English

• Carbon monoxide (CO), endogenously produced from the degradation of haem mediated by haem‐ oxygenases in mammalian cells and tissues, is a crucial signaling gasotransmitter as nitride oxide (NO) and hydrogen sulfide (H₂S)^[1-2]. CO is recognized to be involved in multiple signaling pathways in living organisms thus playing various valuable physiological functions^[3-9]. It should be noted that responses of gaseous CO are highly associated with its intracellular concentrations, most likely, anti-inflammatory, anti-apoptosis, and antiproliferative effects in low concentrations whereas mitochondrial electron‐transport inhibition and toxic effects in high concentrations^[10-11]. Therefore, the delivery of CO in control is highly desired to exploit its precious values in medicine.

  To the issue, CO-releasing molecules (CORMs), first reported by the Motterlini group^[12], have been drawing rising attention. A wide range of CORMs, typically categorized into the metal-based carbonyl complexes^{[9, 13-17]} and the organic compounds^[18-19], are developed because of the concerned properties of solubility, stability, and cytotoxicity for the proposed clinical applications. Ideally, the CORMs are supposed to be as simple as possible from either the easy-going preparation or the minimal metabolic risk of the residue derivative of the CO-release process. Iron-based CORMs would be suitable candidates due to not only iron is an essential element in the body and thus owns perfect, but also iron exists in various oxidation states to remind large numbers of iron carbonyl complexes for screening. Indeed, a massive of monoiron and diiron carbonyls are reported as CORMs^{[13, 20-32]}.

  Among them, a fac-[Fe(CO)₃I₃]^- anion, affording from the reactions between a universal precursor cis‐[Fe(CO)₄I₂] and alkylamines (Scheme 1), is concerned by us due to its facile preparation, appropriate solubility in water, and tunable CO-release^[33]. The solubility and CO-release of the anion would be adjusted by the aminium cations dependent on the hydrophobicity of amines (the logarithm of the octanol/water partition coefficient, lg P value). In particular, iodine radical should be generated during the CO release to induce a high cytotoxicity comparable to cisplatin, a commercial clinical anticancer drug. To expand our interest in the anticancer potency of the anion and example of how the inorganic cations affect its activities, in the present work, five analogues of fac-[Fe(CO)₃I₃]^- anion with a counter ion of inorganic cation, salts 1-5 were prepared (Scheme 1). Their CO-releasing behaviors in media and anticancer activities on bladder cancer cells were evaluated. Moreover, a mechanistic probing into the cytotoxicity of the anion was elucidated.

  Scheme 1

  

  Scheme 1.  Synthetic route for the fac-[Fe(CO)₃I₃]^- anion with an aminium cation or an inorganic cation (1-5) from the universal precursor cis-[Fe(CO)₄I₂]

  下载: 全尺寸图片 幻灯片

  1.   Experimental

  1.1   Chemicals and methods

  All the reactions were operated at an argon atmosphere using Schlenk lines. Degassed dry solvents were prepared by a Phoenix SDS5 system. Sodium iodide (NaI), potassium iodide (KI), magnesium iodide (MgI₂), calcium iodide anhydrous (CaI₂), and ammonium iodide (NH₄I) were purchased from Aladdin company and used as received. The precursor, cis-[Fe(CO)₄I₂] was prepared using the literature method and confirmed by Fourier transform infrared (FTIR) spectra^[34]. The ferroptosis inhibitors, Ferrostatin-1 and Liproxstatin-1, were dissolved in DMSO and performed in a concentration of 1 μmol·L^-1, respectively.

  FTIR spectra were collected on a Nicolet iS10 spectrometer. The samples were dissolved in suitable media and syringed (ca. 0.2 mL) into a cell with the CaF₂ windows for detection. Samples in methanol at a concentration of 1×10^-4 mol·L^-1 were prepared for ultraviolet-visible (UV-Vis) spectra on an Evolution 201 spectrometer. ¹H and ¹³C nuclear magnetic resonance (NMR) spectra were performed on Varian 400-MR equipment in acetone-d₆ solvent. The data, chemical shifts (δ) were referenced to the solvent residues with the values of parts per million:

  $ \delta=\left(\nu_{\text {sample }}-\nu_{\text {ref }}\right) / \nu_{\mathrm{o}} \times 10^6 $

  whereas ν_sample, ν_ref, and ν_o are defined as the resonance frequencies of the sample and the reference (such as tetramethylsilane), and the frequencies of the instrument, respectively. Mass spectra (MS) were collected on a Waters LCT Premier XE instrument in a methanol solution. Iron contents of the salts were measured by inductively coupled plasma atomic emission spectroscopy (ICP-AES) on a 710-ES Spectrometer (Varian, USA). The samples were digested by aqua regia before collection. Salt 1 in ultrapure water (16 mmol·L^-1) with the presence of 5, 5‐dimethyl‐1‐pyrroline N‐oxide (DMPO, 100 mmol·L^-1) was used for the electron paramagnetic resonance (EPR) spectroscopy.

  1.2   Preparation and characterizations of salts 1-5

  In a general procedure, a flask (100 mL) was charged with cis-[Fe(CO)₄I₂] (422 mg, 1 mmol) and dry acetone (20 mL). Iodide salt in acetone solvent (10 mL) was syringed to the precursor which was cooled by an ice bath. The reaction was stirred for 2 h until the precursor was consumed completely, as indicated by the FTIR spectroscopy (Fig.S1, Supporting information). Afterward, the solution was concentrated to ca. 1 mL under vacuum, and dry hexane was added in to precipitate a black solid, which was further washed by hexane (25 mL×3 times). The solid was dissolved in appropriate dichloromethane (DCM), and the DCM solution was filtrated to remove any unwanted byproduct. Finally, the addition of excess hexane crushed out the product.

  fac-Na[Fe(CO)₃I₃] (1). 150 mg of NaI (1 mmol) in 10 mL of acetone was used to afford the black salt 1. Yield: 490 mg (90%). FTIR (DCM, cm^-1): 2 087 (ν_CO), 2 037 (ν_CO). UV-Vis (MeOH, nm): 288 (ε=1.1×10⁴ L·mol^-1·cm^-1), 359 (ε=5.5×10³ L·mol^-1·cm^-1). ¹³C NMR (101 MHz, acetone-d₆): 217.5 (C≡O). MS (ES^-): m/z=520.628 8 (Calcd. for [Fe(CO)₃I₃]^- fragment, 520.632 5).

  fac-K[Fe(CO)₃I₃] (2). 166 mg of KI (1 mmol) in 10 mL of acetone was used to afford the black solid salt 2. Yield: 526 mg (94%). FTIR (DCM, cm^-1): 2 085 (ν_CO), 2 034 (ν_CO). UV-Vis (MeOH, nm): 287 (ε=7.1×10³ L·mol^-1·cm^-1), 359 (ε=3.2×10³ L·mol^-1·cm^-1). ¹³C NMR (101 MHz, acetone-d₆, ): 216.4 (C≡O). MS (ES^-): m/z=520.623 4 (Calcd. for [Fe(CO)₃I₃]^- fragment, 520.632 5).

  fac-Mg[Fe(CO)₃I₃]₂ (3). 139 mg of MgI₂ (0.5 mmol) in 10 mL of acetone was used to afford the black salt 3. Yield: 450 mg (84%). FTIR (DCM, cm^-1): 2 085 (ν_CO), 2 034 (ν_CO). UV-Vis (MeOH, nm): 290 (ε=7.9×10³ L·mol^-1·cm^-1), 358 (ε=4.0×10³ L·mol^-1·cm^-1). ¹³C NMR (101 MHz, acetone-d₆, ): 218.3(C≡O). MS (ES^-): m/z=520.622 2 (Calcd. for [Fe(CO)₃I₃]^- fragment, 520.632 5).

  fac-Ca[Fe(CO)₃I₃]₂ (4). 147 mg of CaI₂ (0.5 mmol) in 10 mL of acetone was used to afford the black salt 4. Yield: 430 mg (79%). FTIR (DCM, cm^-1): 2 086 (ν_CO), 2 035 (ν_CO). UV-Vis (MeOH, nm): 290 (ε=1.3×10⁴ L·mol^-1·cm^-1), 359 (ε=7.1×10³ L·mol^-1·cm^-1). ¹³C NMR (101 MHz, acetone-d₆): 217.0 (C≡O). MS (ES^-): m/z=520.621 7 (Calcd. for [Fe(CO)₃I₃]^- fragment, 520.632 5).

  fac-[NH₄][Fe(CO)₃I₃] (5). 145 mg of NH₄I (1 mmol) in 10 mL of acetone was used to afford the black salt 5. Yield: 480 mg (89%). FTIR (DCM, cm^-1): 2 087 (ν_CO), 2 037 (ν_CO). UV-Vis (MeOH, nm): 286 (ε=5.9×10³ L·mol^-1·cm^-1), 361 (ε=2.6×10³ L·mol^-1·cm^-1). ¹³C NMR (101 MHz, acetone-d₆): 217.4 (C≡O). MS (ES^-): m/z=520.613 7 (Calcd. for [Fe(CO)₃I₃]^- fragment, 520.632 5).

