Conformationally Preorganized High-Affinity Ligands for Copper Biology with Hinged and Rigid Thiophene Backbones

Abstract

Copper-selective ligands are essential tools for probing the affinity of cuproproteins or manipulating the cellular copper availability. They also harbor significant potential as anti-angiogenic agents in cancer therapy or as therapeutics to combat copper toxicity in Wilson’s disease. To achieve the high Cu(I) affinities required for competing effectively with cellular cuproproteins, we recently devised a ligand design based on phosphine sulfide-stabilized phosphine (PSP) donor motifs. Building on this design strategy, we integrated two PSP donors within preorganized ligand architectures composed of either a hinged bithiophene backbone (bithipPS) or a single rigid thiophene bridge (thipPS). Extensive characterization based on X-ray crystal structures, solution NMR data, spectrophotometric titrations, and electrochemical studies established that bithipPS adapts well to the coordination preferences of Cu(I) to form a discrete air-stable mono-nuclear Cu(I) complex with a dissociation constant of 4 zM. In contrast, the wider bite angle of thipPS introduces some strain upon Cu(I) coordination to yield an almost 10-fold lower affinity with a K_d of 35 zM. As revealed by ICP-MS and two-photon excitation microscopy studies with the Cu(I)-selective fluorescent probe crisp-17, both ligands are effective at removing cellular copper from live mouse fibroblasts with rapid kinetics. Altogether, the stability and redox properties of PSP-ligand Cu(I) complexes can be effectively tuned by judicious balancing of their geometrical preorganization and conformational flexibility.

Graphical Abstract

To design chelators with subattomolar dissociation constants suitable for probing copper biology, we integrated two phosphine sulfide-stabilized phosphine donor motifs into preorganized ligand architectures composed of either a hinged bithiophene (bithipPS) or a single rigid thiophene ring system (thipPS). Both ligands form air-stable Cu(I) complexes with dissociation constants of 4 and 35 zM, respectively, while maintaining similar redox properties. Applied to mouse fibroblasts, the chelators selectively removed copper without affecting other divalent metal ions, thus rendering them well suited for probing copper in complex biological systems.

INTRODUCTION

Copper (Cu) is a trace element that plays critical roles in biological processes where it functions as a cofactor in enzymes involved in cellular respiration, antioxidant defense, extracellular matrix formation, or iron acquisition.¹ Increasing evidence suggests that cellular copper concentrations are maintained at low attomolar levels through a polydisperse buffer composed of proteins and other biomolecules that form tight yet kinetically labile complexes with Cu(I),^{2, 3} the predominant oxidation state within the reducing intracellular environment. Unraveling the molecular mechanisms that regulate cellular copper levels requires selective ligands for probing the affinity of cuproproteins and for manipulating the cellular Cu(I) availability. Moreover, copper-selective ligands harbor significant potential as cancer therapeutics^4-9 and for combating copper toxicity in Wilson disease^10-13 and other disorders associated with copper overload.^{14, 15} However, the high affinity of cuproproteins poses significant challenges in designing ligands that can effectively compete for Cu(I) binding.³ To address this challenge, we recently devised a series of high-affinity Cu(I) chelators based on phosphine sulfide stabilized phosphine (PSP) binding motifs (Chart 1).¹⁶ The electron-withdrawing nature of the strongly polarized PS-moiety stabilizes the phosphine donor against protonation and oxidation, while maintaining a strong affinity towards Cu(I). For example, PSP-1 can be recrystallized from boiling water in ambient air and binds Cu(I) with an apparent dissociation constant of 0.8 fM.¹⁶ By combining two PSP motifs in an EDTA-like architecture, we achieved a low-zeptomolar Cu(I) dissociation constant with PSP-2 (K_d = 10 zM). With additional conformation preorganization, dissociation constants in the subzeptomolar range can even be realized. For example, the phenylene- or naphthalene-bridged ligands phenPS and naphPS feature K_d’s of 0.8 zM and 0.6 zM, respectively.¹⁷

Chart 1:

High-affinity Cu(I) Ligands with Phosphine Sulfide-Stabilized Phosphine (PSP) Donors

While molecular preorganization minimizes the entropic penalty associated with the rigidification of the ligand upon metal binding, it may also introduce strain energy if the coordination preferences of the metal center cannot be appropriately satisfied. Ligands with a rigid but hinged backbone architecture may offer a suitable compromise where the axial conformational flexibility of the hinge can accommodate the coordination preferences of the metal center while still maintaining a high degree of structural preorganization.^18-21 Based on this design concept, we devised the ligand bithipPS, in which we integrated two PSP donors into a hinged bithiophene backbone (Chart 1).

For comparison, we also describe thipPS comprised of a single rigid thiophene moiety that bridges the two PSP donors. Compared to phenPS, the 5-membered thiophene ring offers a wider bite-angle, which is expected to reduce the Cu(I) affinity while maintaining a high degree of structural preorganization to discourage formation of oligomeric species as observed for PSP-2.¹⁶ Both ligands were extensively characterized in terms of their coordination chemistry, both in solution and in the solid state, their Cu(I) affinity, redox chemistry, metal ion selectivity, and their ability to remove copper from live mouse fibroblasts.

EXPERIMENTAL SECTION

Synthesis.

[(MCL-2)Cu(I)]PF₆ (Chart S1) was prepared as previously reported.²² All other reagents and materials were from standard commercial sources and used as received. NMR spectra were recorded on a Bruker Avance 500 IIIHD spectrometer equipped with a Prodigy cryoprobe at 298 K except when specified otherwise. Operating frequencies were 500 MHz (¹H) and 203 MHz (³¹P). ¹H chemical shifts are reported relative to 1% (v/v) TMS, and ³¹P spectra were reported relative to 85% D₃PO₄ (internal sealed capillary tube). EI-MS: selected peaks, m/z (intensity).

Bis(dimethylphosphorothioylmethyl)phosphine (3).

