Trigonometric Bundling Disulfide Unit Starship Synergizes More Effectively to Promote Cellular Uptake

A small molecule disulfide unit technology platform based on dynamic thiol exchange chemistry at the cell membrane has the potential for drug delivery. However, the alteration of the CSSC dihedral angle of the disulfide unit caused by diverse substituents directly affects the effectiveness of this technology platform as well as its own chemical stability. The highly stable open-loop relaxed type disulfide unit plays a limited role in drug delivery due to its low dihedral angle. Here, we have built a novel disulfide unit starship based on the 3,4,5-trihydroxyphenyl skeleton through trigonometric bundling. The intracellular delivery results showed that the trigonometric bundling of the disulfide unit starship effectively promoted cellular uptake without any toxicity, which is far more than 100 times more active than that of equipment with a single disulfide unit in particular. Then, the significant reduction in cell uptake capacity (73–93%) using thiol erasers proves that the trigonometric bundling of the disulfide starship is an endocytosis-independent internalization mechanism via a dynamic covalent disulfide exchange mediated by thiols on the cell surface. Furthermore, analysis of the molecular dynamics simulations demonstrated that trigonometric bundling of the disulfide starship can significantly change the membrane curvature while pushing lipid molecules in multiple directions, resulting in a significant distortion in the membrane structure and excellent membrane permeation performance. In conclusion, the starship system we built fully compensates for the inefficiency deficiencies induced by poor dihedral angles.


