Mechanistic and Kinetic Insights into Cellular Uptake of Biomimetic Dinitrosyl Iron Complexes and Intracellular Delivery of NO for Activation of Cytoprotective HO-1

Dinitrosyl iron unit (DNIU), [Fe(NO)2], is a natural metallocofactor for biological storage, delivery, and metabolism of nitric oxide (NO). In the attempt to gain a biomimetic insight into the natural DNIU under biological system, in this study, synthetic dinitrosyl iron complexes (DNICs) [(NO)2Fe(μ-SCH2CH2COOH)2Fe(NO)2] (DNIC–COOH) and [(NO)2Fe(μ-SCH2CH2COOCH3)2Fe(NO)2] (DNIC–COOMe) were employed to investigate the structure–reactivity relationship of mechanism and kinetics for cellular uptake of DNICs, intracellular delivery of NO, and activation of cytoprotective heme oxygenase (HO)-1. After rapid cellular uptake of dinuclear DNIC–COOMe through a thiol-mediated pathway (tmax = 0.5 h), intracellular assembly of mononuclear DNIC [(NO)2Fe(SR)(SCys)]n−/[(NO)2Fe(SR)(SCys-protein)]n− occurred, followed by O2-induced release of free NO (tmax = 1–2 h) or direct transfer of NO to soluble guanylate cyclase, which triggered the downstream HO-1. In contrast, steady kinetics for cellular uptake of DNIC–COOH via endocytosis (tmax = 2–8 h) and for intracellular release of NO (tmax = 4–6 h) reflected on the elevated activation of cytoprotective HO-1 (∼50–150-fold change at t = 3–10 h) and on the improved survival of DNIC–COOH-primed mesenchymal stem cell (MSC)/human corneal endothelial cell (HCEC) under stressed conditions. Consequently, this study unravels the bridging thiolate ligands in dinuclear DNIC–COOH/DNIC–COOMe as a switch to control the mechanism, kinetics, and efficacy for cellular uptake of DNICs, intracellular delivery of NO, and activation of cytoprotective HO-1, which poses an implication on enhanced survival of postengrafted MSC for advancing the MSC-based regenerative medicine.