An oven-dried 2-necked 250 mL round bottom flask equipped with a stirring bar, rubber septum, and a bubbler was purged with argon. Under a steady stream of argon, 100 mL of Et₂O were added via cannula, followed by methyl dichlorophosphite (2.0 g, 15 mmol, weighed in a syringe). The reaction mixture was allowed to cool to −78°C in an acetone/dry ice bath. In a separate 100 mL round bottom flask equipped with a magnetic stir bar was added trimethylphosphine sulfide (3.24 g, 30 mmol) under argon atmosphere, followed by anhydrous THF (~ 70 mL) and TMEDA (4.5 mL, 30 mmol) under vigorous stirring. The resulting solution mixture was cooled to −78°C in an acetone/dry ice bath to afford a white suspension. Under vigorous stirring, n-BuLi (2.5 M in hexanes, 12 mL, 30 mmol) was added dropwise. The resulting clear homogenous solution was transferred to the vigorously stirred methyl dichlorophosphite in anhydrous Et₂O via cannula. A white precipitate started to form almost immediately. After 30 min, the reaction flask was removed from the cooling bath and butanone (1 mL) was added to quench any residual organolithium species. The reaction mixture was diluted with dry Et₂O to 250 mL. The precipitated product methyl bis(dimethylphosphorothioylmethyl)phosphinite 2 was filtered, dried under reduced pressure, and used directly for the next step without further purification.

To an oven-dried 100 mL round bottom flask equipped with a magnetic stir bar was added methyl bis(dimethylphosphorothioylmethyl)phosphinite 2 (2.12 g, ~ 7.68 mmol). The reaction flask was vacuumed and purged with argon, followed by the addition of anhydrous DCM (35 mL). The resulting suspension was cooled to 0°C in an ice bath. Over a period of 10 mins, DIBAL-H (25% in hexanes, ~ 1.2 M, 9.6 mL, 1.5 molar equiv.) was added in small portions. The resulting clear solution was allowed to stir for 30 min at 0°C, and then 1 hour at room temperature. The resulting reaction mixture was poured into a vigorously stirred solution of citric acid (2.95 g, 2.0 molar equiv.) in 12 mL of deionized water (caution: exothermic process with gas evolution). The mixture was diluted with 12 mL of DCM and 25 mL of 1 M citrate buffer (pH ~ 5.5). The combined organic layers were dried over anhydrous Na₂SO₄ and diluted with cyclohexane. The volatiles were removed under reduced pressure, and the crude product was purified via column chromatography with a MTBE-DCM mixture to yield the pure secondary phosphine as a fluffy colorless solid (1.23 g, 33% yield over two steps). ¹H NMR (CDCl₃, 500 MHz) δ 1.84 (d, J = 12.8 Hz, 6H), 1.86 (d, J = 12.8 Hz, 6H), 2.43–2.58 (m, 4H), 3.99 (dtp, J = 214.7, 18.2, 6.8 Hz, 1H). ³¹P{1H} NMR (CDCl₃, 203 MHz) δ –101.5 (t, J = 41.3 Hz), 37.15 (d, J = 41.3 Hz). MS (EI) m/z 246 (M⁺, 43), 153 (33), 139 (49), 138 (40), 93 (57), 75 (100), 61 (19). HRMS (EI) m/z calculated for C₆H₁₇P₃S₂ (M⁺) 245.9985, found 245.9985.

((Thiophene-2,3-diylbis(phosphanetriyl))tetrakis(methylene))tetrakis(dimethylphosphinesulfide), thipPS (5).

An oven-dried 5 mL round bottom flask equipped with a magnetic stir bar was charged with bis(dimethylphosphorothioylmethyl)phosphine 3 (50 mg, 0.20 mmol), 2,3-dibromothiophene (24.6 mg, 0.10 mmol), Pd(OAc)₂ (2.3 mg, 0.010 mmol), 1,1′-bis(diisopropylphosphino)ferrocene (5.1 mg, 0.012 mmol), and anhydrous cesium carbonate (66 mg, 0.20 mmol). The flask was connected to a reflux condenser, capped with a rubber septum, and flushed with a stream of argon. Anhydrous toluene (2.0 mL) was added via a syringe, and the resulting orange heterogenous mixture was refluxed for 48 hours under argon. After cooling to room temperature, the reaction mixture was washed with deionized water, and the aqueous phase was extracted with DCM (3 x 10 mL). The combined organic layers were dried with anhydrous Na₂SO₄, filtered, and concentrated under reduced pressure. The crude product was purified via column chromatography on silica gel using MTBE-DCM (1:9 to 1:5). The purified product was dissolved in boiling DCM and recrystallized in MTBE to afford colorless needle-like crystal clusters (40 mg, 40% yield). ¹H NMR (500 MHz, CD₂Cl₂) δ 0.53 - 0.83 (m, 24H), 1.35 - 1.64 (m, 4H), 1.79 - 1.97 (m, 4H), 6.14 (ddd, J = 5.1, 2.4, 0.9 Hz, 1H), 6.59 (dd, J = 5.1, 1.0 Hz, 1H). ¹H NMR (500 MHz, DMSO-d₆) δ 1.59 - 1.87 (m, 24H), 2.67 - 2.83 (m, 4H), 2.97 (t, J = 14.0 Hz, 2H), 3.07 (t, J = 13.9 Hz, 2H), 7.57 (dd, J = 5.2, 2.3 Hz, 1H), 8.00 (d, J = 5.1 Hz, 1H). ³¹P {¹H} NMR (203 MHz, CD₂Cl₂) δ −63.83 - −61.95 (m, 1P), −60.57 - −58.68 (m, 1P), 35.24 - 36.35 (m, 4P). MS (ESI) m/z 573 (100, [M+H]⁺). HRMS (ESI) m/z calculated for C₁₆H₃₅P₆S₅ ([M+H]⁺) 572.9763, found 572.9747.

(([2,2'-Bithiophene]-3,3'-diylbis(phosphanetriyl))tetrakis(methylene))tetrakis(dimethylphosphinesulfide), bithipPS (7).