Introduction
Efficient and safe cellular delivery strategies currently remain one of the great challenges in chemical biology [1].In particular, the delivery of poorly soluble, highly toxic small molecular entities or natural products and, moreover, large substrates such as nucleic acids and peptide drugs by nanoparticles usually occurs through endocytosis [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16].However, cellular delivery mediated by endocytosis can easily lead to drug retention by the endosomes, which makes release into the cytoplasm more difficult [17].In addition, these delivery systems face serious challenges such as high dosage [18] difficulty in unifying the size of vector-drug complexes [19] and high cytotoxicity.Therefore, endocytosisindependent cellular uptake has attracted increasing attention because of its ability to rapidly distribute drugs in the cytoplasm [20].Currently, dynamic covalent disulfide exchange chemistry at the cell surface is emerging as an innovative strategy to address the above-mentioned challenges [21].Briefly, a drug coupled with the modification of a disulfide unit undergoes disulfide exchange with thiols on the cell surface, causing the drug to covalently adsorbs on the cell surface and be taken up by the cell.Glutathione reduction in the cytoplasm completes the release of the drug in the cytoplasm.Due to the advantages of low molecular weight, lack of charge and non-toxicity, the active disulfide units have been widely used in the field of drug delivery [22][23][24][25][26][27].
Dynamic covalent disulfide exchange on the cell surface is a crucial step in this strategy, which directly impacts drug binding and cellular uptake.Therefore, disulfide modification strategies that can effectively promote the efficiency of the covalent exchange of thiols are the focus of current development.First, Matile's groups [28][29][30] discovered a strong correlation between the CSSC dihedral angle (θ) of the disulfide unit and cellular uptake.All reported disulfide units can be divided into three types (Figure 1A): open-loop relaxed type (I, θ = 90 • ), ring-comforting type (II, III with θ of 35 • and 27 • , respectively), and ringtensing type (IIII, θ = 0 • ).Although there are slight differences in activity between different structures, smaller CSSC dihedral angles generally result in faster disulfide exchange reactions, leading to increased drug attachment to cell surfaces and improved cellular uptake within a shorter time frame [31].This is also proven by the energy profile analysis (Figure S18) of different disulfide units with different CSSC dihedral angles.However, the reduction in the dihedral angle comes with the risk of polymerization of the disulfide unit [32][33][34][35][36]. Particular care is required in peptide and nucleic acid drug synthesis processes, which often involve strong acidic, alkaline, or oxidation-reduction environments to prevent inactivation caused by polymerization.Therefore, a moderate change in the dihedral angle serves as a foundation for ensuring the stability and intracellular uptake efficiency of coupling complexes.Second, the number and tandem mode between disulfide units and drugs also directly affect the cellular uptake efficiency [37].Coupling numerous disulfide units indeed enhances intracellular uptake, but a simple and effective tandem approach needs to be further explored.modification of a disulfide unit undergoes disulfide exchange with thiols on the cell surface, causing the drug to covalently adsorbs on the cell surface and be taken up by the cell.Glutathione reduction in the cytoplasm completes the release of the drug in the cytoplasm.Due to the advantages of low molecular weight, lack of charge and non-toxicity, the active disulfide units have been widely used in the field of drug delivery [22][23][24][25][26][27].
Dynamic covalent disulfide exchange on the cell surface is a crucial step in this strategy, which directly impacts drug binding and cellular uptake.Therefore, disulfide modification strategies that can effectively promote the efficiency of the covalent exchange of thiols are the focus of current development.First, Matile's groups [28][29][30] discovered a strong correlation between the CSSC dihedral angle (θ) of the disulfide unit and cellular uptake.All reported disulfide units can be divided into three types (Figure 1A): open-loop relaxed type (I, θ = 90°), ring-comforting type (II, III with θ of 35° and 27°, respectively), and ring-tensing type (IIII, θ = 0°).Although there are slight differences in activity between different structures, smaller CSSC dihedral angles generally result in faster disulfide exchange reactions, leading to increased drug attachment to cell surfaces and improved cellular uptake within a shorter time frame [31].This is also proven by the energy profile analysis (Figure S18) of different disulfide units with different CSSC dihedral angles.However, the reduction in the dihedral angle comes with the risk of polymerization of the disulfide unit [32][33][34][35][36]. Particular care is required in peptide and nucleic acid drug synthesis processes, which often involve strong acidic, alkaline, or oxidation-reduction environments to prevent inactivation caused by polymerization.Therefore, a moderate change in the dihedral angle serves as a foundation for ensuring the stability and intracellular uptake efficiency of coupling complexes.Second, the number and tandem mode between disulfide units and drugs also directly affect the cellular uptake efficiency [37].Coupling numerous disulfide units indeed enhances intracellular uptake, but a simple and effective tandem approach needs to be further explored.
To further enhance the wide application of open-loop relaxed-type disulfide units in drug delivery, we constructed a trigonometric bundling starship based on a 3,4,5trihydroxyphenyl skeleton with the reported simple small tert-butyl disulfide unit, in which the three contained disulfide units were arranged in the same space at a unique angle (Figure 1B).It is expected to synergistically promote intracellular uptake to compensate for the low exchange rate.To further enhance the wide application of open-loop relaxed-type disulfide units in drug delivery, we constructed a trigonometric bundling starship based on a 3,4,5trihydroxyphenyl skeleton with the reported simple small tert-butyl disulfide unit, in which the three contained disulfide units were arranged in the same space at a unique angle (Figure 1B).It is expected to synergistically promote intracellular uptake to compensate for the low exchange rate.

Chemistry
The synthesis scheme for SS1/3-FITC probes is illustrated in Scheme 1. Specifically, the TsCl reagent was utilized to activate the terminal hydroxyl group of the disulfide unit in the presence of pyridine in CH 2 Cl 2 .The mono-or tri-substituted conjugation products 5a-b were then obtained in high yield using the activated intermediate 2 with mono-or trihydroxybenzoate (3 or 4, respectively) under the K 2 CO 3 condition.After the hydrolysis of methyl benzoate, the linker was extended by classical amide condensation reaction and finally conjugated with FITC under alkaline conditions to obtain fluorescent probes 8a-b (SS1-FITC and SS3-FITC, respectively).Finally, the UV absorbance value of FITC at 520 nm in sodium tetraborate buffer (pH = 9.18) was used to calculate the exact concentration of the fluorescent probe. 1 H NMR, 13 C NMR, and HRMS spectra of all compounds are provided in Figures S1-S12.