■ INTRODUCTION
Mesenchymal stem cell (MSC)-based transplantation therapies have been widely investigated for treating a variety of diseases/ injuries that remain incurable nowadays. 1−6 The harsh microenvironment characterized by (a) inadequate oxygen and nutrient supply, 2,3 (b) excessive proinflammatory cytokines, 7,8 or (c) elevated oxidative stress significantly contributes to death of transplanted MSCs, 9,10 thus compromising the ultimate therapeutic efficacy.In addition to MSC-based therapies, corneal transplantation is another cellbased therapy for sight-restoring in patients with corneal blindness. 11−22 Nevertheless, the safety issue remains a major concern for the clinical application of genetically manipulated cells, 24,34 while the use of cobalt protoporphyrin, a potent inducer of HO-1, also raises concerns regarding metalloporphyrin-induced cytotoxicity or side effects. 35Therefore, strategies that can effectively upregulate HO-1 expression without compromising safety remain to be warranted to potentiate postengrafted MSC viability and therapeutic efficacy.
−59 In addition to the potential biomedical applications, 51 systematic reviews on bioinorganic and coordination chemistry, 52,60,61 synthetic methodology, 62,63 and electronic structure study of biomimetic DNICs were also reported. 64As opposed to the other reported NO-delivery reagents, DNICs [(NO) 2 Fe(μ-SR) 2 Fe(NO) 2 ] feature a unique O 2 -induced mechanism for the release of NO, of which the kinetics are modulated by the side chain of bridging thiolate ligands. 45,47,48,50,55,65 enables the utilization of these endogenous protein vehicles for oral delivery of DNIC/NO into brain. 50nspired by the potential of NO on potentiating MSC/corneal endothelial cell (CEC)-based transplantation therapy, herein, synthetic DNICs [(NO) 2 Fe(μ-SCH 2 CH 2 COOH) 2 Fe(NO) 2 ] (DNIC−COOH) and [(NO) 2 Fe(μ-SCH 2 CH 2 COOCH 3 ) 2 Fe-(NO) 2 ] (DNIC−COOMe) were employed to investigate the structure−reactivity relationship for kinetic and mechanistic control on cellular uptake of DNICs and intracellular delivery of NO (Schemes 1 and 2).As opposed to the slow binding of    As shown in Figure 1a and Scheme 1a, DNIC−COOH and DNIC−COOMe displayed a NO-release reactivity under normoxia condition, which is similar to the reported O 2 -induced release of NO from other dinuclear DNICs [(NO) 2 Fe(μ-SR) 2 Fe(NO) 2 ]. 45,47,48,50,55,65 Based on the pseudo-first-order kinetic model, the t 1/2 (or rate constant) for these processes  OH/ • NO 2 species. 66As opposed to the generation of peroxynitrite and cytotoxic • OH/ • NO 2 during the interaction between O 2 − and NONOates (widely used chemical NO donors), 67 this study unveiled the Inspired by the reported interactions between dinuclear DNICs [(NO) 2 Fe(μ-SR) 2 Fe(NO) 2 ] and R′SH/[SR′] − leading to the assembly of EPR-active mononuclear {Fe(NO) 2 } 9 DNICs [(NO) 2 Fe(SR)(SR′)] − , 50,55,68,69 reactions of DNIC− COOH/DNIC−COOMe with biological thiols, i.e., L-cysteine (Cys) or serum albumin (BSA), were further explored.As shown in Figure 2a,b, treatment of 10 mM of Cys to DNIC− COOH (or DNIC−COOMe) resulted in the formation of a distinctive EPR signal at g = 2.041, 2.035, and 2.015 (or g = 2.041, 2.035, and 2.014, Tables 2 and S2).Similar to the reactions between DNICs and Cys reported in the literature,  condition. 55,65,70Recently, dinuclear DNIC−COOH was reported to be inert toward the Fe 3+ -porphyrin center in metMb, which is in contrast to efficient nitrosylation of the Fe 2+porphyrin center in deoxyMb by dinuclear DNIC−COOH leading to generation of MbNO (Scheme 1d). 53As shown in Figure 3a 2 and S2).Similar to the interactions between DNICs and Cys-containing proteins reported in the literature, these formations of distinctive EPR signals supported the assembly of albumin-bound DNIC [(NO) 2 Fe(SR)(S Cys-albumin )] n− (Schemes 1f and 2a).As opposed to DNIC−COOH, DNIC−COOMe exhibits a rapid binding to BSA, yielding albumin-bound DNIC (Figure S5a−c).Through direct comparisons of the apparent intensity of these EPR features (Figure S3), moreover, elevated formation of albumin-bound DNIC along with the increased %FBS demonstrated a reversible interaction between dinuclear DNIC−COOH/DNIC−COOMe and BSA leading to the generation of albumin-bound DNIC.As shown in Figure S4, pretreatment of N-ethylmaleimide (NEM, a thiol-blocking reagent) to FBS followed by addition of dinuclear DNIC− COOH/DNIC−COOMe into cell culturing media with NEMtreated FBS resulted in the disappearance of distinctive EPR signals. 50,72That is, the Cys-34 (the only thiol in BSA) may serve as the anchoring site for the assembly of albumin-bound DNIC similar to that derived from the reaction of dinuclear DNIC [(NO) 2 Fe(μ-SCH 2 CH 2 OH) 2 Fe(NO) 2 ] and BSA. 50On the other hand, assembly of albumin-bound DNIC [(NO) 2 Fe-(SCH 2 CH 2 COOH)(S Cys-albumin )] − via the reaction between 50 mg/mL of BSA and DNIC−COOH was also evidenced by the shift of IR ν NO absorption peaks from (1781, 1756) cm −1 to (1777, 1747) cm −1 (Figure S3g).Based on the investigations discussed above, the hydrophobic nature of DNIC−COOMe (Log D7.4 = 1.0), in contrast to the hydrophilic nature of deprotonated DNIC−COOH at a neutral pH environment (Log D7.4 = −2.3),may explain its strong and rapid binding to Cys-34 in the hydrophobic pocket of BSA. 73,74−78 Regarding the assembly of albumin-bound DNIC derived from the reaction of dinuclear DNIC−COOH/DNIC− COOMe with BSA, potential NO-delivery reactivity of albumin-bound DNIC was further explored.As shown in Figure S5d−  Regarding the explored binding affinity of dinuclear DNIC− COOMe/DNIC−COOH to Cys and protein-derived Cys residue(s), a thiol-mediated pathway is one of the potential mechanisms for cellular uptake of DNICs.−81 Although the potential cellular uptake of DNIC−COOMe via a passive diffusion process, such a pathway is excluded for DNIC− COOH, considering its dianionic nature under the cell culturing media with a neutral pH environment.−84 Before the whole-cell EPR study, a cell viability study of MSC/N2a/HCEC against different inhibitors was performed to determine the optimal conditions (Figure S9).
During the investigations on DNIC−COOH under a similar methodology, no change of whole-cell EPR signal intensity upon pretreatment of NEM and a slight increase of whole-cell EPR signal intensity using NEM-treated FBS excluded the cellular uptake of DNIC−COOH through the thiol-/albumin-mediated pathway (Figure S10).Upon sequential treatments of  After treatment of DNIC−COOMe to MSC/N2a/HCEC under hypoxia condition, formation and decomposition of mononuclear DNIC [(NO) 2 Fe(SR)(S Cys )] n− /[(NO) 2 Fe(SR)-(S Cys-protein )] n− were also monitored using EPR spectroscopy.As shown in Figure 4c−h, besides the similar t max = 0.5 h for intracellular assembly of mononuclear DNICs, intracellular decompositions of mononuclear DNICs followed t 1/2 = 1.8 ± 0.1 h in MSC, t 1/2 = 2.1 ± 0.4 h in N2a, and t 1/2 = 2.9 ± 0.2 h in HCEC (Table 1).As discussed above, both mononuclear DNICs [(NO) 2 Fe(SR)(S Cys )] n− and [(NO) 2 Fe(SR)-(S Cys-albumin )] n− displayed an enhanced stability under anaerobic and hypoxia conditions.Consequently, the comparable kinetics for intracellular decompositions of mononuclear DNICs in MSC/N2a/HCEC under normoxia and hypoxia conditions implicated an O 2 -independent process for transfer of NO from mononuclear DNICs to Fe 2+ -porphyrin protein(s) (Scheme 2g).In the attempt to explore this NO-transfer process, confocal microscopic study on activation of cGMP as well as real-time quantitative polymerase chain reaction (qPCR)/enzyme-linked immunosorbent assay (ELISA) analyses on transcriptional/ translational regulations of HMOX1 genes/heme oxygenase (HO)-1 proteins in MSC/N2a/HCEC treated with DNIC− COOMe/DNIC−COOH were further performed.

