Articleshttps://doi.org/10.1038/s41565-019-0373-6

Carbon-dot-supported atomically dispersed gold as a mitochondrial oxidative stress amplifier for cancer treatmentNingqiang Gong1,2,3, Xiaowei Ma1, Xiaoxia Ye1, Qunfang Zhou4, Xiaoai Chen1, Xiaoli Tan   5, Shengkun Yao6, Shuaidong Huo1, Tingbin Zhang1, Shizhu Chen1,7, Xucong Teng2, Xixue Hu1, Jie Yu4, Yaling Gan1, Huaidong Jiang6, Jinghong Li   2* and Xing-Jie Liang   1,3*

1Laboratory of Controllable Nanopharmaceuticals, Chinese Academy of Sciences (CAS) Center for Excellence in Nanoscience and CAS Key Laboratory for Biomedical Effects of Nanomaterials and Nanosafety, National Center for Nanoscience and Technology, Beijing, China. 2Department of Chemistry, Key Laboratory of Bioorganic Phosphorus Chemistry & Chemical Biology, Tsinghua University, Beijing, China. 3University of Chinese Academy of Sciences, Beijing, China. 4Department of Interventional Ultrasound, Chinese PLA General Hospital, Beijing, China. 5School of Environment and Chemical Engineering, North China Electric Power University, Beijing, China. 6School of Physical Science and Technology, Shanghai Tech University, Shanghai, China. 7Key Laboratory of Medicinal Chemistry and Molecular Diagnosis of the Ministry of Education, Hebei University, Baoding, China. *e-mail: jhli@mail.tsinghua.edu.cn; liangxj@nanoctr.cn

SUPPLEMENTARY INFORMATION

In the format provided by the authors and unedited.

NaTure NaNoTeCHNoLoGY | www.nature.com/naturenanotechnology

2

Table of Contents 8

Supplementary Figures and Tables. ................................................................................ 4 9

Supplementary Fig 1. Characterization of CDs. ......................................................... 4 10

Supplementary Fig 2. Characterization of the as-synthesized CAT-g. .................. 5 11

Supplementary Fig 3. HAADF image and EELS analyze of CAT-g NPs. .............. 6 12

Supplementary Fig 4. Synthesis and characterization of GSH-rhodamine. .......... 7 13

Supplementary Fig 5. The k3-weighted spectra in k space for Au foil and Au/CDs.14

........................................................................................................................................... 8 15

Supplementary Fig 6. GSH depletion by CAT-g. ....................................................... 9 16

Supplementary Fig 7. HAADF image and EELS analyze of GSH-pre-incubated 17

CAT-g nanoparticle....................................................................................................... 10 18

Supplementary Fig 8. Cell viability and GSH level in HepG-2 cells when treated 19

with CAT-g ..................................................................................................................... 11 20

Supplementary Fig 9. Synthesis and characterization of 1 and 2. ........................ 12 21

Supplementary Fig 10. TEM analysis of MitoCAT-g. .............................................. 13 22

Supplementary Fig 11. Profile of CA release from MitoCAT-g at pH 5.0 and pH 23

7.4. .................................................................................................................................. 14 24

Supplementary Fig 12. Cytotoxicity of various formulations against cancer and 25

normal cells. ................................................................................................................... 15 26

Supplementary Fig 13. Evaluation of the cytotoxicity of MitoCAT-g in SOD2- and 27

CAT-overexpression cells by Cck-8 and colony formation assays ....................... 16 28

Supplementary Fig 14. CLSM images of HepG-2 cells treated with FITC-labeled 29

MitoCAT-g for 0.5 h. The experiments were repeated three times. ...................... 17 30

Supplementary Fig 15. Fluorescence colocalization of MitoCAT-g with 31

Lyso-Tracker or with Mito-Tracker. ............................................................................ 17 32

Supplementary Fig 16. TEM ultrastructure of untreated HepG-2 cancer cells. .. 18 33

Supplementary Fig 17. GSH depletion by MitoCAT-g. ........................................... 19 34

Supplementary Fig 18. MitoCAT-g causes ROS production in the mitochondria of 35

HepG-2 cancer cells. .................................................................................................... 20 36

