Mitochondria targeting combined with methyl modification of novel resveratrol derivatives enhances anti-tumor activity

Abstract

Mitochondria-targeting methyl modification compounds A1–A6 were synthesized by introducing TPP⁺ into the pharmacophore. A4 ((E)-Triphenyl(4-(4-(3,4-dimethylstyryl)phenoxy)butyl)phosphoniumiodide) could selectively accumulate in the mitochondria and exert excellent anticancer activity by both cell cycle arrest in G0/G1 and apoptosis induction in the mitochondrial pathway. The target mitochondria drug design is an effective strategy for exploiting the drug potential for cancer therapy.

---

Cancer is the second leading cause of death worldwide, with a rising incidence rate, and is a serious public health problem, although numerous preventive measures and effective treatments have been developed.¹ So, substantial progress and new therapeutic strategies for it are still needed. In therapeutic oncology, selective targeting of a drug to the appropriate subcellular compartment is an effective way to improve the selectivity, sharpen its effectiveness and limit side effects.² Given the vital role in cancer initiation, metabolism, progression, and metastasis, mitochondria have emerged as an intriguing target for anti-cancer drugs.^3–5

As we know, mitochondria are indispensable for energy production and biomolecule synthesis, and are essential for cancer cells’ rapid proliferation. Furthermore, mitochondria are also responsible for various biological processes including redox balance, calcium homeostasis, cell cycle progression, differentiation, apoptosis and cellular signalling.^6–8 In normal cells, mitochondria conduct a controlled growth-death cycle regulation. However, mitochondrial dysfunction has been observed in numerous cancer cells.^9,10 Obviously, mitochondria play an important role in the survival of cancer cells and have multiple ways to signal overgrowth to the cancer cells. So, mitochondria are potential targets of therapeutic agents against cancer. In addition, due to the higher hyperpolarized mitochondrial membrane of cancer cells (−220 mV) than that of normal cells (−160 mV), certain positively charged molecules can quickly and selectively accumulate inside the cancer's mitochondria, which offer further support for developing mitochondria targeting drugs.^11,12

ROS are mainly derived from mitochondria metabolism, and are necessary for signaling under normal conditions and ROS are crucial to maintain normal cellular homeostasis at low levels.¹³ However, cancer cells generate high levels of ROS to maintain a neoplastic state, while increasing ROS levels may exert an inhibitory effect on cell proliferation by cell cycle arrest and apoptosis induction.¹⁴ So, cancer cells may be more easily undergo a cell death response to drug stimuli in a pro-oxidant manner, which may be an effective therapeutic strategy.

Some natural or synthetic products have been proven to decelerate or reverse carcinogenesis. Resveratrol (trans-3,5,4′-trihydroxy-trans-stilbene, RES), a kind of polyphenolic phytoalexin with a unique skeleton, is well known for its potential against oxidative stress, angiogenesis, inflammation and especially carcinogenesis.¹⁵ However, its therapeutic potential is restricted due to its low bioavailability and rapid metabolism.¹⁶ Therefore, structure modification is often chosen to overcome these hurdles.

According to the heterogeneity of mitochondria and ROS in both normal cells and cancer cells described above, mitochondrial-targeted molecules emerge as an effective means to involve the mitochondrion as an intracellular “reaction chamber” to accumulate bioactive molecules. However, the mitochondrial membrane consists of a double-layered membrane to prevent external substances from crossing it. If penetrated, the molecules are expected to overcome the activation energy.¹¹ A widely used method is conjugation of a lipophilic cation to the bioactive molecules, which can lower this activation to penetrate the membrane and accumulate selectively in the mitochondrial matrix in response to the negative membrane potential.¹⁷ To date, there have been various lipophilic cations, such as alkyltriphenylphosphonium cations, methyltriphenylphosphonium cations, tetraphenylphosphonium cations, triphenylphosphonium cations (TPP⁺), cyanine cations, rhodamine, JC-1, and cationic peptides that can be linked to bioactive compounds to improve their mitochondrial-target uptake and increase their biological activities.^17,18