Steady Delivery of Intracellular NO and Elevated Activation of Cytoprotective HO-1 by DNIC−COOH Enhance the Survival of MSC/HCEC under Stressed Conditions
We then investigated whether the DNIC-induced upregulation of HO-1 was beneficial for the survival of MSCs under stressed conditions.Herein, two in vitro models were employed to simulate the hostile microenvironment in the recipient tissues confronted by the MSCs upon transplantation.In the first scenario, hydrogen peroxide (H 2 O 2 ) was added to the medium to mimic oxidative stress in inflammatory tissues.MSCs were pretreated with DNIC−COOMe or DNIC−COOH for 8 h before exposure to H 2 O 2 , while untreated cells and cells exposed to H 2 O 2 were utilized as controls.As revealed by the results of live/dead staining and CCK-8 assay (Figure 8a,c), significant cell death was observed after H 2 O 2 treatment, suggesting that the elevated oxidative stress in inflamed tissue may prevent successful engraftment of transplanted cells.In contrast, the viability of MSCs could be significantly improved by pretreatment with DNICs (p < 0.005).Furthermore, MSC that received DNIC−COOH exhibited a higher survival rate than that of the DNIC−COOMe group (58.7% vs 35.8%, p < 0.001), indicating the superior cytoprotective potential of DNIC−COOH treatment.As shown in Figure 8e, of interest, the DNIC−COOHprimed HCEC also displayed an enhanced survival under the mimic oxidative stress condition, which is as opposed to that without or with the pretreatment of DNIC−COOMe.
In addition to inflammation-induced oxidative stress, the transplanted cells may also face ischemia-induced hypoxia and limited access to nutrients.To examine the capacity of DNIC in promoting cell survival under such harsh circumstances, cells that were pretreated with DNIC−COOMe or DNIC−COOH were cultured in an oxygen−glucose-deprived (OGD) condition.Compared with the cells grown under a normal condition, deprivation of oxygen and glucose significantly suppressed cell proliferation and induced cell death (Figure 8b,d), indicating the threat of the ischemic microenvironment to the transplanted cells.The viability of the cells that received DNIC−COOMe was comparable to that of OGD group (p > 0.05), suggesting that DNIC−COOMe preconditioning might not achieve cytoprotective effects on cells under ischemic conditions.For the cells that had been pretreated with DNIC− COOH, however, an increase in the number of viable cells was observed, indicating the potential of DNIC−COOH in enhancing cell survival in an ischemic microenvironment.Based on the above observations, DNIC−COOH exhibited a higher efficiency in upregulating HO-1 expression and a better capacity at improving cell viability under hostile conditions and was therefore employed for the following investigations.

Conclusions and Comments
−94 In comparison with NONOates, synthetic DNICs were explored to be more effective in triggering apoptosis of Jurkat cells and enhancing proliferation/migration/tube formation of EA.hy926 human vascular endothelial cells, 45,95 which may be ascribed to the different NO-delivery mechanisms.In terms of mechanisms for delivery of NO to the intracellular NO-responsive targets (i.e., Fe-porphyrin center in sGC), a variety of NONOates feature a pH-dependent nature for release of free NO in the extracellular environment through hydrolysis.Subsequent passive diffusion of freely released NO from extracellular environment to intracellular compartment occurs followed by interaction of delivered NO with Fe 2+porphyrin center in sGC and regulation of the well-known NO-sGC-cGMP pathway.−89 In contrast, (a) superoxide-scavenger nature, (b) • OH/ • NO 2 -free NO-delivery reactivity, (c) modulated mech-anism/kinetic/efficacy for cellular uptake, and (d) direct NO/ NO − -transfer activity (i.e., toward the Fe 2+ /Fe 3+ -porphyrin center of sGC) of dinuclear DNICs [(NO) 2 Fe(μ-SR) 2 Fe-(NO) 2 ] highlight the potential of synthetic DNICs, as a novel type of NO prodrugs, for development of (MSC-based) regenerative medicine and cell therapy.In addition to DNICinduced upregulation of cytoprotective HO-1 in MSC explored in this study, potential activation of neurogenesis, 50 antiinflammation effect, and blood−brain barrier permeability 44 by NO-delivery DNICs prompts our undergoing study on the combination of DNICs and MSC for the treatment of ischemic stroke.

Instruments
All of the EPR measurements were performed at X-band using a Bruker EMXmicro-6/1/S/L spectrometer equipped with a Bruker E4119001 super high sensitivity cavity.X-band EPR spectra were obtained with a microwave power of 0.6456−0.6348mW, frequency at 9.41 GHz, conversion time of 66.68 ms, receiver gain of 30, and modulation amplitude of 10.0 G at 100 kHz.UV−vis spectra were recorded on a PerkinElmer Lambda 365 spectrometer.Fourier transform infrared (FT-IR) spectra were recorded using a sealed solution cell (0.1 mm, CaF 2 windows).Reactions of dinuclear DNIC−COOH/DNIC− COOMe with O 2 in the presence of DTBP were characterized using Trace 1300 Gas Chromatograph in combination with a mass spectrometer with a 5MS column.The confocal microscopic images were recorded using ZEISS LSM 780 or Leica TCS−SP5-X AOBS confocal microscope systems.The absorbance of the assay was recorded using a microplate reader SpectraMax iD3, Molecular Devices, San Jose, CA.