Supplementary Fig 19. Intracellular ROS levels determined by flow cytometry. 21 37

Supplementary Fig 20. Kinetic analysis of MitoCAT-g-induced mitochondrial GSH 38

depletion and ROS production in HepG-2 cancer cells .......................................... 22 39

3

Supplementary Fig 21. Activity of mitochondrial respiratory chain complexes ... 23 40

Supplementary Fig 22. DNA fragmentation in HepG-2 cancer cells. ................... 24 41

Supplementary Fig 23. MitoCAT-g induces apoptosis in HepG-2 cancer cells. . 25 42

Supplementary Fig 24. Mitochondrial depolarization imaged by CLSM. ............. 26 43

Supplementary Fig 25. MitoCAT-g decreases the O2 consumption of cells. ....... 27 44

Supplementary Fig 26. Hemolytic analysis of mouse red blood cells after 45

treatment with MitoCAT-g in PBS. ............................................................................. 28 46

Supplementary Fig 27. Selectivity of MitoCAT-g in the in vivo orthotopic tumour 47

model .............................................................................................................................. 29 48

Supplementary Fig 28. Photographs of H&E-stained sections of tumour and 49

major organs after MitoCAT-g treatment in the subcutaneous tumour model. ... 30 50

Supplementary Fig 29. Photographs of H&E-stained sections of tumour and 51

major organs after MitoCAT-g treatment in the PDX model. ................................. 30 52

Supplementary Fig 30. Levels of apoptosis and proliferation in the tumour tissue.53

......................................................................................................................................... 31 54

Supplementary Fig 31. Immunotoxicity analysis of the blood of mice at 24 h after 55

drug injection. ................................................................................................................ 31 56

Supplementary Table 1. Structural parameters of Au foil and AU/CDs at the Au 57

L3-edge ........................................................................................................................... 32 58

Supplementary Table 2. Blood biochemistry analysis of mice at 24 h after the last 59

administration of MitoCAT-g or saline. ...................................................................... 32 60

61

4

Supplementary Figures and Tables. 62

63

64

65

66 67

Supplementary Fig 1. Characterization of CDs. A), X-ray photoelectron spectroscopy 68

(XPS) of CDs. Experiments were repeated three times. B), Fourier transform infrared 69

spectroscopy (FTIR) of CDs. Experiments were repeated three times. C), dynamic light 70

scattering (DLS) of the synthesized CDs. Data was shown as mean ± SD (n=3). 71

Experiments were repeated three times. D), ICP-MS analysis of the pH-dependent Au 72

adsorption by the CDs at different pH values. Data were shown as mean ± SD (n=3). 73

Experiments were repeated three times. 74

75

76

5

77 78

Supplementary Fig 2. Characterization of the as-synthesized CAT-g. A), Uv-vis 79

spectra of the as-synthesized CDs and CAT-g. Experiments were repeated three times. B), 80

DLS histogram of CAT-g. Data was shown as mean ± SD (n=3). Experiments were 81

repeated three times. C), zeta potential and D), FTIR of the as-synthesized CAT-g. 82

Experiments were repeated three times. 83

84

6

85

86

Supplementary Fig 3. High-Angle Annular Dark Field image and electron energy 87

loss spectroscopy analyze of CAT-g NPs. A), HAADF image of CAT-g NPs. Silicon 88

nitride membrane was used to support the sample. Experiments were repeated three 89

times. B), 2-D array of the EEL spectrum image of a single CAT-g nanoparticle. 90

Experiments were repeated three times. C), EEL spectrum of the red box in A). 91

Experiments were repeated three times. 92

93

94

95

7

96 97

Supplementary Fig 4. Synthesis and characterization of GSH-rhodamine. A), The 98

synthetic route for preparation of GSH-rhodamine, which was used to detect GSH 99

depletion. B), MALDI-TOF MS spectrum of the GSH-rhodamine. Experiments were 100

repeated three times. 101

8

102

103

Supplementary Fig 5. The k3-weighted spectra in k space for Au foil and Au/CDs. A), 104