Due to the stability of TPP⁺ in the biological system, low chemical reactivity toward cellular components and simple synthesis and purification, the biophysics of TPP⁺ across phospholipid bilayers has been extensively studied and it has been proven to facilitate mitochondria targeting by small molecules containing TPP⁺.^17–20 In addition, methylation is widely used to optimize many properties of a drug candidate, and C–H methylation has been proven to play a “Magic Methyl” effect on improving a drug lead's physicochemical properties, such as lengthening the half-life, changing the solubility, improving selectivity, increasing potency, etc.^21,22 Therefore, the aim in this work was to synthesize a novel anticancer compound, which targets mitochondria. On the one hand we hypothesized that butyl triphenylphosphonium was introduced to one benzene ring of the stilbene skeleton molecule to target mitochondria. On the other, methyl groups were decorated on another benzene ring at different locations to obtain molecules with better antitumor activity than that of resveratrol. Then we synthesized the target compounds A1–A6 (Scheme 1) and screened their toxic activity, and then found that A4 showed remarkable tumor inhibitory activity and can accumulate in the mitochondria. Subsequently, the action mechanism research showed that A4 could inhibit tumor cell proliferation through both cycle arrest and apoptosis pathways, which provided important information for the future design of anti-tumor drug molecules.

Scheme 1 Structure of compounds A0–A6 and resveratrol.

First of all, the target compounds A1–A6 were synthesized and confirmed by NMR and HRMS (Scheme S1, for details and characterization see the ESI†). Briefly, we firstly synthesized the prerequisite intermediate Z with p-hydroxybenzaldehyde as starting material by a nucleophilic substitution reaction. Subsequently, we prepared the Wittig–Horner reagents E1–E6 starting from material methyl-benzaldehyde B1–B6 through a reduction reaction, functional group transformation and interaction with triethoxyphosphine. With that, F1–F6 were achieved via a Wittig–Horner reaction between E1–E6 and Z. Lastly, the activation of sodium iodide made the conversion of chloride to iodine, which reacted with triphenyl phosphine to obtain compounds A1–A6. To study the influence of butyl triphenylphosphonium on the toxicity of drug molecules, we also synthesized compound A0. Namely, 1-Iodobutane and triphenylphosphine were heated to reflux for 6 h in toluene to obtain the white solid A0 with 73% yield (data see the ESI†).

With the compounds in hand, the MTT assay²³ was operated to investigate their anti-proliferation effects in HeLa and A549 cells. As shown in Table 1, the cytotoxicity of compounds A1–A6 was significantly improved compared to that of RES, and showed a dose-dependent manner (Fig. 1A and B). The poor activity of compound A0 implied the butyl triphenyl phosphine part of compounds A1–A6 exerted little toxicity under 80 μM (Table 1), which supported that A0 may target mitochondria to improve the compounds’ uptake and increase their anti-tumor effects. In addition, the results showed that the introduction of methyl on the benzene ring also promoted the anti-tumor activity and the methyl modification on different positions showed different activity. A2 with the methyl modification in the 4 position of the benzene ring has superior bioactivity to A3 with that in the 3 position. Compounds A4–A6 modified with two methyl groups showed better bioactivity than those with one methyl. In particular, A4 with 2 methyl groups introduced with one methyl modification in the 4 position of the benzene ring, exhibited the most dominant anticancer activity with an IC₅₀ value of 2.22 ± 0.35 μM in HeLa cells and 2.81 ± 0.24 μM in A549 cells (Table 1). These results indicate that the synthetic A1–A6, especially A4 could efficiently kill the cancer cells.

Table 1 The cytotoxicity of resveratrol and compounds A0–A6

Compounds	IC50^a (μM)
	HeLa	A549
a The results are shown as the mean ± SD for 4 replications in 3 separate trials after treatment with compounds for 24 h.
A0	>80	>80
A1	8.95 ± 1.63	8.33 ± 0.83
A2	4.02 ± 0.35	3.99 ± 0.23
A3	5.13 ± 0.52	7.57 ± 1.65
A4	2.22 ± 0.35	2.81 ± 0.24
A5	4.95 ± 0.58	6.51 ± 0.78
A6	3.92 ± 0.63	6.12 ± 1.10
RES	>80	>80

---

Fig. 1 Anti-tumor proliferative effects of compounds A1–A6. (A) and (B) Compounds A1–A6 showed dose-dependent anti-proliferative effects after HeLa and A549 cell lines were treated with gradient concentrations of compounds A1–A6 for 24 h.