NO-Release Reactivity of DNICs
The NO-release reactivity of DNICs in 50 mM potassium phosphate buffer (pH 7.4) under normoxia condition at 37 °C was investigated using Nitrate/Nitrite Colorimetric Assay Kit (Item No. 780001, Cayman).After addition of 10 μL of 100 mM stock solution of DNIC− COOH (or DNIC−COOMe) in DMSO to 990 μL of 50 mM potassium phosphate buffer (pH 7.4), 25 μM of DNIC−COOH (or DNIC−COOMe) was prepared by the addition of 0.125 mL of this 1 mM solution of DNIC−COOH (or DNIC−COOMe) to 4.875 mL of 50 mM potassium phosphate buffer (pH 7.4).This solution was incubated at 37 °C under normoxia condition for 0, 1, 2, 4, 8, 24, 48, and 72 h, respectively.Then, 40 μL of the solution was collected to assess the release of NO using a Nitrate/Nitrite Colorimetric Assay Kit (Item No. 780001, Cayman).After each 40 μL aliquot of the aqueous solution was mixed with 40 μL of kit assay buffer, 10 μL of Enzyme Cofactor Mixture (Item No. 780012) and 10 μL of nitrate reductase mixture (Item No. 780010) were added before this mixture solution was incubated at room temperature for 1 h.Subsequently, addition of 50 μL of Griess Reagent R1 (Item No. 780018) and 50 μL of Griess Reagent R2 (Item No. 780020) by incubation at room temperature for 15 min results in the formation of UV−vis absorption band at 540 nm.The absorbance at 540 nm was then recorded using a microplate reader (SpectraMax iD3, Molecular Devices, San Jose, CA) with a reference wavelength of 800 nm.According to a calibration curve made with 0, 5, 10, 15, 20, 25, 30, and 35 μM of nitrite standard (Item No. 780016)/ nitrate standard (Item No. 780014), respectively, NO-release reactivity of DNIC−COOH (or DNIC−COOMe) in PBS (pH 7.4) at each time point was further estimated.Assuming that the aerobic degradation of DNIC−COOH (or DNIC−COOMe) follows pseudo-first order kinetics, the half-life for release of NO from DNIC−COOH (or DNIC−COOMe) in 50 mM potassium phosphate buffer (pH 7.4) at 37 °C was determined.Three independent experiments were conducted to measure the average half-life for the release of NO from DNIC−COOH (or DNIC−COOMe).
Under normoxia condition, the NO-release reactivity of 25 μM DNIC−COOH (or DNIC−COOMe) (a) with the presence of 200 μM of Cys in 50 mM potassium phosphate buffer (pH 7.4), (b) in αminimum essential medium (αMEM, Thermo Fisher Scientific, Waltham, MA) without/with the presence of 20% fetal bovine serum (FBS, Gibco), (c) in minimum essential medium (MEM, Hyclone) without/with the presence of 5% FBS, and (d) in human endothelial serum-free medium (HSFM, Gibco) without/with the presence of 2% FBS was evaluated using Nitrate/Nitrite Colorimetric Assay Kit (Item No. 780001, Cayman) under a similar procedure described above.

Reaction of DNIC−COOH (or DNIC−COOMe) with O 2(g) in the Presence of 2,4-Di-tert-butyl Phenol (DTBP)
To a 25 mL Schlenk tube loaded with 5 mL of THF mixture solution of DNIC−COOH (4.4 mg, 0.01 mmol) and DTBP (8.3 mg, 0.04 mmol), 2.4 mL of O 2(g) (0.1 mmol) was added via a gastight syringe at ambient temperature.After this THF mixture solution was stirred overnight, it was used for gas chromatography−mass spectrometry (GC−MS) analysis.The reaction of DNIC−COOMe and 10 equiv of O 2(g) in the presence of 4 equiv of DTBP followed by GC-MS analysis was performed under a similar procedure.

Reactions of DNICs with Cys in PBS (pH 7.4) under Normoxia or Anaerobic Condition
50 μM of DNIC−COOH was prepared via addition of 1 μL of 100 mM stock solution of DNIC−COOH in DMSO to 2 mL of PBS (pH 7.4) with the presence of 10 mM Cys.After this solution was incubated at 37 °C for 0, 1, 2, and 4 h, respectively, EPR spectra of 100 μL aliquots of this solution were measured.Moreover, double integration of the EPR signal was executed to obtain the integrated EPR signal intensity under each condition, which was further utilized to characterize the timedependent formation and degradation of Cys-bound mononuclear DNIC [(NO) 2 Fe(SCH 2 CH 2 COOH)(S Cys )] n− .EPR investigations on the formation and degradation of the Cys-bound DNICs upon incubation of 50 μM of DNIC−COOMe in PBS (pH 7.4) with the presence of 10 mM of Cys were performed in a similar manner.
Reactions of DNIC−COOH/DNIC−COOMe with Cys under anaerobic condition were performed under a similar procedure, while all of the samples were prepared under an anaerobic N 2(g) atmosphere, and the PBS buffer (pH 7.4) was degassed before use.

Reactions of DNICs with metMb/deoxyMb without or with the Presence of Cys (or BSA)
After addition of 30 μL of 500 μM stock solution of DNIC−COOH (in DMSO) to a 4 mL quartz cuvette containing 5 μM metMb and 200 μM Cys in 3 mL of 25 mM PBS (pH 7.4), UV−vis spectra for this mixture solution were measured every 30 min.Transformation of deoxyMb into MbNO was achieved after reaction for 4 h based on the complete shift of UV−vis absorption bands from (410, 502, and 629) nm to (419, 548, and 578) nm.
Reactions of (a

EPR Investigations on Reactions of DNICs under Alternative Cell Culturing Media without/with the Presence of FBS
Under Normoxia Condition.50 μM of DNIC−COOH was prepared via addition of 1 μL of 100 mM stock solution of DNIC− COOH in DMSO to 2 mL of αMEM with the presence of 20% FBS.After this solution was incubated at 37 °C for 0, 0.5, 1, 2, 3, 4, 8, and 24 h, respectively, EPR spectra of 100 μL aliquots of this solution were measured.Moreover, double integration of the EPR signal was executed to obtain the integrated EPR signal intensity under each condition, which was further utilized to characterize the time-dependent formation and degradation of BSA-bound DNIC.EPR investigations on the formation and degradation of protein-bound DNICs upon incubation of ( 1 Under Hypoxia Condition.EPR investigations on the formation and degradation of albumin-bound DNICs upon incubation under hypoxia condition were performed following the procedure described below.αMEM with the presence of 20% FBS were incubated under a hypoxia condition (1% oxygen and 5% CO 2 atmosphere in a hypoxia incubator, Heracell VIOS 160i, Thermo Fisher Scientific) overnight.On the next day, 50 μM DNIC−COOH was prepared via addition of 1 μL of 100 mM stock solution of DNIC−COOH in DMSO to 2 mL of the preconditioned αMEM with the presence of 20% FBS.After incubation of this solution at 37 °C under hypoxia condition for 8 and 24 h, respectively, EPR spectra of 100 μL aliquots were measured.In a similar manner, EPR investigations on ( 1 A O is the A 362 in n-octanol, and A PBS is the A 362 in PBS (pH 7.4) after partitioning.Log D7.4 for DNIC−COOMe was determined in a similar manner.On the other hand, the Log P value for DNIC−COOH was measured based on its distribution in ddH 2 O and n-octanol using a similar procedure. 97,98ll Culture Human umbilical cord blood mesenchymal stem cells (MSC) were purchased from the Bioresource Collection and Research Center, Food Industry Research and Development Institute (Hsinchu, Taiwan) and maintained in αMEM (Thermo Fisher Scientific, Waltham, MA USA) supplemented with 20% FBS (Gibco), 30 mg/mL hygromycin B (Thermo Fisher Scientific), and 4 ng/mL basic fibroblast growth factor (bFGF; Thermo Fisher Scientific).Mouse neuroblast cells (N2a) were cultured in MEM (Hyclone) supplemented with 5% FBS, 1% sodium pyruvate (Gibco), and 1% penicillin/streptomycin (Invitrogen).Human corneal endothelial cells (HCEC) were maintained in human endothelial serum-free medium (HSFM; Gibco) supplemented with 2% FBS and 10 ng/mL bFGF.The cells were cultured at 37 °C with 5% CO 2 in a humidified incubator.