An oven-dried round bottom flask equipped with a magnetic stir bar was charged with bis(dimethylphosphorothioylmethyl)phosphine 3 (76 mg, 0.31 mmol), 3,3’-dibromo-2,2’-bithiophene (50 mg, 0.15 mmol), Pd(OAc)₂ (3.5 mg, 0.015 mmol), 1,1′-bis(diisopropylphosphino)ferrocene (7.7 mg, 0.019 mmol), and anhydrous cesium carbonate (101 mg, 0.31 mmol). The flask was connected to a reflux condenser, capped with a rubber septum, and flushed with a stream of argon. Anhydrous toluene (3.0 mL) was added, and the resulting orange heterogenous mixture was refluxed under argon for 72 hours. After cooling to room temperature, the reaction mixture was washed with deionized water. The aqueous phase was extracted with DCM (3 x 20 mL), and the combined organic layers were dried over anhydrous Na₂SO₄, filtered, and concentrated under reduced pressure. The crude product was purified via column chromatography on silica gel using MTBE-DCM (gradient 1:3 to 1:1). The product was recrystallized from hexanes to afford a colorless crystalline solid (30 mg, 30% yield). ¹H NMR (500 MHz, CDCl₃) δ 1.80 (dd, J = 12.8, 6.8 Hz, 24H), 2.55 (t, J = 13.3 Hz, 4H), 3.11 (t, J = 13.3 Hz, 4H), 7.43 (dd, J = 5.3, 0.8 Hz, 2H), 7.52 (d, J = 5.3 Hz, 2H). ¹H NMR (500 MHz, DMSO-d₆) δ 1.64 (t, J = 13.8 Hz, 24H), 2.72 (t, J = 13.0 Hz, 4H), 3.03 (t, J = 14.1 Hz, 4H), 7.68 (d, J = 5.3 Hz, 2H), 7.84 (d, J = 5.3 Hz, 2H). ³¹P {¹H} NMR (203 MHz, CDCl₃) δ −59.89 - −58.65 (symmetric m, 2P), 35.00 - 36.35 (symmetric m, 4P). MS (ESI) m/z 655 (100, [M+H]⁺). HRMS (ESI) m/z calculated for C₂₀H₃₇P₆S₆ ([M+H]⁺) 654.9640, found 654.9651.

[Cu(I)thipPS]PF₆ (8).

To a solution of thipPS 5 (5.0 mg, 8.73 μmol) in DCM (0.5 mL) was added a solution of [Cu(CH₃CN)₄]PF₆ (3.55 mg, 9.52 μmol) in CH₃CN (0.5 mL). The solvent was slowly evaporated under a steady stream of argon in a room temperature water bath until the solution turned opaque. A few drops of CH₃CN were added to resolubilize all the solids. The resulting solution was allowed to stand until crystals formed. The crystals were filtered through a cotton plug, rinsed with MTBE, and dried under high vacuum overnight (5.5 mg, 80% yield). ¹H NMR (500 MHz, DMSO-d₆, 1.0 mM) δ 1.75 (dd, J = 13.4, 6.0 Hz, 12H), 1.92 (dd, J = 13.5, 5.1 Hz, 12H), 2.94 - 3.20 (m, 4H), 3.42 - 3.65 (m, 4H), 7.69 (dd, J = 4.9, 2.4 Hz, 1H), 8.30 (d, J = 4.9 Hz, 1H). ³¹P{¹H} NMR (203 MHz, DMSO-d₆, 1.0 mM) δ −143.46 (hept, J = 712.0 Hz, 1 P), −44.34 (br, s, 2P), 45.16 (br, s, 4P).

[Cu(I)bithipPS]PF₆ (9).

To a solution of bithipPS 7 (5.0 mg, 7.64 μmol) in DCM (1.0 mL) was added a solution of [Cu(CH₃CN)₄]PF₆ (3.10 mg, 8.32 μmol) in CH₃CN (0.85 mL) under vigorous stirring at room temperature. The solution was continuously stirred until all the solids dissolved. The solvent was slowly evaporated under a steady stream of argon, and a drop of EtOH was added. The solution was allowed to stand in the dark until crystals formed. The crystals were filtered through a cotton plug, rinsed with EtOH, and dried under high vacuum (4.9 mg, 74% yield). ¹H NMR (500 MHz, DMSO-d₆, 1.0 mM) δ 1.42 (dd, J = 67.0, 12.8 Hz, 6H), 1.86 (dd, J = 67.0, 12.8 Hz, 6H), 2.27 (br, s, 2H), 2.77 (br, s, 2H), 3.75 (br, s, 2H), 7.55 (d, J = 5.4 Hz, 2H), 7.86 (d, J = 5.3 Hz, 2H). ³¹P{¹H} NMR (203 MHz, DMSO-d6, 1.0 mM) δ −143.46 (hept, J = 712.0 Hz, 1P), −40.55 (br, s, 2P), 36.31 (br, s, 2P), 47.43 - 56.28 (m, 2P).

[Cu(I)thipPS]Cl (10).

To a suspension of commercially available CuCl (500 mg, ~ 5.0 mmol) in deionized water (6.0 mL) was slowly added conc. HCl with vigorous stirring until the green solid turned colorless. The blue aqueous layer was removed via filtration. The colorless solid was washed with EtOH and dried under high vacuum. Freshly purified CuCl (1.0 mg, 10.1 μmol) and thipPS 5 (5.78 mg, 10.1 μmol) were stirred rapidly in CH₃CN (1.0 mL) for 15 mins under argon atmosphere. The resulting colorless solution was concentrated to dryness under a stream of argon, and the resulting metal complex was dried under high vacuum. ³¹P NMR (203 MHz, 10% D₂O in H₂O, 1.0 mM NaH₂PO₄, 1.0 mM) δ −46.39 (br, s, 2P), 44.86 (br, s, 4P). MS (ESI) m/z 635 (100, M⁺). HRMS (ESI/Orbitrap) m/z calculated for C₁₆H₃₄P₆S₅Cu (M⁺) 634.8980, found 634.8980.

[Cu(I)bithipPS]Cl (11).

Freshly purified CuCl (1.0 mg, 10.1 μmol) and bithipPS 7 (6.61 mg, 10.1 μmol) were stirred rapidly in CH₃CN (2.0 mL) for 15 mins under an argon atmosphere. The resulting pale-yellow solution was concentrated to dryness under a stream of argon, and the product was dried under high vacuum. ³¹P NMR (203 MHz, 10% D₂O in H₂O, 1.0 mM NaH₂PO₄, 10 mM) δ −41.00 (br, s, 2P), 40.83 (br, s, 2P), 51.13 (br, s, 2P). MS (ESI) m/z 717 (100, M⁺). HRMS (ESI/Orbitrap) m/z calculated for C₂₀H₃₆P₆S₆Cu (M⁺) 716.8857, found 716.8859.