  1.3   Monitoring the CO-release of salts 1-5 in media

  Generally, salts containing the fac-[Fe(CO)₃I₃]^- anion (48 μmol) were added into a brown glass vial and dissolved in 3 mL of the medium (DMSO, D₂O, and saline). The saline solution was prepared by dissolving the sodium chloride in D₂O solvent at a content of 0.9%. The vial was sealed under an aerobic atmosphere and shielded from lights. Decomposing of the sample at a normal body temperature (37 ℃) was regularly monitored by FTIR spectroscopy till the CO‐ release was completed. Kinetics of the decomposing process for the five salts were obtained by linear fitting of the values of the natural logarithm of the absorbances (ln A) against the corresponding time. Related rate constant and halftime were calculated by adapting the first-order model. Data were reported as "mean ± standard deviation (SD)" based on three independent experiments (n=3).

  1.4   Effect of iodide on the CO-release of salt 1 in D₂O

  The solution of NaI at the concentrations of 0.05, 0.1, 0.2, 0.3, and 0.4 mol·L^-1, was prepared by dissolution of the solid salt (22.5, 45.0, 89.9, 134.9, and 179.9 mg) in D₂O (10 mL) respectively. Salt 1 (26 mg, 48 mmol) was then dissolved into the NaI solution (3 mL) to result in a solution of salt 1 (16 mmol·L^-1) with the presence of different amounts of NaI. Afterward, the decomposing of the anion in the solution was monitored by following a similar procedure as described in Section 1.3 above. Kinetics were estimated and presented as "mean ± SD" (n=3).

  1.5   pH effect on the CO-release of salt 1 in PBS

  The phosphonate buffer saline (PBS) was made up of a mixture of Na₂HPO₄ and NaH₂PO₄ in a D₂O solvent. The calculated volumes of the Na₂HPO₄ solution (0.2 mol·L^-1) and the NaH₂PO₄ solution (0.2 mol·L^-1), i.e., 0.13 mL vs 1.87 mL, 0.63 mL vs 1.37 mL, 1.62 mL vs 0.38 mL, 1.894 mL vs 0.106 mL, were mixed and diluted by D₂O (2 mL), leading to a PBS (0.1 mol·L^-1) with the pH value tested as 5.7, 6.5, 7.4, and 8.0, respectively. Salt 1 (26 mg, 48 μmol) was then dissolved into the PBS (3 mL) to result in a solution of salt 1 (16 mmol·L^-1) with a different pH value. Afterward, the decomposing of the anion in the solution was monitored by following a similar procedure as described in Section 1.3 above. The corresponding kinetics were estimated and presented as "mean ± SD" (n=3).

  1.6   TMB assay

  The generation of iodine-free radicals derived from the decomposing of the salt 1 was determined by a 3, 3′, 5, 5′-tetramethylbenzidine (TMB) assay as we reported previously^{[33, 35-36]}. Typically, 3 mL of TMB in HAc-NaAc (1.66×10^-4 mol·L^-1, pH=4.5) was treated with a fresh solution of salt 1 (9.19×10^-4 mol·L^-1, MeOH solvent) in the various addition (10, 20, 30, 40, 50, 60, 70, 80, 100 μL). UV-Vis spectra of the solution were collected after each addition. Dose-dependence on the amounts of the fac-[Fe(CO)₃I₃]^- anion against the absorbance of oxidized TMB (A₆₅₀) was obtained.

  1.7   MTT assay

  Anticancer activities of salts 1-5 were evaluated against a human bladder cancer cell line (RT112 cells) as we used before^{[33, 35-36]}. The adhered RT112 cells were grown on a 96-well plate with ca. 10 000 cells per well in Roswell Park Memorial Institute (RPMI)-1640 medium. The salts freshly dissolved in the culture medium (1 mmol·L^-1) were added into the well and diluted by the culture medium to a desired concentration (5, 10, 20, 50, 100, 200, and 400 μmol·L^-1) with 200 μL in total volume for each well. Cells without the treatment of the salts were examined as the control. Six parallel wells were paved for each concentration. The cells were incubated for 24 h. Afterward, the culture medium was aspirated and the methyl thiazolyl tetrazolium (MTT) in the culture medium (40 μL, 5 mg·mL^-1) was added into each well. The cells were incubated for another 4 h before being lysed by DMSO (150 μL). The absorbance of the wells was measured by a microplate reader at 570 nm. Viabilities of the cells were obtained using the corresponding absorbance against that of the control group. IC₅₀ values of these salts were finally reported according to their dose-dependence curves. Cytotoxicity of the residue, derived from the decomposing of salt 1 in water medium for 6 h with the completed CO-release, was also carried out by following the same procedure above.

  To elucidate whether the generated iodine radicals contributed to its anticancer activity, the bladder cancer cells were co-incubated with the salt 1 at various concentrations (0, 15, 30, and 60 μmol·L^-1), with the absence and presence of a radical scavenger, 4-hydroxy‐TEMPO (Tempol, 1 mmol·L^-1) for 24 h, respectively. The cellular viabilities were assessed by the standard MTT assay as described above.

  To probe a ferroptosis-involved pathway, two inhibitors Ferrostatin-1 and Liproxstatin-1 were implied at a concentration of 1 μmol·L^-1. The RT112 cells were incubated with either Ferrostatin-1 alone, Liproxstatin-1 alone, salt 1 alone, or the salt 1 and the inhibitor together for 24 h before the viabilities were evaluated.

  1.8   Trypan blue staining assay

  The bladder cancer cells were grown on a 24-well plate with ca. 20 000 cells per well in the RPMI-1640 medium for 24 h. The adhered cells were then treated with salt 1 in the culture medium at a concentration of 0, 15, 30, and 60 μmol·L^-1, respectively. After being incubated for 4 h, the aged medium was aspirated, and trypan blue in buffer (0.4%) was added into each well (0.5 mL). The cells were stained for another 5 min before being imaged by an inverted microscope (Olympus, CKX41). It should be noted that the dead cells will be stained in blue color and the living cells remain.

  1.9   Measurements of mitochondrial networks and mitochondrial ROS

  The level of reactive oxygen species (ROS) in the mitochondria was detected by the mitoSOX probe and the mitochondrial morphology was examined by the mitoTracker Green probe. The bladder cancer cells were grown in a glass-bottomed (35 mm) confocal dish with about 2×10⁶ cells in the RPMI-1640 medium for 24 h. The adhered cells were then treated with the fac-[Fe(CO)₃I₃]^- anion at a concentration of 15 μmol·L^-1 in the culture medium. Cells without the treatment were performed as a control. After being incubated for 24 h, the cells were stained with several dyes at 37 ℃, firstly the mitochondrial probe, mitoTracker Green (50 nmol·L^-1) for 15 min, and then the mitoSOX (5 μmol·L^-1) for 20 min, finally the Hoechst (1 μg·mL^-1) for 15 min. Afterward, the confocal laser scanning microscope (CLSM) images of the cells were captured by using a Zeiss LSM 800 microscopy with a ×63 oil objective. The mitoTracker Green probe was fluorescent at 512 nm with an excitation of 490 nm, the mitoSOX fluorescent at 572 nm with an excitation of 557 nm, and the Hoechst fluorescent at 455 nm with an excitation of 348 nm.