Cell Viability
Cell viability study of MSC, N2a, and HCEC cells with or without the treatment of DNIC−COOH was performed using a Cell Counting Kit-8 (IMT Formosa New Materials Co., Ltd., Taiwan) according to the manufacturer's protocol.Briefly, cells were seeded into a 96-well plate at a density of 4 × 10 3 cells/well for MSC, 1 × 10 4 cells/well for N2a, and 8 × 10 3 cells/well for HCEC, respectively, and incubated overnight.On the next day, cells were treated with DNIC−COOH at different concentrations and incubated for 24 h.The culture medium was then removed before the cells were washed with DPBS.After addition of fresh medium containing CCK-8 solution to the cells and incubation for 1−3 h, the absorbance at 450 nm was then measured on a microplate reader (SpectraMax iD3, Molecular Devices, San Jose, CA) with a reference wavelength of 650 nm.Three independent experiments were executed, and the results were represented as mean ± SEM% (n = 3) with the untreated cells (control) as 100% viability.Cell viability studies of MSC/N2a/HCEC cells (1) with the treatment of DNIC−COOMe for 24 h, (2) with the treatment of NEM for 0.5 h, (3) with the treatment of chlorpromazine (CPZ; TCI) for 0.5 h, (4) with the treatment of methyl-β-cyclodextrin (MβCD; Thermo Fisher Scientific) for 0.5 h, (5) with the treatment of genistein (Alfa Chemistry) for 0.5 h, (6) with the sequential treatments of ODQ for 1 h and DNIC−COOH for 3 h (MSC)/10 h (N2a)/6 h (HCEC), and (7) with the cotreatment of PTIO and DNIC−COOH for 3 h (MSC)/10 h (N2a)/6 h (HCEC), respectively, were performed in a similar manner.In addition to CCK-8 assay, a live/dead staining was performed using a ViaQuant Viability/ Cytotoxicity Kit (GeneCopoeia, Rockville, MD) according to the manufacturer's instruction.

EPR Investigations on Cellular Uptake of DNICs and Intracellular Formation/Decomposition of Protein-Bound
DNICs in MSC/N2a/HCEC Under Normoxia Condition.Cellular uptake of DNIC−COOH and intracellular formation/decomposition of mononuclear DNICs in MSC/N2a/HCEC cells were investigated using EPR spectroscopy.Briefly, cells were seeded into two (or one) 100 mm dish(es) at a density of 2 × 10 6 cells/dish and incubated at 37 °C overnight.On the next day, cells were treated with 100 μM of DNIC−COOH (or 100 μM of DNIC−COOH to N2a and 75 μM of DNIC−COOH to HCEC) and incubated for 0, 0.5, 1, 2, 3, 4, 8, and 24 h, respectively.After removal of the supernatant solution, cells were washed with DPBS before addition of 1 mL of trypsin (0.05%)/EDTA (0.53 mM) and incubation for 5 min at 37 °C.After the obtained cell suspension solutions from two 100 mm dishes were combined and centrifuged at 100g for 5 min, the supernatant was removed before addition of 100 μL of αMEM to DNIC-treated MSC (or MEM to DNIC-treated N2a and HSFM to DNIC-treated HCEC).This solution was then transferred into EPR quartz tube and frozen in N 2(l) before the EPR measurements.For each measured EPR spectrum, double integration was executed to obtain the integrated EPR intensity under each condition.Moreover, time-dependent decay of integrated EPR signal intensity after treatment of DNIC−COOH was fit to pseudo-first-order kinetics in order to determine the intracellular half-life for EPR-active mononuclear DNIC.Three independent experiments were executed to measure the average half-life for intracellular decomposition of EPR-active mononuclear DNIC after treatment with DNIC−COOH at 37 °C.On the other hand, at the time point with maximum integrated EPR signal intensity, spin quantitation of intracellular EPR-active DNICs was performed using mononuclear DNIC [PPN][(NO) 2 Fe(S 5 )] as a standard according to the methodology used in the previous work. 50Cellular uptake of DNIC−COOMe (2.5 μM to MSC, 10 μM to N2a, or 1 μM to HCEC) and intracellular formation/decomposition of EPR-active mononuclear DNICs were investigated using EPR spectroscopy in a similar manner.
Pretreatments of Inhibitors.Studies of cellular uptake mechanisms were investigated using EPR spectroscopy in combination with pretreatments of alternative inhibitors.After the MSC cells were seeded into two (or one) 100 mm dish(es) at a density of 2 × 10 6 cells/dish and incubated at 37 °C overnight, sequential treatments of 50 μM of NEM for 0.5 h and 100 μM of DNIC−COOH for 2 h were performed before the treated MSC cells were collected for EPR measurement following the procedure described above.Under a similar procedure, (1) N2a