Au foil, B), Au/CDs1, C), Au/CDs2 (CAT-g) and D), Au/CDs3. Experiments were repeated 105

two times. 106

9

107

108

Supplementary Fig 6. GSH depletion by CAT-g. A). 1 mL of reduced GSH (100 μM) 109

was treated with various concentrations of CAT-g for 2 h and the free GSH in the 110

supernatant was determined using a GSH quantification kit. B). The specificity of CAT-g in 111

depleting GSH rather than thiol groups from mitochondrial proteins, cellular proteins or 112

bovine serum albumin (BSA) was evaluated using a fluorescence quenching experiment. 113

Proteins in mitochondria (MP) and whole cells (CP) were extracted, the protein 114

concentration and thiol concentration were determined, and the proteins were labeled with 115

rhodamine (Rh.). After that, 1 mL of mitochondrial protein-rhodamine (MP-Rh.), cellular 116

protein-rhodamine (CP-Rh.) or GSH-rhodamine (GSH-Rh.) with an equivalent thiol group 117

concentration of 100 μM was added to a 5 mL Eppendorf tube containing 100 μg of CAT-g. 118

A mixture of free rhodamine and CAT-g was used as a control. Two hours later, the 119

fluorescence intensity of the mixtures was determined, and the fluorescence quenching 120

effect was evaluated by comparing the fluorescence before and after the addition of NPs. 121

Data is presented as relative fluorescence intensity. C), Schematic showing how the steric 122

10

effect enables CAT-g to selectively deplete GSH rather than -SH groups from proteins. 123

The mean ± s.d. from three independent replicates is shown. ***P < 0.001, calculated by 124

two-way ANOVA. 125

126

127

128

129

130

131

132

133

134

Supplementary Fig 7. High-Angle Annular Dark Field image and electron energy 135

loss spectroscopy analyze of GSH-pre-incubated CAT-g nanoparticle. A), HAADF 136

image of GSH-pre-incubated CAT-g NPs. Silicon nitride membrane was used to support 137

the sample. Experiments were repeated two times. B), and C), 2-D array analyze of the 138

distribution of carbon and sulfur elements in a single GSH-pre-incubated CAT-g 139

nanoparticle. Experiments were repeated two times. D), EEL spectrum of the red box in A). 140

E), Average GSH conjugation number per Au atoms determined using HPLC, data were 141

shown as mean ± SD (n=5). Experiments were repeated five times. We calculated that on 142

average each Au/CDs2 contains 15.5 active Au atoms whereas Au/CDs3 contains 9.6 143

active Au atoms in each nanoparticle. 144

11

145

Supplementary Fig 8. Cell viability and GSH level in HepG-2 cells when treated with 146

CAT-g. A), Cell viability test of HepG-2 cells after CAT-g (2, 4, 8, 16, 32, 64 μg/mL) 147

treatment for 24 h. Experiments were repeated three times and mean ± SD (n=3) was 148

shown. B) and C), Cytoplasm and mitochondrial GSH level after CAT-g treatment for 24 h. 149

Data are presented as nmol/mg protein. Shown are box plots from six independent 150

replicates. The box represents the 25th-75th percentiles. Whiskers represent minimum 151

and maximum. *P < 0.05, ***P < 0.001, by Student’s t test. 152

12

153

154

155

Supplementary Fig 9. Synthesis and characterization of 1 and 2. A), The synthetic 156

route of 1 and 2, starting from 1,1,1- tris(hydroxymethyl)ethane and cinnamaldehyde. CDI, 157

1,1’-carbonyldiimidazole. B) and C), 1H-NMR spectra of 1 and 2, respectively. 158

Experiments were repeated three times. 159

13

160

161

Supplementary Fig 10. TEM analysis of MitoCAT-g. A), Cartoon model and B), TEM 162

images of MitoCAT-g. Experiments were repeated three times. 163

164

165

166

167

168

169

170

171

172

173

174

175

176

14

177

178

Supplementary Fig 11. Profile of CA release from MitoCAT-g at pH 5.0 and pH 7.4. A) 179

Standard curve showing the absorption at 286 nm of increasing concentrations of CA. 180