  1.10   Western blotting

  The bladder cancer cells were grown in a six-well plate with ca. 2×10⁶ cells per well in the RPMI-1640 medium for 24 h. The adhered cells were then treated with the fac-[Fe(CO)₃I₃]^- anion at various concentrations (0, 1, 5, and 10 μmol·L^-1) in the media. After being incubated for 24 h, immunoblotting analysis was performed to elucidate mitochondrial-involved pathways for the cancer cellular death. The bladder cancer cells were lysed with the radioimmunoprecipitation assay (RIPA) buffer containing 1% protease inhibitor and 1% phosphatase inhibitor. Afterward, the supernatants were collected and the proteins′ concentrations were determined by a bicinchoninic acid (BCA) protein assay kit. The proteins were separated by sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS‐PAGE) and blotted onto nitrocellulose membranes. The following primary antibodies were used: Parkin antibody (ab77924, dilution 1∶1 000), and glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) (CST5174, dilution 1∶1 000). After incubation with primary and secondary antibodies, Western blot bands were obtained by using an Amersham Imager 680.

  2.   Results and discussion

  2.1   Synthesis and characterizations of salts 1-5

  The fac-[Fe(CO)₃I₃]^- anion with a counter inorganic cation, salts 1-5 were prepared by reacting a universal precursor cis-[Fe(CO)₄I₂] with the corresponding iodide salts (MI_n) in acetone at mild conditions^[33-39], as depicted in Scheme 1. All the salts were obtained in solid state with a high yield (79%-94%). Differed from the precursor cis-[Fe(CO)₄I₂] with three carbonyl bands at 2 136, 2 090, and 2 072 cm^-1 (Fig.S1, Supporting information), the carbonyl absorption of all the five products shift towards lower frequencies with two characteristic vibrational bands at ca. 2 085 and 2 035 cm^-1, as shown in Fig. 1. The shifts are attributed to the substitution of one CO by a nucleophilic iodide to enhance the electron-density of the metal centre. Additionally, the carbonyl patterns of the salts match the facial orientation of three carbonyls, consistent well with the analogues of aminium salts (2 086, 2 035 cm^-1)^[33]. Therefore, the formation of fac-[Fe(CO)₃I₃]^- anion should be confirmed from their FTIR spectroscopy.

  Figure 1

  

  Figure 1.  FTIR spectra of the salts 1-5 in DCM solvent

  下载: 全尺寸图片 幻灯片

  Nevertheless, salts 1-5 were further structurally determined using UV-Vis, NMR spectroscopies, and mass analysis (Fig.S2-S14). UV-Vis spectra of salts 1-5 (Fig.S2) exhibit two intense bands (ca. 288, 359 nm) shouldering with a weak band (468 nm). The two strong bands are assigned to the charge transfer of ligand-to-metal transitions (LMCT), and the weak band for a d-d electronic transition of the metal centre, respectively. Additionally, carbon NMR spectra of salts 1-5 demonstrate the high symmetry of the fac-[Fe(CO)₃I₃]^- anion, with only one carbonyl peak, ca. 217 observed (Fig.S3-S7). The fac-[Fe(CO)₃I₃]^- anion was also determined by the high-resolution mass analysis, evidence of the signals of ca. 520.62 (m/z) (Fig.S9-S13) for the successful assignment. The selected characterized data are summarised in Table 1. It seems that the inorganic cations did not exhibit a significant electronic effect on the anion, indicative of the UV-Vis and FTIR spectroscopies. Moreover, salts 1-5 are stable under an inert atmosphere in the solid but extremely hygroscopic. This hygroscopicity made their elemental analysis randomly. Alternatively, their purity was detected by ICP-AES. The results show a purity between 96%-99% for the five salts (Table S1).

  Table 1

  

  Table 1.  Selective characterized data of salts 1-5

  下载: 导出CSV

  Salt/Mn+	νCO[a] / cm-1	λmax[b] / nm	δC[c]	m/z of the anion[d]
  1/Na+	2 087, 2 037	288, 359	217.5	520.628 8
  2/K+	2 085, 2 034	287, 359	216.4	520.623 4
  3/Mg2+	2 085, 2 034	290, 358	218.3	520.622 2
  4/Ca2+	2 086, 2 035	290, 359	217.0	520.621 7
  5/NH4+	2 087, 2 037	286, 361	217.4	520.613 7
  [a] FTIR data of the carbonyl group in DCM; [b] UV-Vis data of the complexes in the methanol; [c] 13C NMR of the carbonyl group in the acetone-d6; [d] Mass analysis in methanol. Calculated m/z signal for the anion: 520.632 5.

  2.2   CO-releasing kinetics of salts 1-5 in various media

  The inorganic salts 1-5 have decent solubility in various solvents such as DCM, THF, acetone, DMSO, and water. Although these salts are stable enough in the low-polar solvents (DCM, THF, acetone), they certainly decomposed via solvolysis in the strong-polar solvents. Decomposition associated with CO release of these salts in various media, e.g., DMSO, D₂O, and saline (0.9% NaCl in D₂O solvent) was monitored by FTIR spectroscopy as we used previously^[33-36]. As shown in Fig.S14 (DMSO), Fig.S15 (D₂O), and Fig. 2 (saline), the characteristic carbonyl absorption of salts 1-5 was attenuated steadily in the three media during the stability study. Since there were no new carbonyl species observed in the process, the degradation of the carbonyl absorption should be associated with the free CO liberating. Accordingly, kinetic analysis was successfully adopted by linear fitting of the lnA values against the time. All the salts were abided by the first-order model in these media. Hence, CO‐releasing kinetics of salts 1-5 about the observed constant (k_obs) and halftime (t_1/2) were calculated as summarized in Table 2.

  Figure 2

  

  Figure 2.  Infrared spectral variation of the fac-[Fe(CO)₃I₃]^- anion for salts 1 (a), 2 (b), 3 (c), 4 (d), 5 (e) in saline and the corresponding kinetic plots (f), obtained by linear fitting of the ln A values at 2 090 cm^-1 against reaction time

  下载: 全尺寸图片 幻灯片

  Table 2

  

  Table 2.  CO-releasing kinetics of the fac-[Fe(CO)₃I₃]^- anion for salts 1-5 in DMSO, D₂O, and saline media

  下载: 导出CSV

  Salt/Mn+	kobs / min-1				t1/2 / min
  	DMSO	D2O	Saline		DMSO	D2O	Saline
  1/Na+	(10.4±0.1)×10-3	(14.3±0.1)×10-3	(11.3±0.3)×10-3		66±1	48±1	61±2
  2/K+	(10.2±0.2)×10-3	(13.6±0.9)×10-3	(16.0±1.4)×10-3		68±1	52±4	44±4
  3/Mg2+	(11.2±0.2)×10-3	(10.5±0.2)×10-3	(11.1±0.3)×10-3		62±1	66±1	62±2
  4/Ca2+	(7.6±0.2)×10-3	(11.2±0.7)×10-3	(15.1±0.9)×10-3		91±3	62±4	46±3
  5/NH4+	(10.7±0.2)×10-3	(14.6±0.5)×10-3	(11.9±0.7)×10-3		65±1	48±2	58±4
  Data were presented as "mean±SD" (n=3).

  Previously, we reported the decomposing kinetics of the fac-[Fe(CO)₃I₃]^- anion with an organic aminium cation (Scheme 1)^[33]. In DMSO, the kinetics of them were abided by the first-order model and were relevant to the lg P values of the amines, and in D₂O, their decomposition was complicated with a two-stage process assigned under the aerobic atmosphere. Herein, the analogues of inorganic cations (salts 1-5) performed a similar CO release via solvolysis, but their kinetics were quietly diverse to the aminium salts. In DMSO, the CO release of salts 1-5 was accelerated (2-3-fold) against the aminium salts (t_1/2=116-185 min), whereas in aqueous media, the decomposition of them was slower down (2-3-fold) than the aminium salts and only one‐stage progress rather than two‐stages was evidenced (Fig. 2, Fig.S15, and Table 2). These results are plausibly attributed to an oppositive partition effect of the inorganic cations versus the organic aminiums during their solvolysis in DMSO and aqueous media.