Kinetic Study on Intracellular Release of NO in MSC/N2a/HCEC Cells after Treatment of DNICs
Intracellular release of NO DNIC−COOH was validated using the fluorescence probe 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate (DAF-FM) in combination with fluorescence analysis techniques.Briefly, the MSC cells (or N2a and HCEC cells) were seeded onto a 96-well black plate at a density of 6 × 10 3 cells/well (or 1 × 10 4 cells/well for N2a and 8 × 10 3 cells/well for HCEC cells) and incubated at 37 °C overnight.On the next day, the MSC cells were first treated with 10 μM of DAF-FM (or 5 μM for N2a and HCEC cells) and incubated at 37 °C for 0.5 h.After removal of the supernatant solution, 100 μM of DNIC−COOH was added to the MSC cells (or 100 μM of DNIC−COOH to N2a cells and 75 μM of DNIC−COOH to HCEC cells) and incubated for 0, 1, 2, 4, 6, and 8 h, respectively, in the dark.Fluorescence intensity of the cells was then recorded using a microplate reader (SpectraMax iD3, Molecular Devices, San Jose, CA) with an excitation wavelength at 485 nm and an emission wavelength at 525 nm.Three independent experiments were executed to measure the timedependent change of average fluorescence intensity.Time-dependent change of average fluorescence intensity for the MSC/N2a/HCEC cells with sequential treatment of DAF-FM (10 μM for MSC cells, 5 μM for N2a cells, 5 μM for HCEC cells; 0.5 h) and DNIC−COOMe (2.5 μM for MSC cells, 10 μM for N2a cells, 1 μM for HCEC cells) was determined in a similar manner.
Confocal microscopy images for the MSC/N2a/HCEC cells that received sequential treatment of DAF-FM and DNIC−COOH (or DNIC−COOMe were taken following the procedure described below.Cells were plated in 24-well plates at a density of 2.5 × 10 4 cells/well for MSC cells, 6 × 10 4 cells/well for N2a cells, and 2.5 × 10 4 cells/well for HCEC cells, respectively, on cover glasses (18 mm in diameter) for 24 h and incubated with the DAF-FM (10 μM for MSC cells, 5 μM for N2a cells, 5 μM for HCEC cells) for 0.5 h.After the cell culturing media were removed, DNIC−COOH (500 μM for MSC cells, 100 μM for N2a cells, 75 μM for MSC cells) was then added and incubated for 1 h (MSC), 2 h (N2a), and 6 h (HCEC), respectively.Then, the supernatant solution was removed, and the treated cells were washed thrice with DPBS before the cells were fixed with 4% formaldehyde solution for 15 min at room temperature.After the cells were washed thrice with DPBS followed by membrane permeabilization with 0.25% Triton for 5 min, the cover glasses containing fixed cells were washed thrice with DPBS and mounted onto a microscope slide containing 10 μL of mounting medium (FluoroQuest) with 4′-6-diamidino-2phenylindole (DAPI) for cell nuclei staining.The optical images were captured using a confocal imaging system (ZEISS, LSM 780).Confocal microscopy images of the MSC/N2a/HCEC cells with sequential treatment of DAF-FM (10 μM and 0.5 h for MSC, 5 μM and 0.5 h for N2a, 5 μM and 0.5 h for HCEC) and DNIC−COOMe (12.5 μM and 1 h for MSC, 10 μM and 0.5 h for N2a, 1 μM and 0.5 h for HCEC) were taken under a similar procedure.

Investigation of DNIC-Induced cGMP Formation
The MSC cells were plated on a cover glass (18 mm in diameter) loaded into a 24-well plate at a density of 1.7 × 10 3 cells/well until the desired confluency was reached.After the culture media were removed, 100 μM of DNIC−COOH (or 2.5 μM of DNIC−COOMe) was added to the cell culture and incubated for 2 h.The culture media were removed before the MSC cells were washed three times with PBS, whereas the washed MSC cells were further fixed with 4% paraformaldehyde (PFA) solution for 15 min at room temperature.After the PFA solution was removed, the MSC cells were washed twice with PBS for 3 min and permeabilized using 0.1% Triton X-100 (in PBS) at room temperature for 4 min.Subsequently, the MSC cells were washed twice with PBS for 3 min and incubated with a blocking buffer (2.5% BSA in PBS) at room temperature for 30 min.Then, the MSC cells were incubated with primary antibody solution (1:1000 for rabbit anti-cGMP antibody, purchased from Biorbyt, Cambridge, U.K.) at room temperature for 1 h before the MSC cells were washed three times with PBS for 3 min.Cover glasses containing the fixed MSC cells were mounted onto a microscope slide loaded with Fluoroshield mounting medium with DAPI to stain the nuclei before the optical images were captured using a confocal imaging system (Leica TCS-SP-X AOBS).
After the treatments of DNIC−COOH/DNIC−COOMe to the MSC cells following the procedure described above, the intracellular level of HO-1 protein in DNIC-treated MSC cells was also determined by ELISA (Abcam, Cambridge, MA) according to the manufacturers' protocols.

Assessment of DNIC-Induced Cytoprotective Effect in MSCs and HCECs against Hostile Microenvironment
To investigate the capacity of DNIC preconditioning to protect the MSC cells against elevated oxidative stress, cells were pretreated with 2.5 μM of DNIC−COOMe or 100 μM of DNIC−COOH for 8 h before being incubated with 250 μM H 2 O 2 for another 24 h.Cell viability was evaluated by live/dead staining or CCK-8 assay.In a similar manner, DNIC-induced cytoprotective effect in HCECs was evaluated through pretreatment of 1 μM of DNIC−COOMe or 75 μM of DNIC−COOH for 24 h.
To simulate the microenvironment of ischemic tissues, the MSC cells with or without DNIC priming were cultured using glucose-free DMEM in a humidified hypoxia incubator (1% oxygen; Heracell VIOS 160i, Thermo Fisher Scientific) for 48 h.Cell viability was assessed by live/dead staining or CCK-8 assay.