Chemistry
The synthesis scheme for SS1/3-FITC probes is illustrated in Scheme 1. Specifically, the TsCl reagent was utilized to activate the terminal hydroxyl group of the disulfide unit in the presence of pyridine in CH2Cl2.The mono-or tri-substituted conjugation products 5a-b were then obtained in high yield using the activated intermediate 2 with mono-or trihydroxybenzoate (3 or 4, respectively) under the K2CO3 condition.After the hydrolysis of methyl benzoate, the linker was extended by classical amide condensation reaction and finally conjugated with FITC under alkaline conditions to obtain fluorescent probes 8a-b (SS1-FITC and SS3-FITC, respectively).Finally, the UV absorbance value of FITC at 520 nm in sodium tetraborate buffer (pH = 9.18) was used to calculate the exact concentration of the fluorescent probe. 1 H NMR, 13 C NMR, and HRMS spectra of all compounds are provided in Figures S1-S12.

CCK-8 Assay
Firstly, the CCK-8 assay was used to assess the cytotoxicity of the disulfide unit starship and their intracellular metabolites in cells.To eliminate interference from fluorophores, the intermediate 7a' and 7b' (Scheme 1) of the starship were used for specific exercises.The results showed that both their intracellular glutathione hydrolysis products were not cytotoxic in the above commonly used cancer cells, even at 10 μM (Figure 2).Scheme 1. Synthesis of SS1/3-FITC probe (8a and 8b).

Biological Activity 2.2.1. CCK-8 Assay
Firstly, the CCK-8 assay was used to assess the cytotoxicity of the disulfide unit starship and their intracellular metabolites in cells.To eliminate interference from fluorophores, the intermediate 7a' and 7b' (Scheme 1) of the starship were used for specific exercises.The results showed that both their intracellular glutathione hydrolysis products were not cytotoxic in the above commonly used cancer cells, even at 10 µM (Figure 2).2.2.2.Time-Dependent Cellular Uptake of SS1/3-FITC Probe A series of cell experiments were performed to evaluate the bioactivity of the fluorescent probes.Firstly, the uptake of the fluorescent probes SS1-FITC and SS3-FITC into various cancer cells was monitored using confocal laser scanning microscopy (CLSM).The

Time-Dependent Cellular Uptake of SS1/3-FITC Probe
A series of cell experiments were performed to evaluate the bioactivity of the fluorescent probes.Firstly, the uptake of the fluorescent probes SS1-FITC and SS3-FITC into various cancer cells was monitored using confocal laser scanning microscopy (CLSM).The incubation of A549 cells with a 1 µM SS3-FITC probe in DMEM medium for 4 h at 37 • C resulted in intense and uniform fluorescence emission from the cytoplasm of each cell (Figure 3B), including poorly stained nuclei with DAPI.At the same time, slightly weak fluorescence emission was also detected in the nuclei of some cells with a p value of 0.61 (Figure 3D,E).In sharp contrast, the SS1-FITC probe (1 µM) equipped with a single disulfide unit failed to detect significant fluorescence emission after 4 h of incubation with A549 cells (Figure 3A), which was consistent with the low cellular uptake of open-loop relaxed type disulfide units reported previously [37].To exclude the fluorescence interference of the FITC fluorophore itself, no obvious fluorescence signal was found for the disulfide unit in which the FITC-NH 2 probe was absent under the same conditions (Figure S13).These results indicate that the trigonometric bundling disulfide unit can effectively promote intracellular uptake.Meanwhile, similar results were observed in several other cell lines (HeLa S3, MDA-MB-231, HepG2, HT29 and MCF-7, Figures S14A,E, S15 and S16, respectively).In total, we used six different tumor cell lines to evaluate the cell uptake effect of fluorescent probes.The results from CLSM showed that SS3-FITC probes exhibited strong cell delivery effects on various cell lines, indicating their potential in the field of anti-tumor drug delivery in the future.Among these cell lines, fluorescence intensity analysis revealed that the SS3-FITC probe demonstrated the highest cellular uptake ability in A549 cells (Figure S17).We hypothesize that differences in the number of thiol groups on the cell membrane or variations in expression levels of thiol-rich proteins may contribute to this discrepancy in cellular uptake efficiency.
We also investigated in detail the uptake status of the SS3-FITC probe (1 µM) in A549 cells at different incubation time points (0.5, 1.0, 2.0, 4.0 h) (Figure 3B).It was evident that the SS3-FITC probe rapidly distributed in the cell membrane and cytoplasm within 0.5 h incubation, and the intracellular probe concentration increased significantly with the prolongation of the incubation time (Figure 3C).It was found that the fluorescent signal runs in points around the cell membrane.Once inside the cell, the probe was evenly distributed throughout the cytoplasm and part of the nucleus (Figure 3D,E), which is consistent with the previously described mechanism of disulfide unit intracellular transport [31].In addition, similar results were observed in the HeLa S3 cell line (Figure S14B-E).These results indicate that the designed trigonometric bundling disulfide unit starship not only greatly promotes cellular uptake but also achieves ultrafast cytoplasmic distribution, which has strong potential for clinical application.