Experiments were repeated three times. The correlation between concentration and 181

absorption at 268 nm was determined using Microsoft Excel software. B) CA release at 182

the physiological pH value (pH 7.4) and the pH of the acidic endosomal acidic 183

environment (pH 5.0). Data were shown as mean ± SD (n=3) from three independent 184

experiments. 185

15

186

187

Supplementary Fig 12. Cytotoxicity of various formulations against cancer and 188

normal cells. Human liver cancer cells (HepG-2, BEL-7404, HCCC-9810) and human 189

normal liver cells (L02, QSG-7701) or primary mouse liver cells were treated with 190

equivalent amounts of CDs, CDs-CA, CDs-TPP, CDs-CA-TPP, CA, CAT-g, CAT-g-TPP, 191

CAT-g-CA, 3-hydroxy-4-pentenoate or MitoCAT-g for 24 h. The cell viability was then 192

evaluated using Cck-8 kit (A) or colony formation assay (B, C and D, the experiments 193

were repeated three times). The mean ± s.d. (n=3) from three independent replicates is 194

plotted in A). ***P < 0.001, based on student’s t test. HP represents 195

3-hydroxy-4-pentenoate. 196

16

197

198

Supplementary Fig 13. Evaluation of the cytotoxicity of MitoCAT-g in SOD2- and 199

CAT-overexpression cells by Cck-8 and colony formation assays. A), Cell viability 200

and B), representative colony formation assay results of HepG-2 (SOD2-mito-CAT) cells 201

when treated with MitoCAT-g for 24h. Experiments were repeated three times. The mean 202

± s.d. (n=3) from three independent replicates is plotted in A). NS, not significant. 203

204

205

17

206

207

208

Supplementary Fig 14. CLSM images of HepG-2 cells treated with FITC-labeled 209

MitoCAT-g for 0.5 h. The experiments were repeated three times. 210

211

212

213

214

215

216

217

Supplementary Fig 15. Fluorescence colocalization of MitoCAT-g with Lyso-Tracker 218

or with Mito-Tracker. Quantification of fluorescence colocalization between Lyso-Tracker 219

and MitoCAT-g (Lyso&NPs, black column) or Mito-Tracker and MitoCAT-g (Mito&NPs, red 220

column) after HepG-2 cancer cells were incubated with MitoCAT-g for 2.5 h. 10 cells were 221

analyzed and data were shown as mean ± s.d. (n=10). Data were obtained using 222

Image-pro Plus 6.0 software. 223

224

18

225

226

Supplementary Fig 16. TEM ultrastructure of untreated HepG-2 cancer cells. A) 227

Ultrastructural analysis of untreated HepG-2 cancer cells by TEM (M, mitochondrion; N, 228

nucleus). Experiments were repeated three times. B) Quantitative analysis of the 229

mitochondrion form factor in untreated cells and MitoCAT-g-treated cells. Shown are box 230

plots from mitochondria in 25 cells. The box represents the 25th-75th percentiles. 231

Whiskers represent minimum and maximum. ***P < 0.001, analyzed by two-tailed 232

Student’s t-test. 233

234

19

235 236

Supplementary Fig 17. GSH depletion by MitoCAT-g. A) Mitochondrial GSH level when 237

HepG-2 cancer cells were incubated with TPP-CAT-g-GSH (TPP-CAT-g NPs 238

pre-incubated with GSH to occupy the gold atoms) for 24 h. B) Cells were incubated with 239

MitoCAT-g for 0, 1, 2, 4, 8 and 12 h and then the cellular GSH levels were determined. 240

The mean ± s.d. from three independent replicates is shown. NS, not significant, ***P < 241

0.001, from two-sided Student’s t test. 242

243

20

244

245

Supplementary Fig 18. MitoCAT-g causes ROS production in the mitochondria of 246

HepG-2 cancer cells. A), Cells were treated with various formulations with an equivalent 247

dose of 64 μg/mL for 2.5 h and then the cellular ROS levels were visualized using a 248

fluorescent probe (DCFH-DA). Green channel, DCFH-DA; red channel, Mito-Tracker; 249

scale bar: 20 μm. The top row shows DCFH-DA fluorescence. The bottom row shows 250

merged images. CAT-g-GSH indicates CAT-g nanoparticles pre-incubated with GSH. 251