Encouraged by the above results, we further evaluated the mitochondria localization capability of A4 by the mitochondrial colocalization assay in HeLa cells. Because A4 contains a lipophilic cations TTP⁺, which can tend to accumulate in mitochondria inner transmembrane as described above. After, the cells were co-incubated with A4 and commercial mitochondrial dye (Mito-Tracker Red CMXRos), and then washed with PBS three times before imaging. The fluorescence of the MitoTracker Red was observed in the red channel, and A4 emitted fluorescence in the blue channel. The color pairs of each pixel of both channels displayed a highly correlated plot with the Pearson's coefficient r = 0.868 (Fig. 2A and B), which verified that A4 could target the mitochondria and accumulate in it.

Fig. 2 A4 is enriched in cell mitochondria. (A) The laser colocalization images of HeLa cells co-incubated with A4 and Mito-Tracker Red CMXRos. (B) Intensity scatter plot of A4 and Mito-Tracker Red. Blue channel: 430–470 nm, E_x = 405 nm; Red channel: 570–620 nm, E_x = 561 nm. Scale bar: 15 μm.

Subsequently, we further explore its antiproliferation mechanism. We know that uncontrolled cell division is a typical characteristic of carcinogenesis, and cell cycle arrest, as an important way of controlling the cell growth, was detected by flow cytometry after incubation of A4 for 24 h in HeLa cells. As shown in Fig. 3A, A4 caused a dramatic decrease in the proportion of cells in the S phase, while the proportion in G0/G1 phase increased significantly in a dose-dependent manner with A4 treatment for 24 h. In addition, the cyclin D1, cyclin E1, and p21 as the key regulated factors of the cell cycle progression from G1 to S in response to DNA damage,^24,25 were determined by western blot. In line with expectations, the cyclin D1 and cyclin E1 expression was remarkably decreased after treatment with A4 for 24 h (Fig. 3B and C), whereas the cell cycle inhibitor p21 increased significantly. Obviously, the results above indicated that A4 induced cell cycle arrest in the G0/G1 phase to inhibit cellular proliferation by down-regulation of the cyclin D1 and cyclin E1 expression in HeLa cells.

Fig. 3 A4 exerted anti-tumor activity by both arresting cell cycle and apoptosis. (A) A4 inhibited the cell proliferation by arresting cell cycle in the G0/G1 phase in a dose-dependent manner for 24 h. (B) and (C) A4 down-regulated the expression of cyclin D1 and cyclin E1 and up-regulated the expression of p21 after A4 treatment for 24 h in HeLa cells. (D) and (E) A4 induced apoptosis by affecting the expression of pro-apoptotic protein Bax and the anti-apoptotic protein Bcl-2. (F) The nuclear chromatin condensation and the formation of apoptotic bodies in A4-treated HeLa cells for 24 h with DAPI staining. The white arrow indicates cell shrinkage, chromatin condensation and the formation of apoptotic bodies. (G) A4 induced apoptosis in a dose-dependent manner for 24 h detected by flow cytometry in HeLa cells. The percentage in each quadrant represents the proportion of normal cells, early apoptotic ones, late apoptotic ones and necrotic ones in the order quadrant 3-4-1-2. (H) The caspase-3 activity increased 2 fold at 8 μM after A4 treatment for 24 h. Each experiment was performed at least 3 times, and the asterisks (**) represent the significant differences (P < 0.01) versus the control group.

Apoptosis, a kind of programmed death, is another important means induced by chemotherapy to inhibit the cell proliferation. It is controlled by intracellular or extracellular signals with typical cell morphological alterations. So, the apoptosis effect was detected with fluorescence microscopy and flow cytometry. Firstly, the nucleus morphology with DAPI staining was observed. As shown in Fig. 3F, compared to the control group, HeLa cells presented the hallmarks of apoptosis with cell shrinkage, chromatin condensation and nuclear fragmentation in a dose-dependent manner after A4 treatment for 24 h. HeLa cells were showing an apparent increase in the Annexin V-FITC and PI positive cell subpopulation by flow cytometry assay (Fig. 3G). Subsequently, we analyzed the expression of pro-apoptotic protein Bax and the anti-apoptotic one Bcl-2,²⁶ and the western blot results demonstrated that A4 not only up-regulated Bax compared to the untreated HeLa cells, but also down-regulated the Bcl-2 (Fig. 3D and E). Obviously, the increased Bax/Bcl-2 ratio may be a vital reason for apoptosis induction in the mitochondria intrinsic pathway. Caspase-3, as an “executioner” caspase, is a frequently activated death protease in cell apoptosis.²⁷ So, we tested if the caspase-3 activity changed or not. As shown in Fig. 3H, compared with untreated cells, the caspase-3 activity increased about 2 fold at 8 μM after A4 treatment for 24 h, which supported that A4 could act as a mitochondria-targeting drug to exert its anti-proliferative effect by apoptosis induction.