Concentration-Dependent Cellular Uptake of SS1/3-FITC Probe
To further visualize the comparison of intracellular uptake of the SS3-FITC and SS1-FITC probes, different concentrations of the fluorescent probes were incubated in A549 cells for 1 h and analyzed by CLSM (Figure 4).Apparently, SS3-FITC (0.1 µM) still efficiently entered the cells even at lower concentrations (Figure 4B).However, SS1-FITC (10 µM) still had no positive influence on intracellular delivery at high concentration conditions (Figure 4A).The fluorescence intensity obtained at 0.1 µM using the trigonometric bundling disulfide unit starship is still higher than that obtained at 10 µM using the single disulfide unit-equipped transporter, indicating that the trigonometric bundling disulfide unit starship is much more than 100-fold more active (Figure 4C), which was also found in Figure 3C.The nucleic acid drug constructed by the linear disulfide bond coupling method adopted by Abe's group is upgraded from 1 disulfide unit in tandem to 5 disulfide units in series and then to 10 disulfide units in series, and its activity is about 2-4 times higher [37].This fully proves the synergistic highly efficient intracellular delivery potential of the trigonometric bundling type disulfide unit starship, which fully compensates for the low reaction speed caused by the dihedral angle.
the previously described mechanism of disulfide unit intracellular transport [31].In addition, similar results were observed in the HeLa S3 cell line (Figure S14B-E).These results indicate that the designed trigonometric bundling disulfide unit starship not only greatly promotes cellular uptake but also achieves ultrafast cytoplasmic distribution, which has strong potential for clinical application.

Concentration-Dependent Cellular Uptake of SS1/3-FITC Probe
To further visualize the comparison of intracellular uptake of the SS3-FITC and SS1-FITC probes, different concentrations of the fluorescent probes were incubated in A549 cells for 1 h and analyzed by CLSM (Figure 4).Apparently, SS3-FITC (0.1 μM) still efficiently entered the cells even at lower concentrations (Figure 4B).However, SS1-FITC (10 μM) still had no positive influence on intracellular delivery at high concentration conditions (Figure 4A).The fluorescence intensity obtained at 0.1 μM using the trigonometric bundling disulfide unit starship is still higher than that obtained at 10 μM using the single disulfide unit-equipped transporter, indicating that the trigonometric bundling disulfide unit starship is much more than 100-fold more active (Figure 4C), which was also found in Figure 3C.The nucleic acid drug constructed by the linear disulfide bond coupling method adopted by Abe's group is upgraded from 1 disulfide unit in tandem to 5 disulfide units in series and then to 10 disulfide units in series, and its activity is about 2-4 times higher [37].This fully proves the synergistic highly efficient intracellular delivery