X-ray Structure Determination.

A suitable crystal was selected and mounted on a loop with paratone on a XtaLAB synergy-S diffractometer. The crystal was kept at a steady T = 100.0(1) K during data collection. The structure was solved with the SheIXT 2018/2 solution program using dual methods and by using Olex2 1.3-alpha as the graphical interface.²³ The model was refined with SheIXL 2018/3 using full matrix least squares minimization on F².^{24, 25}

Cu(I) Binding Stoichiometry.

An aqueous stock solution of bicinchoninic acid (BCA, Chart S1) was prepared by dissolving BCA to a concentration of 30 mM in 35 mM aqueous KOH. In a 1-cm path length quartz cuvette equipped with a magnetic stir bar, 15 μL of BCA stock solution was diluted into 2975 μL of PIPES buffer (10 mM PIPES, 0.1 M KCl, pH 7.0, 25°C), followed by CuSO₄ (2.5 μL from a 60 mM solution), and sodium ascorbate (7.5 μL from a 60 mM solution) sequentially. The absorption spectrum was recorded from 450 to 650 nm following a 1-minute mixing period after each addition. The solution was titrated with 5.0 μM aliquots of thipPS or bithipPS from a 2.0 mM stock solution in DMSO under vigorous stirring. Absorbance of the [BCA₂Cu(I)]³⁻ complex was plotted at 562 nm versus molar equivalents of thipPS or bithipPS to total Cu(I).

Molar Ratio Titration with [Cu(I)MCL-2]PF₆.

In a 1-cm path length quartz cuvette equipped with a magnetic stir bar, a 50 μM aqueous solution of thipPS was prepared by diluting the ligand from a 3.0 mM stock solution in DMSO into 3.0 mL of PIPES buffer (10 mM PIPES, 0.1 M KCl, pH 7.0, 25°C) under rapid stirring. A UV-vis absorption spectrum was recorded from 250 nm to 400 nm. [Cu(I)MCL-2]PF₆ (Chart S1) was titrated in 0.1 mol equiv. aliquots. An absorption spectrum was recorded after a 1-minute mixing period following each aliquot. The exact concentration of the thipPS was calibrated against [Cu(I)MCL-2]PF₆. The stock solution of bithipPS was prepared by dissolving the ligand in anhydrous DMF under ultrasonication. A 40 μM aqueous solution of bithipPS was prepared by diluting the ligand from a 3.0 mM stock solution in DMF into 3.0 mL PIPES buffer (10 mM PIPES, 0.1 M KCl, pH 7.0, 25°C) under rapid stirring. The exact concentration of the bithipPS was calibrated against [Cu(I)MCL-2]PF₆. Absorbance of DMF was subtracted to derive a more accurate spectrum.

Stability Constants of Cu(I) Complexes.

An aqueous stock solution (60 mM) of bathocuproinedisulfonic acid disodium salt hydrate (BCS, Chart S1) was prepared and calibrated by competition titration against MCL-1 (Chart S1), as previously reported.^{17, 22} BCS (25 μL from a 60 mM stock solution), sodium ascorbate (7.5 μL from a 60 mM stock solution), and CuSO₄ (1.5 μL from a 60 mM stock solution) were diluted sequentially into 2966 μL of PIPES buffer (10 mM PIPES and 0.1 M KCl, pH 7.0, 25°C) in a 1-cm path length quartz cuvette equipped with a magnetic stir bar. A UV-vis absorption spectrum was recorded from 400 to 600 nm following a 1-minute mixing period after each addition. The solution was titrated with 5.0 μM aliquots of thipPS from the stock solution. An ensuing absorption spectrum was acquired after a 5-minute equilibration period following each aliquot. The titration was conducted in triplicate, and each data set was analyzed by nonlinear least-squares fitting over the spectral range from 600 nm to 400 nm using the SPECFIT software package.²⁶

Electrochemistry.

The redox potentials of [thipPSCu(I)]PF₆ and [bithipPSCu(I)]PF₆ complexes were determined by cyclic voltammetry in a 0.1 M methanolic tetrabutylammonium hexafluorophosphate (Bu₄NPF₆) solution. All measurements were conducted under argon atmosphere in a single-compartment cell with a glassy carbon working electrode, a Pt counter electrode, and a nonaqueous Ag/AgNO₃ reference electrode (10 mM AgNO₃ in CH₃CN). The half-wave potentials (E_1/2) were referenced to ferrocenium as the external standard.²⁷ Measurements were performed with a scan rate of 20 mV/s at 25°C.

ICP-MS Analysis of Cellular Metal Contents.

Mouse fibroblast (3T3) cells were seeded at a density of ~ 2 x 10⁵ cells/mL in DMEM, supplemented with 10% bovine calf serum, 1% penicillin-streptomycin solution, and 50 μM of CuCl₂. The cells were growing at 37°C under an atmosphere of humidified air containing 5% CO₂. The high copper culture medium was changed every two days until the cells reached confluence. When confluent, the 3T3 cells were washed with prewarmed PBS followed by an incubation with DMEM containing 10 μM of PSP-2, thipPS, or bithipPS (diluted from 3.0 mM stock solutions in DMSO) respectively. For the negative controls, an equal volume of analytical grade DMSO was used. After a 5-hour incubation, the cells were washed with prewarmed PBS and the cell density was determined with a hemocytometer. The 3T3 cells were pelleted in Eppendorf Safe-Lock tubes at 4500 x g for 5 mins and the supernatants were aspirated. The cell pellets were acidified with 100 μL of concentrated nitric acid (trace metal grade) and heated at 90°C in a heating block for 2 hours. Then 30% hydrogen peroxide (50 μL) was added to each sample, followed by heating at 90°C for another hour. The resulting samples were diluted with 1850 μL of ultrapure water to 2.0 mL for ICP-MS (Agilent 7900 ICP-MS) analysis.

Ratiometric TPEM Imaging.