  2.3   Dose-dependent effect of iodide on the CO- release of fac-[Fe(CO)₃I₃]^- anion

  Concerning the iodide involved in the decomposition of fac-[Fe(CO)₃I₃]^- anion^[33], a dose-dependent effect of iodide on the CO-releasing process was evaluated from the FTIR spectroscopical monitoring. Given salt 1 as an example (Fig. 3 and Table 3), the increasing amounts of iodide in the solution did not change anything but suppressed the degradation of the anion in serious. For example, in Table 3, k_obs of salt 1 at 0.4 mol·L^-1 of NaI (2.2×10^-3 min^-1) was delayed by 6.5-fold compared to that in the absence of NaI (14.3×10^-3 min^-1). Surprisingly, fitting the kinetics of the anion against their corresponding concentrations of presented NaI salt afforded an excellent linear relationship (Fig. 4, R²=0.99). The dose-dependent effect of iodide reinforces the iodide should be associated in the decomposition process, and renders a new strategy for its controllable CO release from the extra addition of iodine ion.

  Figure 3

  

  Figure 3.  Infrared spectral variation of the fac-[Fe(CO)₃I₃]^- anion at the presence of various concentrations of NaI: 0.05 (a), 0.1 (b), 0.2 (c), 0.3 (d) and 0.4 (e) mol·L^-1 in saline and the corresponding kinetic plots (f), obtained by linear fitting of the ln A values at 2 090 cm^-1 against reaction time

  下载: 全尺寸图片 幻灯片

  Table 3

  

  Table 3.  Kinetic data of degradation of salt 1 at various concentrations of iodine ions

  下载: 导出CSV

  Parameter	cNaI / (mol·L-1)
  	0	0.05	0.1	0.2	0.3	0.4
  kobs / min-1	(14.3±0.1)×10-3	(8.0±0.1)×10-3	(5.5±0.2)×10-3	(3.9±0.1)×10-3	(2.9±0.1)×10-3	(2.2±0.1)×10-3
  t1/2 / min	48±1	87±1	127±5	178±5	239±8	330±16
  Data were presented as "mean±SD" (n=3).

  Figure 4

  

  Figure 4.  Plot of the halftime of the fac-[Fe(CO)₃I₃]^- anion (1, 16 mmol·L^-1) in D₂O solvent against the concentration of NaI

  Linear relationship: t_1/2=677.05cNaI + 49.85, R²=0.99.

  下载: 全尺寸图片 幻灯片

  2.4   pH effect on the stability of fac-[Fe(CO)₃I₃]^- anion

  Since the body fluid and cytoplasm from different tissues and cells are diverse in acidity, CO releasing behavior of the fac-[Fe(CO)₃I₃]^- anion was investigated to shed some light on its stability relevant to a pH. Taking salt 1 as a representative example, the anion in PBS with different pH values (5.7, 6.5, 7.4, and 8.0) was scrutinized by FTIR spectroscopy, as presented in Fig.S16. Related kinetics were performed by linear fitting to remind again a first-order model for its CO release (Fig. 5). From the data presented in Table 4, it is noteworthy that the anion maintained a comparable CO release in the acidic PBS (pH=5.7, 6.5) as for the neutral aqueous media (Table 2), but degraded quickly in the basic PBS (pH=7.4, 8.0). The accelerated decomposition is plausibly attributed to an appended substitution launched by a rich hydroxyl group in the basic media. Collectively, the stability of fac-[Fe(CO)₃I₃]^- anion performs divergence in the acidic media against the basic solution, demonstrative of a promising pH response to its CO-release.

  Figure 5

  

  Figure 5.  Representatively kinetic plots of the fac-[Fe(CO)₃I₃]^- anion in PBS solution with different pH values

  下载: 全尺寸图片 幻灯片

  Table 4

  

  Table 4.  Kinetic data of the degradation of fac-[Fe(CO)₃I₃]^- anion in PBS solution at different pH environments

  下载: 导出CSV

  Parameter	pH
  	5.7	6.5	7.4	8.0
  kobs / min-1	(12.1±0.1)×10-3	(13.9±0.1)×10-3	(31.0±0.2)×10-3	(28.3±0.1)×10-3
  t1/2 / min	57±1	50±1	22±1	24±1
  Data were presented as "mean±SD" (n=3).

  2.5   Cytotoxicity of the salts 1-5

  To evaluate the cytotoxicity of the inorganic salts 1-5, a human bladder cancer cell line (RT112 cell) was selected as we used previously^{[33, 35-36]}. The adhered RT112 cells were treated with these salts for 24 h. Afterward, their cellular viabilities were determined by MTT kits as represented in Fig.S17. IC₅₀ values of these salts were estimated from the dose-dependent curves and summarized in Table 5. All these salts exhibit cytotoxicity towards the RT112 cells, within an IC₅₀ value ranging from 25 to 43 μmol·L^-1. It seems the cation would modulate the cytotoxicity of the salt to some extent to fall into a preference order: Na⁺ > K⁺ > Mg²⁺ > Ca²⁺ > NH₄⁺. On the other hand, these inorganic salts had lower cytotoxicity than that for the analogous organic aminium salts^[33] plausibly due to their inhibited CO-release in the aqueous media (Section 2.2 and Table 2). As summarized in Table 5, salts 1-5 possess comparative cytotoxicity for some tricarbonyl iodo iron complexes ([Fe(CO)₃I₂(py)]), but higher toxic than that for the tricarbonyl bromo iron complexes ([Fe(CO)₃Br₂(py)])^[35], indicative of the iodine involved in. Furthermore, to verify the cytotoxicity, a trypan blue staining assay was further adopted, from which the death cells were stained in blue to be visualized with the leaving of the living cells. As shown in Fig. 6, significant cellular deaths were evidenced in the presence of the anion for salt 1. In particular, the cell deaths were dose-dependent on the fac-[Fe(CO)₃I₃]^- anion.

  Table 5

  

  Table 5.  IC₅₀ values of salts 1-5 and selected complexes against RT112 cells^*

  下载: 导出CSV

  Complex	IC50 / (μmol·L-1)	Reference
  Na[Fe(CO)3I3] (1)	24.8±1.3	This work
  K[Fe(CO)3I3] (2)	26.8±0.8	This work
  Mg[Fe(CO)3I3]2 (3)	31.7±1.9	This work
  Ca[Fe(CO)3I3]2 (4)	33.3±2.4	This work
  NH4[Fe(CO)3I3] (5)	43.0±4.3	This work
  [EtNH3][Fe(CO)3I3]	15.4±1.1	[33]
  [PrNH3][Fe(CO)3I3]	20.0±1.0	[33]
  [Fe(CO)3I2(py)]	27.4±1.8	[35]
  [{Fe(CO)3I2}2(bipy)]	97.2±5.7	[35]
  [Fe(CO)3Br2(py)]	137.9±5.4	[35]
  [{Fe(CO)3Br2}2(bipy)]	468.7±16.1	[35]
  [Fe(CO)3I2(NH2Et)]	> 100	[36]
  *MTT, 24 h.

  Figure 6

  

  Figure 6.  Trypan blue staining of the bladder cancer cells incubated with the fac-[Fe(CO)₃I₃]^- anion in various concentrations for 4 h

  下载: 全尺寸图片 幻灯片

  2.6   Mechanistic probing the cytotoxicity of fac-[Fe(CO)₃I₃]^- anion

  To pave a mechanistic probing of the cytotoxicity of the anion, an inspection of the decomposing of the anion therefore the species generated from is necessary. The species derived from salt 1 in aqueous were representatively examined. Firstly, the diiodine molecule was evidenced by a purple color of the tetrachloromethane (CCl₄) addition (Fig.S18). Subsequently, mass analysis of the solution after a course of about 4 h was performed, detective of the generated anions of I^-, and [FeI₃]^- from the decomposition, Fig.S19 particularly, to confirm whether a free radical species was involved in the decomposition or not, a solution of salt 1 with the presence of a radical scavenger DMPO, was determined by the EPR spectroscopy (Fig.S20). It is noticed that no radical signals are evidenced. Therefore, a reactive oxygen species plausibly produced from an iron- induced Fenton reaction during the CO release should be ruled out. However, it does not matter whether an iodine radical should be relevant or not since the detection of the iodine radical by DMPO is still struggled. On the other hand, the evidenced diiodine reminds the generation of iodine radicals.