Mitochondria, as the cell's power plant, can produce the energy necessary for proliferation. Anticancer drugs could selectively disrupt cancerous mitochondria and the depolarization of mitochondrial membrane potential (MMP, Δψ_m) is a crucial step during the apoptotic process.⁵ Besides, the mitochondrial play a crucial role in cell cycle regulation and a low mitochondrial Δψ_m can drive the cell cycle arrest to interrupt cell cycle progression.²⁶ Subsequently, we determine Δψ_m by rhodamine123 to investigate if mitochondria involved in the apoptotic induction and cell cycle arrest by A4. The flow cytometry analysis showed that A4 could induce a dramatic Δψ_m decrease in a dose- and time-dependent manner (Fig. 4A and B). The fluorescence intensity weakened dramatically after A4 treatment for 18 h (Fig. 4C) detected by fluorescence images, which implied A4 could affect the mitochondrial dysfunction by changing the Δψ_m.

Fig. 4 A4 affects the mitochondrial dysfunction. (A) The Δψ_m was decreased after A4 treatment for 18 h in HeLa. (B) A4 was able to induce dose- and time- dependent MMP decrease in HeLa cells. (C) The rhodamine123 fluorescence intensity declined markedly when treated with A4 for 18 h in a dose-dependent manner. (D) ATP level decreased markedly after treating with a gradient concentration of A4 for 24 h in HeLa cells. (E)–(G) ROS level significantly increased after HeLa cells were treated with A4 for 6 h in a dose-dependent manner. (H) The cells’ viability improved after the DTT (10 mM) was co-incubated with A4. (I) DTT (10 mM) reversed the repression of mitochondrial membrane potential by A4 to some extent. (J) and (K) A4 caused the cytosolic Ca²⁺ release. (L) The fluorescence intensity of dose-dependent changes of Fluo-3 signal in HeLa cells in response to A4. The asterisks (*/**) represent the significant differences (P < 0.05/P < 0.01) versus the control group.

As we know, the mitochondria can generate the energy from ATP through oxidative phosphorylation. So, ATP level is a reliable sign of assessing mitochondrial function. Subsequently, the total ATP level was detected by the Luminometer ATP Assay. As shown in Fig. 4D, the ATP level was remarkably decreased after A4 treatment for 24 h in a dose-dependent manner.

The mitochondria are also the major source of ROS, and one of the strategies for anti-cancer is in enhanced generation of ROS to affect the cell cycle progression and trigger apoptosis.⁵ So, we tested if the ROS levels had changed by flow cytometry and fluorescence imaging after HeLa cells were incubated with indicated concentrations A4 and DCFH-DA as an indicator. As expected, the results showed that A4 could sharply increase the ROS level in a dose-responsive and time-responsive manner (Fig. 4E–G). Furthermore, the cells’ proliferative viability improved after addition of a ROS scavenger, dithiothreitol (DTT) co-incubated with A4 (Fig. 4H). Therefore, we suspected that ROS induced by A4 could cause cytotoxicity by oxidative stress to inhibit the proliferation. More interestingly, the mitochondrial membrane potential was reversed to some extent after dithiothreitol (DTT) was co-incubated with A4 (Fig. 4I), which implied ROS induced by A4 could serve as intracellular signaling agents to regulate mitochondrial function by a potential feedback mechanism.