Mouse fibroblast (3T3) cells were grown on poly-L-lysine treated glass bottom culture dishes (MatTek) to 80% confluency in DMEM growth media supplemented with 10% bovine calf serum, and 1% penicillin-streptomycin. Before imaging, the media was replaced with pre-warmed colorless DMEM (10% fetal bovine serum, 1% penicillin-streptomycin, and 1% sodium pyruvate) containing 1.0 μM of crisp-17 (Chart S1).² The cells were incubated at 37°C under an atmosphere of humidified air containing 5% CO₂ for 15 mins. Images were acquired using a Zeiss LSM confocal NLO 710 microscope equipped with a femtosecond pulsed Ti:sapphire laser at 37°C. Scanning fluorescence micrographs were acquired with excitation at 880 nm and emission simultaneously collected over two channels with bandpass ranges of 479-536 nm (BP1) and 611-750 nm (BP2). Image J was used to analyze the change in fluorescence emission ratio over time as previously described.^{2, 28} The emission ratio for the region of interest was averaged at each time point.

RESULTS AND DISCUSSION

Synthesis.

In previous studies, we developed syntheses for various PSP ligands by substitution of the corresponding phosphine halides or phosphinites using either lithiated trimethylphosphine sulfide (1) or an aryl-magnesium halide Grignard derivative as the nucleophile.^{2, 16} This approach relies on in situ generation of the organometallic species and the electrophilic phosphorus building block, both of which are cumbersome to handle due to their moisture and oxygen sensitivity. Employing the conditions established by Murata and Buchwald,²⁹ we also installed C-P bonds in aryl-substituted PSP ligands via a Pd-catalyzed cross coupling reaction from the corresponding halide derivatives and the secondary phosphine 3.¹⁷ The latter is a bench stable solid that can be recrystallized from boiling water. For the synthesis of thipPS 5, we thus performed a double cross-coupling reaction of phosphine 3 with 2,3-dibromothiophene 4 to afford the target ligand in 40% yield (Scheme 1). Likewise, the reaction of 3,3'-dibromo-2,2'-bithiophene 6 with phosphine 3 produced bithipPS 7 in 30% yield. The somewhat moderate yields are primarily due to competitive reductive dehalogenation, regardless of the nature of the aryl halide starting material. To further optimize the cross-coupling conditions, we examined the effect of the base on the reaction yield of bithipPS 7. Replacing Cs₂CO₃ with anhydrous K₃PO₄ resulted in a modest improvement from 30% to 41% yield (Table S1).

Scheme 1.

Synthesis of thipPS and bithipPS

X-ray Structures.

To investigate the degree of structural preorganization of the ligands towards Cu(I) binding, we determined the crystal structures of thipPS and bithipPS, as well as the corresponding Cu(I) complexes [Cu(I)thipPS]PF₆ (8) and [Cu(I)bithipPS]PF₆ (9). Single crystals of thipPS suitable for crystallographic characterization were obtained by diffusing methyl tert-butyl ether into a saturated solution of thipPS in dichloromethane. Similarly, bithipPS crystallized in a biphasic mixture of dichloromethane and methyl tert-butyl ether and yielded colorless plate-shaped crystals. The complexes [Cu(I)thipPS]PF₆ (8) and [Cu(I)bithipPS]PF₆ (9) were isolated from a mixture of dichloromethane-acetonitrile (1:1) containing the respective ligands and 1.1 mol equiv. of [Cu(I)(CH₃CN)₄]PF₆. Pertinent geometrical parameters of the structures are compiled in Tables 1 and 2 and additional crystallographic data are provided with the Supporting Information. Structural representations and atom numbering schemes are shown in Figures 1 and 2.

Table 1.

Selected Bond Distances (Å) and Angles (deg) in the Crystal Structures of thipPS 5 and bithipPS 7.

	thipPS 5		bithipPS 7
P1-C1	1.831(2)	P1-C2	1.837(3)
P2-C2	1.827(2)	P2-S1	1.9639(11)
P3-S1	1.9605(7)	P3-S2	1.960(2)
P1···P2	3.4128(6)	P1···P1	6.000(3)
P1-C1-C2	124.16(16)	P1-C2-C1	125.1(2)
P2-C2-C1	123.28(16)	C2-C1-C1-C2	160.79(6)

Table 2.

Selected Bond Distances (Å) and Angles (deg) Describing the Cu(I) Coordination Geometry in the Crystal Structures of [Cu(I)₂(thipPS)₂](PF₆)₂ 8 and [Cu(I)bithipPS]PF₆ 9.

	[Cu(I)2(thipPS)2] 8		[Cu(I)bithipPS] 9
Cu1-P1	2.3104(18)	Cu1-P1	2.2657(4)
Cu1-P2	2.2791(17)	Cu1-P2	2.2695(4)
Cu1-S1	2.3959(17)	Cu1-S1	2.3373(4)
Cu1-S2’	2.2868(18)	Cu1-S2	2.3282(4)
P1-C1	1.806(4)	P1-C2	1.8180(14)
P2-C2	1.827(4)	P2-C18	1.8196(15)
P3-S1	1.989(2)	P3-S1	1.9920(5)
P1-C1-C2	121.0(2)	P1-C2-C1	120.56(11)
P2-C2-C1	122.2(2)	P2-C18-C17	119.69(11)
P1···P2	3.286(2)	P1···P2	3.631(6)
P1-Cu1-P2	91.45(6)	P1-Cu1-P2	106.362(15)
S1-Cu1-S2’	116.55(7)	S1-Cu1-S2	108.514(19)

Figure 1.

ORTEP drawing and atom numbering scheme for crystal structures of (A) thipPS 5 and (B) [Cu(I)₂(thipPS)₂](PF₆)₂ 8. Ellipsoids shown represent 50% probability. H atoms and counterions are omitted for clarity.

Figure 2.

ORTEP drawing and atom numbering scheme for crystal structures of (A) bithipPS 7 and (B) [Cu(I)bithipPS]PF₆ 9. Ellipsoids shown represent 50% probability. H atoms and counterions are omitted for clarity.