  Thus, a facile assay of TMB was performed to detect the radical generation for the inorganic salts herein, as we adopted previously^{[33, 35-36]}. It is noticed that the colorless TMB was oxidized in the presence of fac-[Fe(CO)₃I₃]^- anion, with a blue color observed, and a new absorption band at 650 nm appeared (Fig. 7), indicative of the generation of iodine radical. For the generation, a dose-dependence on the anion should be further confirmed from the TMB assay. Particularly, the iodine radical is proven to respond to the high toxicity of the anion from an MTT assay by using a free radical scavenger, Tempol. As indicated in Fig. 8, the cytotoxicity of the anion heavily is inhibited with the engagement of Tempol. On the other hand, halide tricarbonyl iron complexes without any production of iodine radicals are well biocompatibility, e.g., [Fe(CO)₃I₂(NH₂Et)]^[36], and [Fe(CO)₃Br₂(py)]^[35] as presented in Table 5. Additionally, the low cytotoxicity of the residue derived from the completed CO-release of the anion, i.e., an IC₅₀ value of ca. 249 μmol·L^-1 for the residue of salt 1 (Fig.S21), would not agree more with the generated radical involved in the toxic of the anion. It should be noteworthy that the scavenger did not suppress the anticancer activity of the salts completely probably due to a CO-involving gaseous therapy since the decomposition of the fac-[Fe(CO)₃I₃]^- anion produced not only radicals but also toxic CO when in a high level.

  Figure 7

  

  Figure 7.  UV-Vis spectra (a) of TMB in HAc-NaAc buffer (pH=4.5) with various additions of salt 1, and the plot (b) of the absorption of oxidized TMB versus the concentration of salt 1

  Determination of the production of iodine radical with a dose-dependent effect of the fac-[Fe(CO)₃I₃]^- anion by a TMB assay.

  下载: 全尺寸图片 幻灯片

  Figure 8

  

  Figure 8.  Viabilities of RT112 cells treated with various amounts of the fac-[Fe(CO)₃I₃]^- anion for 24 h in the absence and presence of Tempol (1 mmol·L^-1)

  下载: 全尺寸图片 幻灯片

  It is recognized that CO is a versatile gaseous transmitter in cells. In a high concentration, CO should inhibit mitochondrial electron transport and protein expression thus to lead cytotoxicity for cancer cells^[40]. Therefore, not only the generation of iodine radical but also CO‐release from the decomposition of the fac‐[Fe(CO)₃I₃]^- anion would modulate the oxidative stress thus the level of ROS and regulate a cellular death, whereas the energy manufacturer as mitochondria is the crucial target for both of them^[41-43]. Mitochondria is an essential organelle in cells that participates in various biological processes, including energy production and cellular death. Dysfunctions in mitochondrial function and oxidative stress are also associated with the occurrence and development of cancer. Therefore, it is meaningful to detect a level of ROS in the mitochondria by a suitable probe such as MitoSOX so that the mitochondrial-related pathway(s) can be uncovered. From the CLSM analysis in Fig. 9, it should be noticed that the mitochondrial ROS level of the bladder cancer cells following treatment of the anion was upregulated in significance. Particularly, mitochondrial fission was evidenced by the enlargement of the anion-treated cell (Fig. 9b). Collectively, the decomposition of the anion with the production of radicals and CO induced a remarkable upregulation of ROS and nonnegligible fission to damage the mitochondrial dynamics and homeostasis.

  Figure 9

  

  Figure 9.  CLSM images of the staining bladder cancer cells incubated with the absence (a) and presence of fac-[Fe(CO)₃I₃]^- anion (15 μmol·L^-1) (b) for 24 h

  Inset: the corresponding enlarged images.

  下载: 全尺寸图片 幻灯片

  Subsequently, to elucidate the plausible mitochondrial-involved pathway(s) for the cancer cellular death, immunoblotting analysis was carried out. A signaling protein, Parkin, encoded by PARK2 and generally participated in the mitochondrial homeostasis and oxidative stress response, was reasonably exampled by Western blotting, with the GAPDH loaded as a control. As depicted in Fig. 10, the Parkin protein in the bladder cancer cells was expressed in a high expression level with the increment of fac-[Fe(CO)₃I₃]^- anion. Mechanistically, these results support the Parkin protein‐ involved pathway(s) for the regulatory of mitochondrial fission, energy metabolism, degradation, and mitophagy to facilitate cellular death, which is stimulated by the production of iodine radicals and therapeutic CO in the decomposition of fac-[Fe(CO)₃I₃]^- anion.

  Figure 10

  

  Figure 10.  Immunoblotting analysis (Western blotting) of Parkin protein in bladder cancer cells following the treatment of fac-[Fe(CO)₃I₃]^- anion for 24 h

  下载: 全尺寸图片 幻灯片

  Since the anion contains an iron(Ⅱ) centre, ferroptosis, a cell death driven by an iron-dependent lipid peroxidation^[44-45], might be also induced by the anion. To confirm this, two radical-trapping agents, Ferrostatin-1^[46] and Liproxstatin-1^[47] were adopted to suppress the ferroptosis from productively interrupting the lipid peroxidation. As depicted in Fig. 11, the cellular deaths were significantly reverted with the co-incubation of the two inhibitors with salt 1, to lead an increment in viability of about 35% and 27%, respectively. The results reinforce that the anion should induce a promising ferroptosis accompanying the cancer therapy.

  Figure 11

  

  Figure 11.  Inhibitors Ferrostatin-1 (1 μmol·L^-1) and Liproxstatin-1 (1 μmol·L^-1) suppress heavily the ferroptosis of RT112 cells driven by the fac-[Fe(CO)₃I₃]^- anion (MTT, 24 h)

  下载: 全尺寸图片 幻灯片

  3.   Conclusions

  In summary, five salts of the fac-[Fe(CO)₃I₃]^- anion containing an inorganic cation, fac-M[Fe(CO)₃I₃]_n were prepared and structurally determined. Decomposing of these salts with CO-release in several media, e.g., DMSO, D₂O, saline, and PBS was investigated by FTIR spectroscopy. Related kinetics were estimated and abided by the first‐order model. Compared to the analogues of organic aminiums, diverse stability of the inorganics is observed in the DMSO and aqueous solution probably due to the distinguish solvation effect of the inorganic cations against the aminiums. Additionally, the CO-release of the fac-[Fe(CO)₃I₃]^- anion can be manipulated by the precise addition of iodine and response to a basic media. Cytotoxicity assessments of the five salts reveal severe toxicity against a bladder cancer cell line (RT112). The toxicity of them is lower than that for the analogues of organic aminiums. Since the toxicity is related to the decomposition of the anion, their inhibited cytotoxicity is attributed to a sluggish CO-release of the inorganic salts against the aminium salts^[33]. Given together, varying the counter ion of fac-[Fe(CO)₃I₃]^- anion from an organic aminium to an inorganic cation unambiguously affects the stability thus the cytotoxicity of the anion.