Cellular Uptake of SS3-FITC Probe with Thiols Inhibitors
To further elucidate the cellular uptake mechanism of the trigonometric bundling SS3-FITC probe, the following experiments were performed.Previous studies showed that the disulfide unit on the probe first undergoes a chemical exchange reaction with the thiol group of the proteins on the cell membrane.Therefore, the number and status of thiol groups of the proteins on the cell membrane play a key role in the intracellular delivery of the probe.In this study, cells were pretreated with several inhibitors for 0.5 h to reduce

Cellular Uptake of SS3-FITC Probe with Thiols Inhibitors
To further elucidate the cellular uptake mechanism of the trigonometric bundling SS3-FITC probe, the following experiments were performed.Previous studies showed that the disulfide unit on the probe first undergoes a chemical exchange reaction with the thiol group of the proteins on the cell membrane.Therefore, the number and status of thiol groups of the proteins on the cell membrane play a key role in the intracellular delivery of the probe.In this study, cells were pretreated with several inhibitors for 0.5 h to reduce the amount and/or state of thiol groups on the cell surface, and intracellular uptake of the probe was investigated (Figure 5A-C).Firstly, the inhibitors N-Ethylmaleimide (NEM) [38] and sodium iodoacetate (SIA) [29] were used to pre-treat the thiol groups on the surface of A549 cells, and the uptake capacity of SS3-FITC was decreased by 93% and 86%, respectively.Secondly, the uptake capacity of SS3-FITC was reduced by 73% by using DTNB to pre-treat A549 cells to convert thiol groups on the cell surface to a disulfide state.These results confirmed that SS3-FITC starship is an endocytosis-independent internalization mechanism via a dynamic covalent disulfide exchange mediated by thiols on cell surface proteins.

Cellular Uptake of SS3-FITC Probe with Thiols Inhibitors
To further elucidate the cellular uptake mechanism of the trigonometric bundling SS3-FITC probe, the following experiments were performed.Previous studies showed that the disulfide unit on the probe first undergoes a chemical exchange reaction with the thiol group of the proteins on the cell membrane.Therefore, the number and status of thiol groups of the proteins on the cell membrane play a key role in the intracellular delivery of the probe.In this study, cells were pretreated with several inhibitors for 0.5 h to reduce the amount and/or state of thiol groups on the cell surface, and intracellular uptake of the probe was investigated (Figure 5A-C).Firstly, the inhibitors N-Ethylmaleimide (NEM) [38] and sodium iodoacetate (SIA) [29] were used to pre-treat the thiol groups on the surface of A549 cells, and the uptake capacity of SS3-FITC was decreased by 93% and 86%, respectively.Secondly, the uptake capacity of SS3-FITC was reduced by 73% by using DTNB to pre-treat A549 cells to convert thiol groups on the cell surface to a disulfide state.These results confirmed that SS3-FITC starship is an endocytosis-independent internalization mechanism via a dynamic covalent disulfide exchange mediated by thiols on cell surface proteins.