What's more, mitochondria are crucial organelles in Ca²⁺ homeostasis regulation, and elevation in intracellular Ca²⁺-triggered apoptosis has been proved.^8,28 So, concentration of intracellular free Ca²⁺([Ca²⁺]_i) was measured using Fluo-3 AM as the indicator by flow cytometry and fluorescence imaging after HeLa cells were treated with A4 for 6 h. As shown in Fig. 4J–L, [Ca²⁺]_i increased 2 fold and a similar result was observed by fluorescence imaging in a concentration-dependent manner after treatment with A4 for 6 h, which indicated that the intracellular free Ca²⁺ was associated with A4-induced apoptosis.

Conclusions

In summary, in this work a series of mitochondrial-targeting stilbene compounds A1–A6 were synthesized and A4 presented excellent anti-cancer potential with a methyl modification in the 4 position of the benzene ring. It can accumulate selectively in the mitochondrial and exert its anti-cancer activity by both cell cycle arrest in the G0/G1 phase and apoptosis induction with decreased ATP, elevated ROS and increased cytosolic Ca²⁺. This study further supports that targeting mitochondria is an excellent strategy for enhancing a drug's molecular toxicity in cancer treatment.

Author contributions

XZH and KYF designed the study and finished the manuscript; WJN and CMN performed the experiments. GC and YYZ participated in the inhibitory mechanism experiments, WZY participated in the synthesis of compounds, ZXL analysed the data. All authors read and approved the final manuscript.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

All the authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by the Basic Scientific Research Funds of Hebei province (grant numbers JYT2020006). The funding body played no role in the design of the study; collection, analysis, interpretation of data and in writing the manuscript.