  To probe a mechanistic insight into the cytotoxicity of the anion, a set of experiments were performed to elucidate the generation of iodine radicals and released CO on the cellular deaths of the bladder cancer cells. Firstly, a dose-dependence on the anion for the iodine radical production was established from the TMB assay. Subsequently, both the produced radicals and CO from the decomposition of the anion were proved to contribute to its cytotoxicity as verified from a free radical scavenger, Tempol-involved MTT assay. Moreover, a distinguished increment in the ROS level for the mitochondria following the treatment of the anion was proved from the MitoSOX kit, whereas mitochondrial fission was evidenced from the CLSM microscopy. On the other hand, the protein expression of the oxidative stress response and mitophagy-related protein, Parkin was significantly upregulated, indicative of the immunoblotting analysis. Particularly, a ferroptosis should be induced by the anion as verified by the ferroptosis inhibitor assays. Therefore, a mechanistic understanding of the cytotoxicity of the fac-[Fe(CO)₃I₃]^- anion is proposed, which stimulates the decomposing of the anion with the production of radicals and CO and thus impacts mitochondria-related activities such as dynamics and fission, evoke a ferroptosis pathway, to further lead severe cellular damage even cell death.

  Supporting information is available at http://www.wjhxxb.cn

  
  
• 1. [1]

     Kim H P, Ryter S W, Choi A M K. CO as a cellular signaling molecule[J]. Annu. Rev. Pharmacol. Toxicol., 2006, 46:  411-449. doi: 10.1146/annurev.pharmtox.46.120604.141053

  2. [2]

     Tsuchihashi S, Busuttil R W, Kupiec-Weglinski J W. Heme oxygenase system//Jean-Franç ois D, Pierre-Alain C, Christian T, Rolf G. Signaling pathways in liver diseases. Berlin, Heidelberg: Springer Berlin Heidelberg, 2005: 291-298

  3. [3]

     Ling K, Men F, Wang W C, Zhou Y Q, Zhang H W, Ye D W. Carbon monoxide and its controlled release: Therapeutic application, detection and development of carbon monoxide-releasing molecules (CO-RMs)[J]. J. Med. Chem., 2017, 61(7):  2611-2635.

  4. [4]

     Motterlini R, Otterbein L E. The therapeutic potential of carbon monoxide[J]. Nat. Rev. Drug Discov., 2010, 9(9):  728-743. doi: 10.1038/nrd3228

  5. [5]

     Romao C C, Blä ttler W A, Seixas J D, Bernardes G J L. Developing drug molecules for therapy with carbon monoxide[J]. Chem. Soc. Rev., 2012, 41(9):  3571-3583. doi: 10.1039/c2cs15317c

  6. [6]

     Yang X X, Lu W, Hopper C P, Ke B W, Wang B H. Nature's marvels endowed in gaseous molecules I: Carbon monoxide and its physiological and therapeutic roles[J]. Acta Pharm. Sin. B, 2021, 11(6):  1434-1445. doi: 10.1016/j.apsb.2020.10.010

  7. [7]

     Pena A C, Pamplona A. Heme oxygenase-1, carbon monoxide, and malaria - the interplay of chemistry and biology[J]. Coord. Chem. Rev., 2022, 453:  214285. doi: 10.1016/j.ccr.2021.214285

  8. [8]

     Choi H I, Zeb A, Kim M S, Rana I, Khan N, Qureshi O S, Lim C W, Park J S, Gao Z G, Maeng H J, Kim J K. Controlled therapeutic delivery of CO from carbon monoxide-releasing molecules (CORMs)[J]. J. Control. Release, 2022, 350:  652-667. doi: 10.1016/j.jconrel.2022.08.055

  9. [9]

     Zhou Y Z, Yang T, Liang K, Chandrawati R. Metal-organic frameworks for therapeutic gas delivery[J]. Adv. Drug Deliv. Rev., 2021, 171:  199-214. doi: 10.1016/j.addr.2021.02.005

  10. [10]

      Wegiel B, Hanto D W, Otterbein L E. The social network of carbon monoxide in medicine[J]. Trends Mol. Med, 2013, 19(1):  3-11. doi: 10.1016/j.molmed.2012.10.001

  11. [11]

      Otterbein L E, Foresti R, Motterlini R. Heme oxygenase-1 and carbon monoxide in the heart: The balancing act between danger signaling and pro-survival[J]. Circ. Res., 2016, 118(12):  1940-1959. doi: 10.1161/CIRCRESAHA.116.306588

  12. [12]

      Motterlini R, Clark J E, Foresti R, Sarathchandra P, Mann B E, Green C J. Carbon monoxide-releasing molecules - characterization of biochemical and vascular activities[J]. Circ. Res., 2002, 90(2):  E17-E24.

  13. [13]

      Jiang X J, Xiao Z Y, Zhong W, Liu X M. Brief survey of diiron and monoiron carbonyl complexes and their potentials as CO-releasing molecules (CORMs)[J]. Coord. Chem. Rev., 2021, 429:  213634. doi: 10.1016/j.ccr.2020.213634

  14. [14]

      Fairlamb I J S, Lynam J M. Chapter 7 - carbon monoxide-releasing molecules: design principles inspired by mechanism, enabling activity to be controlled and tuned//Hirao T, Moriuchi T[J]. Advances in Bioorganometallic Chemistry. Amsterdam: Elsevier, 2019, :  137-154.

  15. [15]

      Ford P C. Metal complex strategies for photo-uncaging the small molecule bioregulators nitric oxide and carbon monoxide[J]. Coord. Chem. Rev., 2018, 376:  548-564. doi: 10.1016/j.ccr.2018.07.018

  16. [16]

      Mann B E. 3.29 - Signaling molecule delivery (CO)//Reedijk J, Poeppelmeier K. Comprehensive Inorganic Chemistry Ⅱ (Second Edition). Amsterdam: Elsevier, 2013: 857-876

  17. [17]

      Rimmer R D, Pierri A E, Ford P C. Photochemically activated carbon monoxide release for biological targets. Toward developing air-stable photoCORMs labilized by visible light[J]. Coord. Chem. Rev., 2012, 256(15/16):  1509-1519.

  18. [18]

      Abeyrathna N, Washington K, Bashur C, Liao Y. Nonmetallic carbon monoxide releasing molecules (CORMs)[J]. Org. Biomol. Chem., 2017, 15(41):  8692-8699. doi: 10.1039/C7OB01674C

  19. [19]

      Ji X Y, Wang B H. Strategies toward organic carbon monoxide prodrugs[J]. Acc. Chem. Res., 2018, 51(6):  1377-1385. doi: 10.1021/acs.accounts.8b00019

  20. [20]

      Nakae T, Hirotsu M, Nakajima H. CO Release from N, C, S-Pincer iron(Ⅲ) carbonyl complexes induced by visible-to-NIR light irradiation: Mechanistic insight into effects of axial phosphorus ligands[J]. Inorg. Chem., 2018, 57(14):  8615-8626. doi: 10.1021/acs.inorgchem.8b01407

  21. [21]

      Ou J, Zheng W H, Xiao Z Y, Yan Y P, Jiang X J, Dou Y, Jiang R, Liu X M. Core-shell materials bearing iron(Ⅱ) carbonyl units and their CO-release via an upconversion process[J]. J. Mater. Chem. B, 2017, 5(41):  8161-8168. doi: 10.1039/C7TB01434A

  22. [22]

      Sitnikov N S, Li Y C, Zhang D F, Yard B, Schmalz H G. Design, synthesis, and functional evaluation of CO-releasing molecules triggered by penicillin G amidase as a model protease[J]. Angew. Chem. Int. Ed., 2015, 54(42):  12314-12318. doi: 10.1002/anie.201502445

  23. [23]

      Jiang X J, Chen L M, Wang X, Long L, Xiao Z Y, Liu X M. Photoinduced carbon monoxide release from half-sandwich iron(Ⅱ) carbonyl complexes by visible irradiation: Kinetic analysis and mechanistic investigation[J]. Chem.-Eur. J., 2015, 21(37):  13065-13072. doi: 10.1002/chem.201501348

  24. [24]

      Long L, Jiang X J, Wang X, Xiao Z Y, Liu X M. Water-soluble diiron hexacarbonyl complex as a CO-RM: Controllable CO-releasing, releasing mechanism and biocompatibility[J]. Dalton Trans., 2013, 42(44):  15663-15669. doi: 10.1039/c3dt51281a

  25. [25]