Molecular Dynamics Simulations Analysis
Having recognized the potential impact of the trigonometric bundling disulfide unit starship on improving cellular uptake, we further examined the deformation of lipid bilayers induced by key intermediates 6a and 6b (Scheme 1), which did not present interference from fluorophores.Based on the 500 ns long classical molecular dynamics (MD) simulation trajectories, it was evident that 6a and 6b exhibited different cell membrane distortion behaviors.The local membrane thickness maps revealed that 6a slightly impacted the membrane thickness, while 6b could significantly change the membrane curvature (Figure 6A).In addition, our simulations demonstrated that both 6a and 6b were inserted into the cell membrane via disulfide units.From the viewpoint of structure, the linear structure of 6a is similar to that of lipid molecules.The insertion of 6a would not drastically impact the alignment of the lipid molecules.Meanwhile, 6b could simultaneously squeeze lipid molecules in multiple directions, resulting in significant distortion of the membrane structure (Figure 6B).Moreover, the collapse of the lipid bilayer allowed 6b to penetrate deeper into the membrane.The orientation of the fatty acid chains of lipids around the small molecules can be monitored using the lipid deuterium order parameter (S cd ) [39].The S cd value of 0 indicates a completely random alignment, while a value of 1 indicates alignment perpendicular to the membrane plane [40].As shown in Figure 6C, two lipid tails tended to have more random orientations in the presence of 6b, meaning that 6b had a stronger disruption effects on the lipid bilayer.Note that disrupting the steady state of the lipid bilayer is a prerequisite for small molecule transport through the membrane; we believe that 6b had better membrane permeation performance than 6a [41].
the membrane structure (Figure 6B).Moreover, the collapse of the lipid bilayer allowed 6b to penetrate deeper into the membrane.The orientation of the fatty acid chains of lipids around the small molecules can be monitored using the lipid deuterium order parameter (Scd) [39].The Scd value of 0 indicates a completely random alignment, while a value of 1 indicates alignment perpendicular to the membrane plane [40].As shown in Figure 6C, two lipid tails tended to have more random orientations in the presence of 6b, meaning that 6b had a stronger disruption effects on the lipid bilayer.Note that disrupting the steady state of the lipid bilayer is a prerequisite for small molecule transport through the membrane; we believe that 6b had better membrane permeation performance than 6a [41].were shown in the bottom panel (the head groups and tail groups of lipid molecules were colored in sea blue and white; the carbon, oxygen, nitrogen, and sulfur atoms in 6a and 6b were colored in green, red, blue, and yellow, respectively).(C) The lipid order parameters of (Scd) of the palmitic chain (up) and oleic chain (down) in lipid molecules.The carbon index started from the carboxyl group of the fatty acid chain.The legend "control" indicates that nothing inserts in membrane.

General Method
All chemicals were purchased from commercial suppliers.The 1 H and 13 C NMR spectra of synthesized compounds were characterized on a 400 MHz spectrometer (Bruker, Billerica, MA, USA), and high-resolution mass spectra (HRMS) were recorded on a mass spectrometer (Waters, Milford, MA, USA).The fluorescence images of cultured cancer were shown in the bottom panel (the head groups and tail groups of lipid molecules were colored in sea blue and white; the carbon, oxygen, nitrogen, and sulfur atoms in 6a and 6b were colored in green, red, blue, and yellow, respectively).(C) The lipid order parameters of (S cd ) of the palmitic chain (up) and oleic chain (down) in lipid molecules.The carbon index started from the carboxyl group of the fatty acid chain.The legend "control" indicates that nothing inserts in membrane.

General Method
All chemicals were purchased from commercial suppliers.The 1 H and 13 C NMR spectra of synthesized compounds were characterized on a 400 MHz spectrometer (Bruker, Billerica, MA, USA), and high-resolution mass spectra (HRMS) were recorded on a mass spectrometer (Waters, Milford, MA, USA).The fluorescence images of cultured cancer cells were captured by confocal laser scanning microscope (CLSM, Leica, Wetzlar, Germany, TCS SP8) and analyzed by ImageJ (https://imagej.net/software/imagej/(accessed on 5 July 2024)).The CCK-8 assay was performed on a multifunctional microplate reader (Synergy H1, Bio-Tek, Winooski, VT, USA).