The [Cu(I)thipPS] complex 8 crystallized as a dimer with C_i symmetry. Each Cu center adopts a distorted tetrahedral coordination geometry, where both trivalent P atoms coordinate to Cu to form a 5-membered ring (Figure 1). However, only one of the four phosphine sulfide donors is bound to Cu, whereas a second phosphine sulfide moiety is provided by the adjacent ligand of the dimer, resulting in an interlocked structure similar to that of [(PSP-1)₂Cu₂](PF₆)₂.¹⁶ The Cu1-P1/2 bond distances (2.310 and 2.279 Å) are comparable to those in [Cu(I)phenPS], suggesting similar Cu-P bond strengths. Likewise, the Cu1-S2’ bond distance to the bridging phosphine sulfide donor is with 2.287 Å in line with expectation;¹⁷ however, the Cu center is displaced by 0.44Å above the thiophene ring plane to accommodate the geometrical constraints imposed by the 5-membered chelate ring of the phosphine sulfide arm. In addition, the Cu1-S1 bond is significantly elongated (2.395 Å), indicating a weaker interaction, presumably due to ring strain.

The x-ray structure of ligand 5 revealed a conformational preorganization similar to that of phenPS;¹⁷ however, the geometry of the five-membered thiophene ring results in a wider bite angle and thus a larger nonbonding P···P distance of 3.413 Å compared to 3.070 Å in phenPS. Consequently, the P1-C1-C2 and P2-C2-C1 bond angles are rendered more acute upon Cu(I) complexation (Table 1), likely accompanied by an increase in strain energy. At the same time, the wider bite angle of the free ligand reduces the electron repulsion between the phosphorous lone pairs, thus resulting in a lower strain energy in the free ligand compared to phenPS. Both effects are expected to lower the stability of the Cu(I) complex of thipPS relative to phenPS (vide infra). Moreover, the wider bite angle is likely responsible for favoring a dimeric over monomeric structure as coordination of the second phosphine sulfide moiety (P6-S4) would result in additional buildup of ring strain.

In contrast to the dimeric structure 8, [Cu(I)bithipPS]PF₆ 9 crystallized as a discrete monomeric complex with C₂ symmetry. Similar to [Cu(I)phenPS],¹⁷ the Cu(I) center adopts a distorted tetrahedral geometry in which the coordination planes, defined by S-Cu-S and P-Cu-P, intersect at an angle of 67.5°, thus indicating substantial flattening compared to an ideal tetrahedral geometry. Contrary to complex 8, the Cu-P and Cu-S bond lengths are uniform averaging 2.268 ± 0.003 Å and 2.333 ± 0.006 Å, respectively, indicating similar bond strengths around the tetrahedral Cu(I) center.

The free ligand 7 adopts a nearly anti-periplanar conformation with a torsion angle (C2–C1···C1–C2) of 160.8°, such that the thiophene sulfur atom is brought within 3.264 Å distance to the trivalent phosphorous atom, corresponding to ~83% of the sum of the van der Waals radii. Upon Cu(I) complexation, the bithiophene backbone is twisted out of plane with an axial torsion angle of 71.3°. Concomitantly, the P1-C2-C1 bond angle decreases from 125.1° to 120.6°, indicating some buildup of strain energy despite the axial conformational flexibility of the bithiophene backbone. At the same time, the average P-Cu bond length of 2.268 ± 0.003 Å remains, within experimental error, identical compared to [Cu(I)phenPS] (2.277 ± 0.014 Å), while the P-Cu-P bond angle of 106.36 is significantly wider compared to [Cu(I)phenPS] (90.5°) or [Cu(I)thipPS] (91.5°). Altogether, the bithipPS ligand scaffold can better accommodate the coordination requirements for Cu(I) compared to thipPS, thus minimizing steric and ring strains to favor a discrete monomeric structure.

Cu(I) Coordination Chemistry in Solution.

To investigate the Cu(I)-binding stoichiometry of the two ligands thipPS and bithipPS in solution, we conducted molar-ratio titrations in PIPES buffer (10 mM, pH 7.0, 0.1 M KClO₄, 25°C) using BCA as a colorimetric indicator.¹⁷ The bidentate ligand BCA forms with Cu(I) a purple 2:1 complex (λ_max = 562 nm and ε = 7900 M⁻¹cm⁻¹), with a stability constant of logβ₂ = 17.7 at pH 7.0 (0.1 M KClO₄, 25°C).²² The [BCA₂Cu(I)]³⁻ complex was prepared from CuSO₄ (50 μM) through in situ reduction with sodium ascorbate (150 μM) in the presence of 150 μM of BCA. Titration of this solution with either ligand yielded a linear decrease in absorbance, which leveled off at an equimolar ligand/Cu(I) concentration (Figure S1), thus indicating formation of a 1:1 complex as the predominant species.

To elucidate the coordination mode of the two ligands in more detail, we acquired ¹H NMR spectra in the presence and absence of equimolar amounts of [Cu(I)(CH₃CN)₄]PF₆ in DMSO-d₆. As shown in Figure 3A, the two constitutionally distinct dimethylphosphorothioylmethyl arms of thipPS produce two pairs of doublets of doublets between 2.6-3.3 ppm that can be assigned to the methylene groups engaging in geminal H─C─H coupling and coupling to the neighboring P nuclei of the phosphine sulfide moiety. As observed for phenPS and PSP-1, coupling with the trivalent P nuclei appears to be negligible.^{16, 17} Moreover, the four pairs of enantiotopic methyl groups couple with the neighboring P nuclei of the heterotopic phosphine sulfide moieties (J_PH = 13.1 Hz) to produce a total of four distinct doublets between 1.6-1.8 ppm.

Figure 3.

¹H NMR spectra of (A) thipPS 5 and (B) bithipPS 7 in the absence and presence of 0.5 or 1.0 mol equiv. of [Cu(I)(CH₃CN)₄]PF₆ in DMSO-d₆ at 25°C. The solvent signals are labeled with an asterisk.

In the presence of equimolar Cu(I), all signals were shifted downfield relative to those of free thipPS 5. Although the crystal structure implies that the four dimethylphosphorothioylmethyl substituents are constitutionally heterotopic, the complexity of the coupling pattern did not increase compared to the free ligand, indicating that the phosphine sulfide groups exchange dynamically at a significantly faster rate relative to the NMR time scale.