      Romanski S, Ruecker H, Stamellou E, Guttentag M, Neudoerfl J M, Alberto R, Amslinger S, Yard B, Schmalz H G. Iron dienylphosphate tricarbonyl complexes as water-soluble enzyme-triggered CO-releasing molecules (ET-CORMs)[J]. Organometallics, 2012, 31(16):  5800-5809. doi: 10.1021/om300359a

  26. [26]

      Atkin A J, Fairlamb I J S, Ward J S, Lynam J M. CO Release from norbornadiene iron(0) tricarbonyl complexes: Importance of ligand dissociation[J]. Organometallics, 2012, 31(16):  5894-5902. doi: 10.1021/om300419w

  27. [27]

      Jackson C S, Schmitt S, Dou Q P, Kodanko J J. Synthesis, characterization, and reactivity of the stable iron carbonyl complex[Fe(CO)(N4Py)](ClO₄)₂: Photoactivated carbon monoxide release, growth inhibitory activity, and peptide ligation[J]. Inorg. Chem., 2011, 50(12):  5336-5338. doi: 10.1021/ic200676s

  28. [28]

      Fairlamb I J S, Lynam J M, Moulton B E, Taylor I E, Duhme-Klair A K, Sawle P, Motterlini R. η¹-2-pyrone metal carbonyl complexes as CO-releasing molecules (CO-RMs): A delicate balance between stability and CO liberation[J]. Dalton Trans., 2007, 33:  3603-3605.

  29. [29]

      Schlawe D, Majdalani A, Velcicky J, Heßler E, Wieder T, Prokop A, Schmalz H G. Iron-containing nucleoside analogues with pronounced apoptosis-inducing activity[J]. Angew. Chem. Int. Ed., 2004, 43(13):  1731-1734. doi: 10.1002/anie.200353132

  30. [30]

      姜秀娟, 肖志音, 龙莉, 陈丽梅, 张立秋, 刘小明. 水溶性二铁六羰基配合物与生物相关分子的相互作用及其促进一氧化碳释放[J]. 无机化学学报, 2022,38,(5): 913-920. JIANG X J, XIAO Z Y, LONG L, CHEN L M, ZHANG L Q, LIU X M. Interactions of a water-soluble diiron hexacarbonyl complex with biologically relevant molecules and their promotion in CO-Release[J]. Chinese J. Inorg. Chem., 2022, 38(5):  913-920.

  31. [31]

      罗佳彬, 郭晋忠, 肖志音, 钟伟, 李雪明, 刘小明. 含二齿膦配体的二羰基铁化合物的制备及光诱导释放一氧化碳性能[J]. 无机化学学报, 2022,38,(7): 1241-1251. LUO J B, GUO J Z, XIAO Z Y, ZHONG W, LI X M, LIU X M. Preparation of dicarbonyl iron compounds with a bidentate phosphine and their CO release behaviors upon irradiation[J]. Chinese J. Inorg. Chem., 2022, 38(7):  1241-1251.

  32. [32]

      张俊蝶, 姜秀娟, 肖志音, 陈丽梅, 王雪梅, 刘小明. CO释放体系中沉淀的预防及利用配体交换反应调控CO释放速率[J]. 无机化学学报, 2022,38,(8): 1593-1600. ZhANG J D, JIANG X J, XIAO Z Y, CHEN L M, WANG X M, LIU X M. Preventing CO-releasing systems from forming precipitates and tuning CO-releasing rate via ligand exchange reaction[J]. Chinese J. Inorg. Chem., 2022, 38(8):  1593-1600.

  33. [33]

      Yang X Q, Jin J, Guo Z M, Xiao Z Y, Chen N W, Jiang X J, He Y, Liu X M. The monoiron anion fac-[Fe(CO)₃I₃]^- and its organic aminium salts: Their preparation, CO-release, and cytotoxicity[J]. New J. Chem., 2020, 44(25):  10300-10308. doi: 10.1039/D0NJ01182G

  34. [34]

      Xiao Z Y, Jiang R, Jin J, Yang X Q, Xu B Y, Liu X M, He Y B, He Y. Diiron(Ⅱ) pentacarbonyl complexes as CO-releasing molecules: Their synthesis, characterization, CO-releasing behaviour and biocompatibility[J]. Dalton Trans., 2019, 48(2):  468-477. doi: 10.1039/C8DT03982H

  35. [35]

      Guo Z M, Jin J, Xiao Z Y, Chen N W, Jiang X J, Liu X M, Wu L F, He Y, Zhang S H. Four iron(Ⅱ) carbonyl complexes containing both pyridyl and halide ligands: Their synthesis, characterization, stability, and anticancer activity[J]. Appl. Organomet. Chem., 2021, 35(1):  e6045. doi: 10.1002/aoc.6045

  36. [36]

      Guo J Z, Guo Z M, Xiao Z Y, Jin J, Yang X Q, He Y, Liu X M. Further exploration of the reaction between cis-[Fe(CO)₄I₂] and alkylamines: An aminium salt of fac-[Fe(CO)₃I₃]^- or an amine-bound complex of fac-[Fe(CO)₃I₂(NH₂R)]?[J]. Organomet. Chem., 2021, 35(8):  .

  37. [37]

      Xiao Z Y, Wei Z H, Long L, Wang Y L, Evans D J, Liu X M. Diiron carbonyl complexes possessing a {Fe(Ⅱ)Fe(Ⅱ)} core: Synthesis, characterisation, and electrochemical investigation[J]. Dalton Trans., 2011, 40(16):  4291-4299. doi: 10.1039/c0dt01465f

  38. [38]

      Hieber W, Bader G. Reaktionen und Derivate des Eisencarbonyls, II.: Neuartige Kohlenoxyd-Verbindungen von Eisenhalogeniden[J]. Ber. Dtsch. Chem. Ges., 1928, 61(8):  1717-1722. doi: 10.1002/cber.19280610825

  39. [39]

      Pankowski M, Bigorgne M. Syntheses and isomerization of halocarbonyliron complexes: [FeX(CO)_5-nL_n]⁺, FeX₂(CO)_4-nL_n and[FeX₃(CO)₃]^- (L=PMe₃; n=1, 2, 3; X=Cl, Br, I)[J]. J. Organomet. Chem., 1977, 125(2):  231-252. doi: 10.1016/S0022-328X(00)89443-7

  40. [40]

      Szabo C. Gasotransmitters in cancer: From pathophysiology to experimental therapy[J]. Nat. Rev. Drug Discov., 2016, 15(3):  185-203. doi: 10.1038/nrd.2015.1

  41. [41]

      Wang X S, Zeng J Y, Li M J, Li Q R, Gao F, Zhang X Z. Highly stable iron carbonyl complex delivery nanosystem for improving cancer therapy[J]. ACS Nano, 2020, 14(8):  9848-9860. doi: 10.1021/acsnano.0c02516

  42. [42]

      Piantadosi C A. Carbon monoxide, reactive oxygen signaling, and oxidative stress[J]. Free Radic. Biol. Med., 2008, 45(5):  562-569. doi: 10.1016/j.freeradbiomed.2008.05.013

  43. [43]

      Zuckerbraun B S, Chin B Y, Bilban M, d'Avila J d C, Rao J, Billiar T R, Otterbein L E. Carbon monoxide signals via inhibition of cytochrome c oxidase and generation of mitochondrial reactive oxygen species[J]. Faseb J., 2007, 21(4):  1099-1106. doi: 10.1096/fj.06-6644com

  44. [44]

      Stockwell B R, Jiang X J. The chemistry and biology of ferroptosis[J]. Cell Chem. Biol., 2020, 27(4):  365-375. doi: 10.1016/j.chembiol.2020.03.013

  45. [45]

      Jiang X J, Stockwell B R, Conrad M. Ferroptosis: Mechanisms, biology and role in disease[J]. Nat. Rev. Mol. Cell Biol., 2021, 22(4):  266-282. doi: 10.1038/s41580-020-00324-8

  46. [46]

      Dixon S J, Lemberg K M, Lamprecht M R, Skouta R, Zaitsev E M, Gleason C E, Patel D N, Bauer A J, Cantley A M, Yang W S, Morrison B, Stockwell B R. Ferroptosis: An iron-dependent form of nonapoptotic cell death[J]. Cell, 2012, 149(5):  1060-1072. doi: 10.1016/j.cell.2012.03.042