Statistical Analysis
Data are expressed as the mean ± SEM for all of the in vitro study and the mean ± SD for the rest.Statistical analyses were performed using GraphPad Prism software (version 8.2; San Diego, CA).For a comparison of two groups, an unpaired, two-tailed Student's t test was used.One-way analysis of variance (ANOVA) with Tukey's correction was employed for comparisons of three or more groups.Differences were considered to be significant at p < 0.05.
Scheme 1. Mechanisms for Reversible Interactions between DNICs and Biological Thiols, Release of Free NO, and Direct Transfer of NO toward Fe n+ -Porphyrin Center in Myoglobin

Figure 1 .
Figure 1.NO-release profiles of DNIC−COOH (25 μM, blue circle) and DNIC−COOMe (25 μM, red square) at pH 7.4 (a) without or (b) with the presence of 200 μM Cys under normoxia condition at 37 °C, which were fitted to pseudo-first-order kinetics (dashed line).Data show the mean ± standard deviation (SD) from three independent experiments.

2 a
Concentrations of the reactants are DNIC−COOH = 25 μM, DNIC−COOMe = 25 μM, deoxyMb = 5 μM, and metMb = 5 μM.200 μM of Cys was used for the study of NO-delivery reactivity of DNICs, while 10 mM of Cys was used for the study of degradation of DNICs.b Obtained from the reaction of DNIC−COOH/DNIC−COOMe and Cys.c Obtained from the reaction of DNIC−COOH/DNIC−COOMe and BSA in αMEM with 20% FBS, MEM with 5% FBS, or HSFM with 2% FBS.d Obtained from treatments of DNIC−COOH/DNIC−COOMe to MSC, N2a, or HCEC.e Half-life (t 1/2 ) for release of NO from DNICs monitored using total nitrate/nitrite assay.f Reaction time for complete conversion of deoxyMb/metMb into MbNO monitored using UV−vis spectroscopy.g Half-life (t 1/2 ) for degradation of DNICs monitored using EPR spectroscopy.DNIC−COOH to the cysteine residue in bovine serum albumin (BSA) yielding BSA-bound DNIC, the fast BSA-binding kinetics of DNIC−COOMe may rationalize the rapid cellular uptake of DNIC−COOMe into MSC, human CEC (HCEC), and neuro-2a cells (N2a) through a thiol-mediated pathway.In comparison, DNIC−COOH displays a slow cellular uptake process through endocytosis instead of a thiol-mediated pathway, followed by steady intracellular release/transfer of NO.Of importance, this steady kinetics for intracellular delivery of NO featured by DNIC−COOH reflects on the extended/ elevated overexpression of HO-1 and enhanced survival of DNIC−COOH-primed MSC/HCEC against hostile microenvironments.(Cys) Regulate the Release of NO and Transfer of NO toward Fe n+ -porphyrin Center (n = 2 or 3)

Figure 4 .
Figure 4. Time-dependent change of EPR spectra for MSCs treated with (a) DNIC−COOH (blue) and (b) DNIC−COOMe (red).Formation and decay of intracellular mononuclear DNIC upon treatment of DNIC−COOH (blue) and DNIC−COOMe (red) to (c, d) MSCs, (e, f) N2a,and (g, h) HCEC, respectively, whereas the decay is fitted to pseudo-first-order kinetics (dashed line).Data show the mean ± standard error of the mean (SEM) (n = 3).EPR monitoring on mononuclear DNIC derived from the treatments of DNIC−COOMe under hypoxia conditions is depicted in gold in the corresponding figures.
f, delayed decays of the EPR signals at g av = 2.03 were observed after incubation of dinuclear DNIC−COOH/DNIC− COOMe in cell culturing media with FBS under hypoxia conditions (∼1% O 2 ).In contrast, the accelerated degradations of albumin-bound DNICs in cell culturing media with FBS under normoxia conditions suggested an O 2 -induced mechanism for the release of NO (Scheme 1g and Figure S5g).During the reaction of metMb with dinuclear DNIC−COOH/DNIC− COOMe in the presence of BSA, no change of UV−vis absorption bands at (409, 504, 631) nm excluded the NO −transfer reaction despite the formation of mononuclear DNIC [(NO) 2 Fe(SR)(S Cys-albumin )] n− (Figure S6).As opposed to the NO − -transfer reactivity of mononuclear DNIC [(NO) 2 Fe(SR)-(S Cys )] n− to metMb, this absent NO − -transfer reactivity of albumin-bound DNIC [(NO) 2 Fe(SR)(S Cys-albumin )] n− may be ascribed to the steric hindrance between metMb and albumin.Based on the cell viability study shown in Figure S7, both DNIC−COOH and DNIC−COOMe displayed a dosedependent cell-proliferation effect on N2a/HCEC, whereas DNIC−COOH features an improved biocompatibility with MSC/N2a/HCEC in comparison with DNIC−COOMe.Accordingly, treatment of 100 μM of DNIC−COOH (or 2.5 μM of DNIC−COOMe) to MSC, treatment of 100 μM of DNIC−COOH (or 10 μM of DNIC−COOMe) to N2a, and treatment of 75 μM of DNIC−COOH (or 1 μM of DNIC− COOMe) to HCEC were determined as the optimal conditions for further in vitro investigations.Upon treatment of DNIC−COOH to MSC, formation of distinctive EPR signal at g = 2.041, 2.034, and 2.015 observed in the whole-cell EPR spectra indicated the cellular uptake of DNIC−COOH followed by intracellular transformation into mononuclear DNIC [(NO) 2 Fe(SR)(SR′)] n− (Figure 4a and Tables 2 and S2).Based on the reactivity study discussed above and other reported literature, intracellular Cys or other Cyscontaining proteins are proposed to facilitate this assembly of mononuclear DNIC [(NO) 2 Fe(SCH 2 CH 2 COOH)(SR′)] n− (SR′ = S Cys or S Cys-protein , Scheme 2b).As shown in Figures 4a,b and S8 and Table 2, effective cellular uptake of dinuclear DNIC−COOH/DNIC−COOMe by MSC, N2a, and HCEC, respectively, followed by intracellular conversion into mononuclear DNIC [(NO) 2 Fe(SR)(SR′)] − (SR = SCH 2 CH 2 COOH or SCH 2 CH 2 COOMe; SR′ = S Cys or S Cys-protein ) was evidenced by the generations of similar EPR features at g av = 2.03.According to time-dependent change of whole-cell EPR spectra, DNIC−COOMe displayed rapid cellular uptake and intracellular transformation processes with t max = 0.5 h upon treatments to MSC, N2a, and HCEC, respectively (Figures 4c−h and S8).In contrast, DNIC−COOH exhibited delayed

