the literature with a little modification. 3 The 40 mg as-prepared SP was dissolved in 8 mL anhydrous ethanol and the solution was sonicated to disperse nanoparticles. The 8 mg Azo-COOH was added into the mixture and sonicated 5 min. The solution was then allowed to stir under room temperature for >24 h under dark. The nanoparticles were separated from solution by centrifugation and washed with water and ethanol respectively. The particles were dried under vacuum. The conjugation of the azobenzene groups on surface of DMSN was confirmed by X-ray photoelectron spectroscopy (XPS) and Fourier transformation infrared spectroscopy (FTIR) spectrum. 6

Capping ASP with β-cyclodextrin (CASP)
The dried 20 mg ASP was dissoved in 10 mL distilled water firstly, then β-CD was added into the solution. The mixture was then stirred and sonicated. After complete dessolvation of CD, the solution was allowed to stir at 4 ͦ C > 24 h to maximize the association of CD and the formation of the pseudorotaxanes. After that, the nanoparticles were separated from solution via centrifugation and then washed using

Inductively coupled plasma mass spectrometry (ICP-MS) instrumentation for quantification of SP
ICP-MS measurements were performed on a TheroScientific Xseries Ⅱinductively coupled plasma mass spectrometer. 0.5 mL of fresh aqua regia was added to the 10 µL sample solution and then the sample was diluted to 10 mL with distilled water.
The amount of 106 Pd and 28 Si were tracked by ICP-MS. The 28 Si was come from SiO2 and 106 Pd was from Palladium nanoparticles. Therefore, The Pd load amount in DMSN was estimated with the following formula:

SP induced allylcarbamate cleavage inside vial
The allylcarbamate cleavage of N-alloc-coumarin, 2, in vitro was carried out to assess the catalytic efficiency of SP and CASP. Brifely, the SP (100 µg mL -1 ) and 10 µM Nalloc-coumarin (40 mM in DMSO) were mixed in water, PBS and DMEM medium respectively at 37 °C . After a period of time, the mixture was centritigated to recover In the groups labeled with *, the CASP was treated with UV light.

Light-gated catalysis of SP and CASP in solution
The catalysts SP or CASP (100 µg mL -1 ) were mixed with 10 µM of N-alloc- The changes of spectra indicated that the Azo modified on the SP still retained the photoresponse activity.

Recyclable test of SP and CASP
The nanocatalysts were collected from the solution after the catalytic reaction had In the groups labeled with *, the recycled CASP was not treated with UV light.

Cellular uptake of the nanocatalysts tracked by ICP-MS
Nanocatalysts SP and CASP (40 µg mL -1 ) were incubated with pre-seeded HeLa cells  Additionally, the biocompatibility of low dose UV light and Vis light were also comfirmed in this study. The incubated cells were exposed to 365 nm UV light (0.12 W cm -2 ) or Vis light for 0, 1, 2, 4, 6, 10 min. After 24 h, the cells were washed with 100 µL PBS three times and treated with MTT solution for 4 h. Subsequently, the optical density (OD) was read at a wavelength of 490 nm as mentioned above.

SP induced allylcarbamate cleavage inside living cells
Experiments analysed by flow cytometry HeLa cells were plated in 6-well paltes.

LC-MS study of light-gated allylcarbamate cleavage in HeLa cells
The Reaction treated by CASP only. Data were presented as mean ± s.d (n=3).

LC-MS study of Suzuki-Miyaura Reaction in HeLa cells
The reaction was carried out on the HeLa cells in 6-well plates using the same The toxicity profiles of 5FU and Pro-5FU were also investigated at different concentrations by performing a cell viability assay. The toxicity profiles of 5FU and Pro-5FU were investigated by performing a cell viability assay ( Figure S31b). Cells were treated with 5FU or Pro-5FU at various concentrations from 10 nM to 1 mM.
Although 5FU showed toxicity as its concentration increased, Pro-5FU retained a high cellular viability at all concentrations studied.
As shown in Figure S31c, cells that were incubated with CASP + UV light showed elevated toxicity at a higher concentration of Pro-5FU, while CASP alone retained ∼100% cell viability even at higher concentrations of Pro-5FU. As expected, Pro-5FU was not toxic at any concentrations used. These results indicating that the toxicity was coming from the intracellular conversion of Pro-5FU into 5FU by gated-catalysis, but not the catalyst itself.

Coumarin fluorescence standard curve
We can estimate the concentration of the product after catalytic reaction, using the fluorescence intesnsity coming from the coumarin product. For this purpose, a fluorescence standard curve for coumarin was obtained. The stock solution of coumarin (0.1 mM) was prepared in water. Successive dilution was done, and the fluorescence intensity of prepared solutions were directly measured by fluorescence spectrometer.
Supplementary Figure 32. The standard curve of coumarin. Data were presented as mean ± s.d (n=3).

Rhodamine 110 fluorescence standard curve
The fluorescence standard curve for rhodamine 110 was also obtained as described above.