  47. [47]

      Angeli J P F, Schneider M, Proneth B, Tyurina Y Y, Tyurin V A, Hammond V J, Herbach N, Aichler M, Walch A, Eggenhofer E, Basavarajappa D, Rådmark O, Kobayashi S, Seibt T, Beck H, Neff F, Esposito I, Wanke R, Förster H, Yefremova O, Heinrichmeyer M, Bornkamm G W, Geissler E K, Thomas S B, Stockwell B R, O'Donnell V B, Kagan V E, Schick J A, Conrad M. Inactivation of the ferroptosis regulator Gpx4 triggers acute renal failure in mice[J]. Nat. Cell Biol., 2014, 16(12):  1180-1191. doi: 10.1038/ncb3064

• 

  Scheme 1  Synthetic route for the fac-[Fe(CO)₃I₃]^- anion with an aminium cation or an inorganic cation (1-5) from the universal precursor cis-[Fe(CO)₄I₂]

  下载: 全尺寸图片 幻灯片
  

  Figure 1  FTIR spectra of the salts 1-5 in DCM solvent

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Infrared spectral variation of the fac-[Fe(CO)₃I₃]^- anion for salts 1 (a), 2 (b), 3 (c), 4 (d), 5 (e) in saline and the corresponding kinetic plots (f), obtained by linear fitting of the ln A values at 2 090 cm^-1 against reaction time

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Infrared spectral variation of the fac-[Fe(CO)₃I₃]^- anion at the presence of various concentrations of NaI: 0.05 (a), 0.1 (b), 0.2 (c), 0.3 (d) and 0.4 (e) mol·L^-1 in saline and the corresponding kinetic plots (f), obtained by linear fitting of the ln A values at 2 090 cm^-1 against reaction time

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Plot of the halftime of the fac-[Fe(CO)₃I₃]^- anion (1, 16 mmol·L^-1) in D₂O solvent against the concentration of NaI

  Linear relationship: t_1/2=677.05cNaI + 49.85, R²=0.99.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Representatively kinetic plots of the fac-[Fe(CO)₃I₃]^- anion in PBS solution with different pH values

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Trypan blue staining of the bladder cancer cells incubated with the fac-[Fe(CO)₃I₃]^- anion in various concentrations for 4 h

  下载: 全尺寸图片 幻灯片
  

  Figure 7  UV-Vis spectra (a) of TMB in HAc-NaAc buffer (pH=4.5) with various additions of salt 1, and the plot (b) of the absorption of oxidized TMB versus the concentration of salt 1

  Determination of the production of iodine radical with a dose-dependent effect of the fac-[Fe(CO)₃I₃]^- anion by a TMB assay.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  Viabilities of RT112 cells treated with various amounts of the fac-[Fe(CO)₃I₃]^- anion for 24 h in the absence and presence of Tempol (1 mmol·L^-1)

  下载: 全尺寸图片 幻灯片
  

  Figure 9  CLSM images of the staining bladder cancer cells incubated with the absence (a) and presence of fac-[Fe(CO)₃I₃]^- anion (15 μmol·L^-1) (b) for 24 h

  Inset: the corresponding enlarged images.

  下载: 全尺寸图片 幻灯片
  

  Figure 10  Immunoblotting analysis (Western blotting) of Parkin protein in bladder cancer cells following the treatment of fac-[Fe(CO)₃I₃]^- anion for 24 h

  下载: 全尺寸图片 幻灯片
  

  Figure 11  Inhibitors Ferrostatin-1 (1 μmol·L^-1) and Liproxstatin-1 (1 μmol·L^-1) suppress heavily the ferroptosis of RT112 cells driven by the fac-[Fe(CO)₃I₃]^- anion (MTT, 24 h)

  下载: 全尺寸图片 幻灯片

  Table 1.  Selective characterized data of salts 1-5

  Salt/Mn+	νCO[a] / cm-1	λmax[b] / nm	δC[c]	m/z of the anion[d]
  1/Na+	2 087, 2 037	288, 359	217.5	520.628 8
  2/K+	2 085, 2 034	287, 359	216.4	520.623 4
  3/Mg2+	2 085, 2 034	290, 358	218.3	520.622 2
  4/Ca2+	2 086, 2 035	290, 359	217.0	520.621 7
  5/NH4+	2 087, 2 037	286, 361	217.4	520.613 7
  [a] FTIR data of the carbonyl group in DCM; [b] UV-Vis data of the complexes in the methanol; [c] 13C NMR of the carbonyl group in the acetone-d6; [d] Mass analysis in methanol. Calculated m/z signal for the anion: 520.632 5.

  下载: 导出CSV

  Table 2.  CO-releasing kinetics of the fac-[Fe(CO)₃I₃]^- anion for salts 1-5 in DMSO, D₂O, and saline media

  Salt/Mn+	kobs / min-1				t1/2 / min
  	DMSO	D2O	Saline		DMSO	D2O	Saline
  1/Na+	(10.4±0.1)×10-3	(14.3±0.1)×10-3	(11.3±0.3)×10-3		66±1	48±1	61±2
  2/K+	(10.2±0.2)×10-3	(13.6±0.9)×10-3	(16.0±1.4)×10-3		68±1	52±4	44±4
  3/Mg2+	(11.2±0.2)×10-3	(10.5±0.2)×10-3	(11.1±0.3)×10-3		62±1	66±1	62±2
  4/Ca2+	(7.6±0.2)×10-3	(11.2±0.7)×10-3	(15.1±0.9)×10-3		91±3	62±4	46±3
  5/NH4+	(10.7±0.2)×10-3	(14.6±0.5)×10-3	(11.9±0.7)×10-3		65±1	48±2	58±4
  Data were presented as "mean±SD" (n=3).

  下载: 导出CSV

  Table 3.  Kinetic data of degradation of salt 1 at various concentrations of iodine ions

  Parameter	cNaI / (mol·L-1)
  	0	0.05	0.1	0.2	0.3	0.4
  kobs / min-1	(14.3±0.1)×10-3	(8.0±0.1)×10-3	(5.5±0.2)×10-3	(3.9±0.1)×10-3	(2.9±0.1)×10-3	(2.2±0.1)×10-3
  t1/2 / min	48±1	87±1	127±5	178±5	239±8	330±16
  Data were presented as "mean±SD" (n=3).

  下载: 导出CSV

  Table 4.  Kinetic data of the degradation of fac-[Fe(CO)₃I₃]^- anion in PBS solution at different pH environments

  Parameter	pH
  	5.7	6.5	7.4	8.0
  kobs / min-1	(12.1±0.1)×10-3	(13.9±0.1)×10-3	(31.0±0.2)×10-3	(28.3±0.1)×10-3
  t1/2 / min	57±1	50±1	22±1	24±1
  Data were presented as "mean±SD" (n=3).

  下载: 导出CSV

  Table 5.  IC₅₀ values of salts 1-5 and selected complexes against RT112 cells^*

  Complex	IC50 / (μmol·L-1)	Reference
  Na[Fe(CO)3I3] (1)	24.8±1.3	This work
  K[Fe(CO)3I3] (2)	26.8±0.8	This work
  Mg[Fe(CO)3I3]2 (3)	31.7±1.9	This work
  Ca[Fe(CO)3I3]2 (4)	33.3±2.4	This work
  NH4[Fe(CO)3I3] (5)	43.0±4.3	This work
  [EtNH3][Fe(CO)3I3]	15.4±1.1	[33]
  [PrNH3][Fe(CO)3I3]	20.0±1.0	[33]
  [Fe(CO)3I2(py)]	27.4±1.8	[35]
  [{Fe(CO)3I2}2(bipy)]	97.2±5.7	[35]
  [Fe(CO)3Br2(py)]	137.9±5.4	[35]
  [{Fe(CO)3Br2}2(bipy)]	468.7±16.1	[35]
  [Fe(CO)3I2(NH2Et)]	> 100	[36]
  *MTT, 24 h.

  下载: 导出CSV
•