Concentration-Dependent Fluorescence Microscopy Measurement of the Cancer Cells Treated with the SS1-FITC and SS3-FITC Probe
The A549 cells (1 × 10 4 cells) were seeded onto 35 mm glass-bottom culture dishes (Biosharp) and grown in the complete medium (DMEM, high glucose, 10% FBS, 1% antibiotics) 24 h before probe treatment.The cells were treated with SS1-FITC (1 and 10 µM) and SS3-FITC (0.1 and 1 µM) probes, respectively, in DMEM at 37 • C for 1.0 h.After that, the cells were washed with PBS (1 mL × 3) and fixed with 4% PFA for 15 min.All cells were stained with DAPI for 15 min followed by PBS wash (1 mL × 3).The cells were immediately analyzed using a fluorescence microscope (Leica TCS SP8).

Fluorescence Microscopy Measurement of the Cancer Cells Pretreated with Thiols Inhibitors
The A549 cells (1 × 10 4 cells) were seeded onto 35 mm glass-bottom culture dishes (Biosharp) and grown in the complete medium (DMEM, high glucose, 10% FBS, 1% antibiotics) 24 h before probe treatment.The cells were pretreated with N-ethylmaleimide (1.2 mM), sodium iodoacetate (1.2 mM), and DTNB (1.2 mM), respectively, in DMEM at 37 • C for 0.5 h.After that, the cells were then treated with SS3-FITC (1 µM) probes in DMEM at 37 • C for 1 h.The cells were washed with PBS (1 mL × 3) and fixed with 4% PFA for 15 min.All cells were stained with DAPI for 15 min followed by PBS wash (1 mL × 3).The cells were immediately analyzed using a fluorescence microscope (Leica TCS SP8).

Molecular Dynamics Simulations
The force field parameters (Lennard-Jones parameters and charge parameters) of 6a and 6b were generated using the Antechamber protocol.The lipid bilayer model, composed of 200 palmitoyl-oleoyl-phosphocholine (POPC) molecules (100 per leaflet), was constructed using the CHARMM-GUI (https://charmm-gui.org/ (accessed on 5 July 2024)) Membrane Builder.The TIP3P water model was added to hydrate the lipid bilayer, and all MD systems were neutralized by a concentration of 0.15 M NaCl.For each simulation box, 6a or 6b was initially immersed in the bulk solution and 1.5 nm higher than the membrane plane.
All-atom MD simulations were performed in the NPγT ensemble using the AMBER18 software with the GPU acceleration.The Lipid14 force field and Generalized Amber Force Field (GAFF) were used for the POPC and 6a/6b, respectively.A cutoff value of 10 Å was used to treated the nonbonded interactions.The Particle Mesh Ewald (PME) algorithm was used to compute long-range electrostatic interactions.Pressure coupling was maintained at 1 bar using semi-isotropic coupling with the Berendsen barostat.The temperature was controlled by the Nose-Hoover algorithm.The constant surface tension is used in statistical ensembles for the simulations of the aqueous/membrane interface (xy plane).The surface tension value is 40 dyne/cm.All simulations started with a 5000-step energy minimization using a hybrid steepest descent/conjugated gradient algorithm.After minimization, each system was gradually heated from 0 to 310 K in 2 ns.Before pro-duction runs, all systems were equilibrated for 10 ns.Finally, 500 ns long productive MD simulations were carried out for data collection.The time step for the all-atom MD simulation was 2 fs.
Lipid order parameters.The calculation of lipid order parameters is theoretically analogous to the nuclear magnetic resonance spectroscopy of deuterated lipid chains.In MD simulations, the order parameters can be computed by the following equation: where θ is the angle formed between the carbon-hydrogen bond vector and the z-axis (normal to the monolayer plane).S CD can range from −0.5 to 1. S CD tending to 1 indicates the tail is perfectly perpendicular to the monolayer plane, whereas S CD tending to 0 suggests complete randomness.In the limit value of −0.5, tails are oriented parallel to the plane.