Experiments were repeated three times. B), Cells were treated with MitoCAT-g (64 μg/mL) 252

and an oxidation-insensitive probe (DCF-DA) was used as control dye to demonstrate that 253

the enhanced fluorescence of DCFH-DA originates from the oxidation of the dye. 254

Experiments were repeated three times. C) and D), Images of HepG-2 and HepG-2 255

(SOD2-mito-CAT) cells treated with MitoCAT-g (64 μg/mL). Green fluorescence results 256

from activation of DCFH-DA. Experiments were repeated three times. 257

258

21

259 260

Supplementary Fig 19. Intracellular ROS levels determined by flow cytometry. 261

HepG-2 cancer cells were treated with increasing concentrations of MitoCAT-g for 2.5 h. 262

Cellular ROS levels were determined using flow cytometry (ROS probe: DCFH-DA). 263

Experiments were repeated three times. 264

265

266

267

268

269

270

271

272

273

274

275

276

277

22

278

Supplementary Fig 20. Kinetic analysis of MitoCAT-g-induced mitochondrial GSH 279

depletion and ROS production in HepG-2 cancer cells. A) HepG-2 cells were 280

incubated with MitoCAT-g (64 μg/mL) and the mitochondria were isolated and lysed at 281

different time points. The GSH concentration in the lysates was then determined. The 282

time-course of ROS production in mitochondria was evaluated by measuring the 283

fluorescence of DCFH-DA in the lysate. The mean ± s.d. (n=3) from three independent 284

replicates is shown. Experiments were repeated three times. B), GSH levels were 285

determined in mitochondria and cytoplasm after MitoCAT-g treatment for 2 h. Shown are 286

box plots from six independent experiments. The box represents the 25th-75th percentiles. 287

Whiskers represent minimum and maximum. NS indicates not significant, ***P < 0.001, 288

from two-sided Student’s t test. 289

290

291

23

292

293

Supplementary Fig 21. Activity of mitochondrial respiratory chain complexes. 294

HepG-2 cells were incubated with MitoCAT-g (64 μg/mL) for 24 h and the mitochondria 295

were then isolated and lysed. Complex I activity was determined using a 296

ubiquinone-reduction method; Complex II activity was measured using a 297

2,6-dichlorophenolindophenol (DCIP) reduction method; Activity of complex III was 298

measured with a cytochrome c reduction method; Complex IV activity was determined 299

using a cytochrome c oxidation method. The mean ± s.d. (n=3) from three independent 300

replicates is shown. NS, not significant, ***P < 0.001, from Student’s t test. 301

302

24

303

304

Supplementary Fig 22. DNA fragmentation in HepG-2 cancer cells. A), Cells were 305

incubated with MitoCAT-g (16, 32 and 64 μg/mL) for 24 h and the level of nucleosomal 306

DNA fragmentation was detected. Experiments were repeated three times. B), HPLC 307

analysis of 8-hydroxy-2’-deoxyguanosine (8-OHdG) production in mitochondrial DNA after 308

cells were treated with MitoCAT-g (64 μg/mL). The box represents the 25th-75th 309

percentiles. Whiskers represent minimum and maximum. Experiments were repeated 310

three times. ***P < 0.001, from Student’s t test. 311

312

313

314

315

316

317

318

319

25

320 321

Supplementary Fig 23. MitoCAT-g induces apoptosis in HepG-2 cancer cells. Cells 322

were incubated with various formulations for 24 h, then stained with Propidium Iodide and 323

Annexin V-FITC. Stained cells (8000 for each group) were subjected to flow cytometric 324

analysis to determine the distribution of cells. Experiments were repeated three times. 325

326

327

328

329

330

26

331

332

Supplementary Fig 24. Mitochondrial depolarization imaged by CLSM. HepG-2 333

cancer cells were incubated with different amounts of MitoCAT-g (2, 4, 8, 16, 32 and 64 334

μg/mL) for 24 h. The MMP was then evaluated by staining the cells with JC-1 dye and 335

observing them using confocal microscopy. J-monomers show green fluorescence while 336

J-aggregates show red fluorescence. Experiments were repeated three times. Scale bar: 337