To gauge the kinetics of Cu(I) exchange between free thipPS 5 and its Cu(I) complex, we further acquired a ¹H NMR spectrum in the presence of 0.5 mol equiv. of Cu(I). Under these conditions, two distinct sets of sharp resonances were observed, both for the aliphatic and thiophene protons. Even when the temperature was raised to 65°C, all signals remained sharp (Figure S6), indicating a slow exchange kinetics relative to the NMR time scale and thus a high barrier for the Cu(I) self-exchange equilibrium.

By virtue of its C₂-symmetry, bithipPS produced a ¹H NMR spectrum analogous to phenPS and PSP-2, with two overlapping doublets arising from the two diastereotopic methyl groups attached to each phosphine sulfide moiety and a pair of doublets of doublets from the bridging methylene units of the two dimethylphosphorothioylmethyl arms (Figure 3B).^{16, 17} In the presence of equimolar Cu(I), the methyl groups resonances split into broadened doublet of doublets, indicating formation of a C₂-symmetric complex analogous to the x-ray structure, in which only two out of the four phosphine sulfide moieties are coordinated. Upon heating to 65°C, the resonances in the aromatic region remained sharp; yet the aliphatic signals continued to broaden and finally started to coalesce towards a single wide and unstructured peak (Figure S6). The observed temperature-dependent line broadening indicates that the rate for intramolecular coordinative exchange of the dimethylphosphorothioylmethyl groups is significantly slower compared to thipPS or phenPS. The higher kinetic barrier might be due to significant conformational reorganization of the backbone, which must undergo an axial twisting motion in the course of the coordinative exchange process.

Analogous to thipPS, addition of 0.5 mol equiv. of Cu(I) yielded a spectrum with two sets of resonances that correlated with the free ligand 7 and [Cu(I)bithipPS] complex 9. Heating to 65°C did not lead to signal broadening in the aromatic region, consistent with a high kinetic barrier for Cu(I) self-exchange as observed for thipPS (Figure S6).

Dimer Formation.

To investigate whether [Cu(I)thipPS] may form a dimeric complex in aqueous solution, we performed a series of ³¹P NMR and UV-vis spectroscopic studies. Because [Cu(I)thipPS]PF₆ does not dissolve at millimolar concentrations in aqueous buffer, the dimerization studies were conducted with the more soluble [Cu(I)thipPS]Cl complex. When [Cu(I)-thipPS]Cl was diluted from 10 mM to 1.0 mM in aqueous phosphate buffer, the ³¹P NMR revealed a downfield shift of the phosphine sulfide phosphorous resonance centered around 45 ppm (Figure S9A). This observation is consistent with a concentration-dependent dimerization; however, only an averaged signal could be observed due to the fast exchange kinetics relative to the NMR timescale. Moreover, an overlay of the UV-vis spectra of [Cu(I)thipPS], acquired at concentrations ranging from 50 μM to 1.5 mM, showed small but significant changes of the molar absorptivity (Figure S10); however, a spectral deconvolution analysis indicated that dimerization did not pass the mid-point and thus it was not possible to determine a reliable dimerization constant. Based on these data we can conclude that logK_dimer is less than 2, which is significantly weaker compared to the dimerization constant of [Cu(I)PSP-1] (logK_dimer = 3.7, 25°C).¹⁶ In contrast to [Cu(I)thipPS], neither the ³¹P NMR nor UV-vis spectra of complex 9 indicated any significant degree of dimerization up to 10 mM complex concentration (Figure S9B).

Stability Constants of Cu(I) Complexes.

To investigate the strength of the Cu(I)-thipPS interaction in aqueous solution, we performed a spectrophotometric titration at constant pH in PIPES buffer using [Cu(I)MCL-2]PF₆ (logK_Cu(I)L of 13.08 ± 0.13) as the source of Cu(I) ions.²² Consistent with strong and quantitative Cu(I) chelation, addition of [Cu(I)MCL-2] PF₆ produced a linear response in the absorption spectrum with a sharp saturation point at 1.0 mol equiv. of Cu(I) (Figure 4). These data suggest that under Cu(I) limiting conditions, thipPS forms a tight 1:1 complex with an affinity exceeding that of MCL-2 by several orders of magnitude. Addition of Cu(I) beyond equimolar concentrations produced smaller but nevertheless distinct spectral changes (Figure S4), indicating coordination of a second equivalent of Cu(I) to a weaker binding site. Nonlinear least-squares fitting with a 1:1 binding model yielded a logK₂ of 12.21 ± 0.08 (K_d = 0.62 ± 0.12 pM) for this second binding site (Figure S4).

Figure 4.

Evaluation of binding stoichiometry to Cu(I) in PIPES buffer (pH 7.0, 10 mM, 0.1 M KCl, 25°C). (A) Molar ratio titration of thipPS (50 μM) with [Cu(I)MCL-2]PF₆. Inset: Experimental (black circles) and calculated (red trace) absorbances at 280 nm. (B) Molar ratio titration of bithipPS (40 μM) with [Cu(I)MCL-2]PF₆. Inset: Experimental (black circles) and calculated (red trace) absorbances at 280 nm.

To determine the complex stability constant for binding of the first equivalent of Cu(I), we employed BCS as a competitor ligand, which offers a significantly higher affinity compared to MCL-2 (logβ₂ (BCS) = 20.93 ± 0.05, pH 7.0, 0.1 M KClO₄, 25 °C).¹⁷ In addition, BCS forms with Cu(I) a brightly colored 2:1 complex (λ_max = 483 nm and ε = 13300 M⁻¹ cm⁻¹),³⁰ which simplifies the spectral deconvolution during data analysis. Thus, titration of 30 μM of Cu(I) with thipPS in the presence of excess BCS (500 μM) resulted in a gradual decrease of the 483 nm absorption band (Figure S7). Nonlinear least-squares fitting from 400 to 600 nm yielded a logK value of 19.46 ± 0.04 (pH 7.0, 10 mM PIPES, 0.1 M KCl, 25°C) corresponding to a K_d of 34.7 ± 3.3 zM (Figure S7).