Conclusions
In conclusion, here we developed a novel conjugation method to enhance the drug delivery effect for the disulfide units based on dynamic covalent disulfide exchange chemistry, especially for those that are less reactive, such as open-loop relaxed type disulfide units.In the case of the tert-butyl substituted disulfide unit, we synthesized the trigonometric bundling disulfide unit starship based on the 3,4,5-trihydroxyphenyl skeleton.Cell imaging experiments showed that this trigonometric bundling disulfide unit starship synergizes more effectively to promote cellular uptake more quickly, and it completely compensates for the disadvantage regarding the low reactivity of the tert-butyl substituted disulfide unit.Furthermore, studies on thiol inhibition of the cell membrane revealed that this starship is also an endocytosis-independent internalization mechanism.Therefore, this disulfide unit starship has the potential to be an effective drug delivery agent.Research is currently underway into the delivery of macromolecular nucleic acids.

Figure 1 .Figure 1 .
Figure 1.(A) Brief description of disulfide unit performance between structure and corresponding dihedral angle.(B) The trigonometric bundling disulfide unit starship strategy used.

Figure 3 .
Figure 3. CLSM images of A549 cells after incubation with 1 μM SS1/SS3-FITC probe together with DAPI to stain the nuclei for (A,B); Quantitative analysis of fluorescence intensity and colocalization analysis are shown in (C,D), respectively.Scatterplot analysis of the merge image in the last view with the white arrow symbol is shown in (E).Scar bar: 10 μm.

Figure 3 .
Figure 3. CLSM images of A549 cells after incubation with 1 µM SS1/SS3-FITC probe together with DAPI to stain the nuclei for (A,B); Quantitative analysis of fluorescence intensity and colocalization analysis are shown in (C,D), respectively.Scatterplot analysis of the merge image in the last view with the white arrow symbol is shown in (E).Scar bar: 10 µm.Int.J. Mol.Sci.2024, 25, x FOR PEER REVIEW 7 of 16

Figure 4 .
Figure 4. CLSM images of A549 cells after 1 h incubation with different concentrations of SS1/SS3-FITC probe together with DAPI to stain the nuclei for (A,B).Quantitative analysis of fluorescence intensity is shown in (C).Scar bar: 10 μm.

Figure 4 .
Figure 4. CLSM images of A549 cells after 1 h incubation with different concentrations of SS1/SS3-FITC probe together with DAPI to stain the nuclei for (A,B).Quantitative analysis of fluorescence intensity is shown in (C).Scar bar: 10 µm.

Figure 4 .
Figure 4. CLSM images of A549 cells after 1 h incubation with different concentrations of SS1/SS3-FITC probe together with DAPI to stain the nuclei for (A,B).Quantitative analysis of fluorescence intensity is shown in (C).Scar bar: 10 μm.

Figure 6 .
Figure 6.(A) The membrane distortion induced by 6a and 6b.The local membrane thickness maps were shown in the top panel (color bar unit: nm).(B) The molecular details sounded by 6a and 6bwere shown in the bottom panel (the head groups and tail groups of lipid molecules were colored in sea blue and white; the carbon, oxygen, nitrogen, and sulfur atoms in 6a and 6b were colored in green, red, blue, and yellow, respectively).(C) The lipid order parameters of (Scd) of the palmitic chain (up) and oleic chain (down) in lipid molecules.The carbon index started from the carboxyl group of the fatty acid chain.The legend "control" indicates that nothing inserts in membrane.

Figure 6 .
Figure 6.(A) The membrane distortion induced by 6a and 6b.The local membrane thickness maps were shown in the top panel (color bar unit: nm).(B) The molecular details sounded by 6a and 6bwere shown in the bottom panel (the head groups and tail groups of lipid molecules were colored in sea blue and white; the carbon, oxygen, nitrogen, and sulfur atoms in 6a and 6b were colored in green, red, blue, and yellow, respectively).(C) The lipid order parameters of (S cd ) of the palmitic chain (up) and oleic chain (down) in lipid molecules.The carbon index started from the carboxyl group of the fatty acid chain.The legend "control" indicates that nothing inserts in membrane.