References

1. F. Bray, M. Laversanne and H. Sung, et al., Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA Cancer J. Clin., 2024, 74(3), 229–263,  DOI:10.3322/caac.21834.
2. L. Biasutto, A. Mattarei and M. La Spina, et al., Strategies to target bioactive molecules to subcellular compartments. Focus on natural compounds, Eur. J. Med. Chem., 2019, 181, 111557,  DOI:10.1016/j.ejmech.2019.07.060.
3. B. Yan, L. Dong and J. Neuzil, Mitochondria: an intriguing target for killing tumour-initiating cells, Mitochondrion, 2016, 26, 86–93,  DOI:10.1016/j.mito.2015.12.007.
4. Y. Chen, H. Zhang and H. J. Zhou, et al., Mitochondrial redox signaling and tumor progression, Cancers, 2016, 8(4), 40,  DOI:10.3390/cancers8040040.
5. L. D. Zorova, P. A. Abramicheva and N. V. Andrianova, et al., Targeting mitochondrial for cancer treatment, Pharmaceutics, 2024, 16(4), 444,  DOI:10.3390/pharmaceutics16040444.
6. J. X. Tan and T. Finkel, Mitochondria as intracellular signaling platforms in health and disease, J. Cell Biol., 2020, 219(5), e202002179,  DOI:10.1083/jcb.202002179.
7. Y. Q. Tan, X. Zhang and S. Zhang, et al., Mitochondria: the metabolic switch of cellular oncogenic transformation, Biochim. Biophys. Acta, 2021, 1876, 188534,  DOI:10.1016/j.bbcan.2021.
8. A. Takeuchi, B. Kim and S. Matsuoka, The destiny of Ca (2⁺) released by mitochondria, J. Phys. Sci., 2015, 65, 11–24,  DOI:10.1007/s12576-014-0326-7.
9. T. C. Kenny and K. Birsoy, Mitochondria and Cancer, Cold Spring Harbor Perspect. Med., 2024, 1, a041534,  DOI:10.1101/cshperspect.a041534.
10. G. R. Shubha, Mitochondrial Changes in Cancer, Handb. Exp. Pharmacol., 2017, 240, 211–227,  DOI:10.1007/164_2016_40.
11. M. T. Jeena, S. Kim and S. Jin, et al., Recent Progress in Mitochondria-Targeted Drug and Drug-Free Agents for Cancer Therapy, Cancer, 2019, 12(1), 4,  DOI:10.3390/cancers12010004.
12. L. D. Zorova, V. A. Popkov and E. Y. Plotnikov, et al., Mitochondrial membrane potential, Anal. Biochem, 2018, 552, 50–59,  DOI:10.1016/j.ab.2017.07.009.
13. E. C. Cheung and K. H. Vousden, The role of ROS in tumour development and progression, Nat. Rev. Cancer, 2022, 22(5), 280–297,  DOI:10.1038/s41568-021-00435-0.
14. S. Saikolappan, B. Kumar and G. Shishodia, et al., Reactive oxygen species and cancer: A complex interaction, Cancer Lett., 2019, 452, 132–143,  DOI:10.1016/j.canlet.2019.03.020.
15. M. S. Bhuia, R. Chowdhury and M. A. Akter, et al., A mechanistic insight into the anticancer potentials of resveratrol: Current perspectives, Phytother. Res., 2024, 38(8), 3877–3898,  DOI:10.1002/ptr.8239.
16. E. Wenzel and V. Somoza, Metabolism and bioavailability of trans-resveratrol, Mol. Nutr. Food Res., 2005, 49, 472–481,  DOI:10.1002/mnfr.200500010.
17. J. Zielonka, J. Joseph and A. Sikora, et al., Mitochondria-Targeted Triphenylphosphonium-Based Compounds: Syntheses, Mechanisms of Action, and Therapeutic and Diagnostic Applications, Chem. Rev., 2017, 117(15), 10043–10120,  DOI:10.1021/acs.chemrev.7b00042.
18. R. A. Smith, R. C. Hartley and M. P. Murphy, Mitochondria-targeted small molecule therapeutics and probes, Antioxid. Redox Signaling, 2011, 15, 3021–3038,  DOI:10.1089/ars.2011.3969.
19. M. F. Ross, G. F. Kelso and F. H. Blaikie, et al., Lipophilic triphenylphosphonium cations as tools in mitochondrial bioenergetics and free radical biology, Biochemistry, 2005, 70, 222–230,  DOI:10.1007/s10541-005-0104-5.
20. M. M. Shchepinova, A. G. Cairns and T. A. Prime, et al., MitoNeoD: A mitochondria-targeted superoxide probe, Cell Chem. Biol., 2017, 24, 1285–1298.e12,  DOI:10.1016/j.chembiol.2017.08.003.
21. D. Aynetdinova, M. C. Callens and H. B. Hicks, et al., Installing the “magic methyl”-C–H methylation in synthesis, Chem. Soc. Rev., 2021, 50, 5517–5563,  10.1039/d0cs00973c.
22. H. Schönherr and T. Cernak, Profound methyl effects in drug discovery and a call for new C–H methylation reactions, Angew. Chem., Int. Ed., 2013, 52(47), 12256–12267,  DOI:10.1002/anie.201303207.
23. R. F. Hussain, A. M. Nouri and R. T. Oliver, A new approach for measurement of cytotoxicity using colorimetric assay, J. Immunol. Methods, 1993, 160, 89–96,  DOI:10.1016/0022-1759(93)90012-V.
24. A. Sancar, L. A. Lindsey-Boltz and K. Unsal-Kaçmaz, et al., Molecular mechanisms of mammalian DNA repair and the DNA damage check-points, Annu. Rev. Biochem., 2004, 73, 39–85,  DOI:10.1146/annurev.biochm.73.01130.07372.
25. A. Karimian, Y. Ahmadi and B. Yousefi, Multiple functions of p21 in cell cycle, apoptosis and transcriptional regulation after DNA damage, DNA Repair, 2016, 42, 63–71,  DOI:10.1016/j.dnarep.2016.04.008.
26. L. Lalier, F. Vallette and S. Manon, Bcl-2 Family Members and the Mitochondrial Import Machineries: The Roads to Death, Biomolecules, 2022, 12(2), 162,  DOI:10.3390/biom12020162.
27. E. Eskandari and C. J. Eaves, Paradoxical roles of caspase-3 in regulating cell survival, proliferation, and tumorigenesis, J. Cell Biol., 2022, 221(6) DOI:10.1083/jcb.202201159.
28. T. Hua, M. Robitaille and S. J. Roberts-Thomson, et al., The intersection between cysteine proteases, Ca²⁺ signalling and cancer cell apoptosis, Biochim. Biophys. Acta, Mol. Cell Res., 2023, 1870(7), 119532,  DOI:10.1016/j.bbamcr.2023.119532.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4nj04125a

‡ Jiang-Nan Wang and Mei-Nuo Chen contributed equally to this work.

---