Figure 7 .
Figure 7. (a) Results of real-time PCR showing the temporal profile of HMOX1 gene after treatments of DNIC−COOH (blue) and DNIC−COOMe (red) to MSC. *p < 0.05, **p < 0.01, and ***p < 0.001.(b) HMOX1 mRNA level in MSC after alternative treatments for 3 h.***p < 0.001 for comparison with the control group.(c) Results of ELISA showing the temporal profile of intracellular HO-1 protein level without (black) or with treatments of DNIC−COOH (blue) and DNIC−COOMe (red).*p < 0.05 and ***p < 0.001 for comparison with the control group; ## p < 0.01 for comparison with the DNIC−COOMe group.Results of real-time PCR showing the temporal profile of HMOX1 gene after treatments of DNIC− COOH (blue) and DNIC−COOMe (red) to (d) N2a and (e) HCEC.Relative HMOX1 mRNA level in (f) MSC at 4 h, (g) N2a at 10 h, and (h) HCEC at 6 h after treatments of DNIC−COOH without or with the PTIO/ODQ.Data show the mean ± SEM (n = 3).
, elevated formation of cGMP observed in the DNIC−COOMe-treated (or DNIC−COOH-treated) MSC suggested that the DNIC-induced nitrosylation of Fe-porphyrin center in sGC enhanced the conversion of GTP into cGMP.Moreover, treatment of DNIC−COOH to MSC triggered maximal 49.5 ± 2.9-fold activation of HMOX1 gene at 3 h and 32.0 ± 4.4-fold expression of HO-1 protein at 8 h (Figure 7a−c).In comparison, reduced activations of HMOX1 gene (18.8 ± 2.3-fold at 3 h) and HO-1 protein (8.1 ± 0.9-fold at 6 h) were observed in the DNIC−COOMe-treated MSC, although similar kinetic profiles for regulations of HMOX1 gene and HO-1 protein were observed.Moreover, absent regulations of HMOX1 gene were observed in the MSC treated with degraded DNIC− COOH/DNIC−COOMe.Presumably, rapid cellular uptake of DNIC−COOMe followed by fast intracellular release/transfer of NO may overwhelm the capacity and kinetics of NO-sGC-cGMP-HO-1 pathway.In contrast, compatible kinetics for intracellular NO-delivery reactivity of DNIC−COOH may rationalize the enhanced activations of HO-1.Parallel qPCR investigations on transcriptional regulations of HMOX1 gene by DNIC−COOH/DNIC−COOMe in N2a and HCEC were collected in Figure 7d,e.Of importance, DNIC−COOH, as opposed to DNIC−COOMe, triggered enhanced transcrip-

Figure 8 .
Figure 8. (a, b) Representative live/dead images of MSCs and (c, d) their corresponding cell viability following various treatments.Scale bar, 200 μm.(e) Cell viability of HCEC under alternative treatments.***p < 0.001 for comparison with the control group.
−h, cotreatment of PTIO to MSC/HCEC and pretreatment of ODQ to MSC/N2a/HCEC, respectively, resulted in no significant perturbation on the transcriptional activation of HMOX1 gene by DNIC−COOH.Regarding the successful scavenging of free NO released from DNIC−COOH by PTIO (Figure S12), the null effect of PTIO on DNIC− COOH-induced regulation of HMOX1 gene in MSC/HCEC suggested the importance of NO-transfer reactivity of mononuclear DNIC [(NO) 2 Fe(SR)(S Cys )] n− (Scheme 2g).That is, direct transfer of NO from mononuclear DNIC [(NO) 2 Fe(SR)(S Cys )] n− toward Fe 2+ -porphyrin center, instead of the released NO, may trigger activation of sGC for regulation of downstream HO-1.During the treatment of higher concentration of PTIO (100 μM) to N2a, reduced transcriptional activation of HMOX1 gene by DNIC−COOH supports the fact that free NO released from mononuclear DNIC [(NO) 2 Fe(SR)(SR′)] n− participates in the modulation of HO-1 in N2a.On the other hand, upon oxidation of the Fe 2+ -porphyrin center within sGC to an Fe 3+ state by pretreatment with ODQ, retained activation of HMOX1 gene by DNIC−COOH (Figure 7f−h) highlighted the critical role of intracellular assembly of mononuclear DNIC [(NO) 2 Fe(SR)(S Cys )] n− on direct transfer of NO − to the Fe 3+ -porphyrin center of sGC (Scheme 2g,k).
n− /[(NO) 2 Fe(SR)-(S Cys-albumin )] n− . 5. On the basis of qPCR/ELISA analyses and cell viability study, steady endocytic uptake of DNIC−COOH and sustained intracellular release/transfer of NO trigger enhanced upregulation of HO-1 and cytoprotective effects.Thus, the survival of DNIC−COOH-primed MSC/HCEC was improved under stressed conditions mimicking the hostile microenvironments after cell transplantation.

Table 2 .
EPR Parameters for Mononuclear and Dinuclear DNICs Explored in This Study Obtained from the reaction of DNIC−COOH/DNIC−COOMe and Cys.b Obtained from the reaction of DNIC−COOH/DNIC−COOMe and BSA in αMEM with 20% FBS, MEM with 5% FBS, or HSFM with 2% FBS.c Obtained from treatments of DNIC−COOH/DNIC−COOMe to MSC, N2a, or HCEC. a