50 μm. 338

27

339

340

Supplementary Fig 25. MitoCAT-g decreases the O2 consumption of cells. A), Digital 341

picture of the self-referencing electrode. B) Bright-field microscopy image showing the 342

self-referencing electrode positioned next to a single HepG-2 cell. Experiments were 343

repeated three times. C), Group data (ten cells for each group) for basal respiration 344

oxygen flux levels of single untreated HepG-2 cancer cells (control) or single cells treated 345

with different concentrations of MitoCAT-g (2, 4, 8, 16, 32 and 64 μg/mL) for 24 h. Data 346

were shown as mean±SD (n=10). 347

28

348

Supplementary Fig 26. Hemolytic analysis of mouse red blood cells after treatment 349

with MitoCAT-g in PBS. The percentage of hemolysis was quantified by the release of 350

hemoglobin into the medium, then plotted as a function of the concentration of MitoCAT-g. 351

Water and PBS were used as positive control and negative control, respectively. 352

Experiments were repeated three times. 353

354

29

355

Supplementary Fig 27. Selectivity of MitoCAT-g in the in vivo orthotopic tumour 356

model. HepG-2 and HepG-2 (SOD2-mito-CAT) orthotopic liver tumour models were 357

constructed and the mice were treated 8 times at 2-day intervals by image-guided 358

intratumoural injection of MitoCAT-g (64 μg/mouse). After that, the mice were killed, and 359

the tumour tissue and tumour-adjacent liver were isolated and lysed. GSH and GSSG 360

levels were determined and the GSH/GSSG ratio was calculated (A-C) in tumour tissues 361

and tumour-adjacent liver in the HepG-2 model. The levels of malonaldehyde (MDA，D 362

and G), protein carbonyl (E and H) and 8-OHdG (F and I) in tumour tissues and 363

tumour-adjacent liver were evaluated in both the HepG-2 and HepG-2 (SOD2-mito-CAT) 364

models. Shown are box plots with whiskers from 5 to 95 percentile (n=6). The box 365

represents the 25th-75th percentiles. Whiskers represent minimum and maximum. NS 366

indicates not significant, ***P < 0.001, from Student’s t test. 367

30

368

369

Supplementary Fig 28. Photographs of H&E-stained sections of tumour and major 370

organs after MitoCAT-g treatment in the subcutaneous tumour model. Tumours and 371

organs (heart, liver, spleen, kidney and lung) were isolated after the mice were killed at 372

day 31. The scale bar represents 200 μm. Experiments were repeated two times. 373

374

375

376

377

378

379

380

Supplementary Fig 29. Photographs of H&E-stained sections of tumour and major 381

organs after MitoCAT-g treatment in the PDX model. Tumours and organs (heart, liver, 382

spleen, kidney and lung) were isolated after the mice were killed at day 94. Scale bar: 200 383

μm. Experiments were repeated two times. 384

385

31

386

387

Supplementary Fig 30. Levels of apoptosis and proliferation in the tumour tissue. 388

Bcl-2, Ki-67 and caspase-3 immunohistochemical (IHC) staining of tumour tissues to 389

evaluate the proliferation of tumour cells. TUNEL assay was also used to evaluate the 390

DNA fragmentation level in tumour tissue. Scale bar: 200 μm. Experiments were repeated 391

two times. 392

393

394

395

396

397

398

Supplementary Fig 31. Immunotoxicity analysis of the blood of mice at 24 h after 399

drug injection. Serum levels of interleukin-1β (IL-1β), IL-6 and tumour necrosis factor-α 400

(TNF-α) were determined using ELISA kits. The data are presented as the means ± s.d. 401

(n=3). ***P < 0.001 (P value from Students’ t test); NS, not significant. 402

403

32

Supplementary Table 1. Structural parameters of Au foil and AU/CDs at the Au 404

L3-edge 405

406

407

408

409

Supplementary Table 2. Blood biochemistry analysis of mice at 24 h after the last 410

administration of MitoCAT-g or saline. 411

412

413

414

ALT: Alanine aminotransferase 415

AST: Aspartate transaminase 416

BUN: Blood urea nitrogen 417

CREA: Creatinine 418

Experiments were repeated three times and mean±SD (n=3) was shown. 419

420