The molar ratio titration of bithipPS with [Cu(I)MCL-2]PF₆ produced a clean set of isosbestic points at 287 and 323 nm with a sharp saturation point at equimolar concentration of Cu(I), consistent with formation of a complex with 1:1 stoichiometry (Figure 4B). In contrast to thipPS, the spectrophotometric titration of bithipPS with [Cu(I)MCL-2]PF₆ yielded only insignificant spectral changes beyond equimolar Cu(I) concentrations, indicating neglible binding of a second equivalent Cu(I) under these conditions (Figure 4B). To determine the 1:1 bithipPS-Cu(I) complex stability constant, we employed the same approach as described for thipPS using 1.5 mM of BCS as competitor ligand (Figure S8). Non-linear least-squares fitting of the titration data yielded a logK value of 20.40 ± 0.08 (K_d = 3.98 ± 0.81 zM), approximately one order of magnitude higher compared to thipPS.

Electrochemistry.

Due to the limited solubility of the Cu(I) complexes of thipPS and bithipPS in aqueous buffer, we determined their redox stability in methanolic solution by cyclic voltammetry using Bu₄NPF₆ as the background electrolyte. The corresponding Cu(I) complexes were formed in situ in acetonitrile from [Cu(I)(CH₃CN)₄]PF₆ and then diluted into methanol. As shown in Figure 5, slow-scan cyclic voltammetry (20 mV/s) indicated a quasi-reversible one-electron redox process for [Cu(I)-thipPS]PF₆ with a half-wave potential of E_1/2 = 0.258 vs. Fc^+/0 and a peak-to-peak separation of ΔE_P = 89 mV. Likewise, [Cu(I)bithipPS]PF₆ also revealed a quasi-reversible one electron redox process (ΔE_P = 62 mV) with a half-wave potential of E_1/2 = 0.226 vs. Fc^+/0 (Figure 5). Compared to [Cu(I)phenPS]PF₆ with E_1/2 = 0.152 vs. Fc^+/0 (0.1 M Bu₄NPF₆ in MeOH),¹⁷ both thipPS and bithipPS form Cu(I) complexes with significantly higher redox stability under the same conditions. Given the more positive half-wave potential (32 mV) but lower Cu(I) stability constant of thipPS, the Cu(II) complex of thipPS is still more than 2-fold less stable compared to bithipPS. This might be due to the shortened coordination bonds around the Cu(II) center, which impose an additional enthalpic penalty upon oxidation of the already strained [Cu(I)thipPS]PF₆ complex. On the other hand, the conformationally flexible backbone of bithipPS can better accommodate the coordination requirements and alleviate strain energy around the Cu(II) center.

Figure 5.

Cyclic voltammograms of (A) thipPS (0.4 mM) and (B) bithipPS (0.4 mM) in the presence of [Cu(I)(CH₃CN)₄]PF₆ (0.36 mM) in methanol at 25°C (0.1 M Bu₄NPF₆, glassy carbon electrode, 20 mV/s scan rate, and a nonaqueous Ag⁺/Ag reference electrode with Fc⁺/Fc as the external reference).

Metal Ion Selectivity and Competitive Chelation in Live Mouse Fibroblasts.

In agreement with the high Cu(I) selectivity of the previously reported PSP ligands, thipPS and bithipPS showed negligible binding of Mn(II), Fe(II), or Zn(II), even at 1.0 mM metal ion concentrations as assessed by UV-vis spectroscopy (Figure S11). Moreover, acidification with 0.1M HClO₄ had no effect on the UV-vis spectrum, indicating both low basicity of the ligands and hydrolytic inertness of the phosphine sulfide bond.

To evaluate the selectivity towards Cu(I) binding within a complex biological environment, we incubated live mouse fibroblasts with 10 μM thipPS or bithipPS over 5 hours. For comparision, we also tested the effect of the high-affinity Cu(I) chelator PSP-2 under the same conditions (Figure 6). Because cells grown in basal medium contain only a small fraction of copper compared to zinc and iron, we supplemented the growth medium with 50 μM CuCl₂, which was removed prior to incubation with the chelators. Consistent with the spectroscopic studies, both thipPS and bithipPS proved as effective as PSP-2 at removing copper from cells, whereas statistically insignificant changes were observed for manganese, iron, and zinc. Despite the large differences in Cu(I) stability constants, all three ligands were equally effective at selectively depleting cellular copper from an average of 8.9 ± 0.7 fg/cell to 3.1 ± 0.5 fg/cell. Moreover, a cell viability assay indicated low toxicity for both ligands at concentrations up to 25-50 μM. In contrast, the membrane-permeant heavy metal chelator TPEN resulted in significantly reduced cell viability at concentrations as low as 1.0 μM (Figure S12).

Figure 6.

ICP-MS analysis of cellular Mn (A), Fe (B), Cu (C), and Zn (D) contents of 3T3 mouse fibroblast cells (average of ~ 5 x 10⁶ cells) grown in media supplemented with 50 μM CuCl₂. Addition of PSP ligands (10 μM) causes reduction of total cellular copper content after 5 hours (mean ± standard deviation, n = 9 replicates).

Visualizing Competitive Copper Chelation in Live Cells by Ratiometric Two-Photon Excitation Microscopy.

While the selectivity studies with mouse fibroblast cells demonstrated a remarkable ability of thipPS and bithipPS to sequester copper without altering zinc, iron, or manganese levels, it remains unclear whether the ligands crossed the plasma membrane to remove copper directly from the intracellular sites or whether their activity is limited to extracellular chelation. To address this question, we performed a series of live cell imaging experiments using the Cu(I)-selective fluorescent probe crisp-17 (K_d = 8 aM) for tracking changes in labile intracellular Cu(I) by two-photon excitation microscopy as previously reported.² To this end, adherent mouse fibroblasts were grown in basal medium and then incubated with crisp-17 (1.0 μM). Ratiometric micrographs were derived from fluorescence intensity images collected over two emission channels (479-536 nm, and 611-750 nm) with two-photon excitation at 880 nm (Figure 7A). Cells were treated on stage with 10 μM of CuGTSM (Chart S1),^{31, 32} which was readily reduced upon cellular uptake to release Cu(I) as manifested by the large increase of the fluorescence ratio from an average of ~ 0.45 ± 0.04 to ~1.93 ± 0.19. The fluorescence ratio changed instantaneously and plateaued within minutes. Subsequent addition of thipPS, bithipPS, or PSP-2 resulted in complete reversal of the CuGTSM-induced ratio increase and restored the initial baseline ratio to ~0.43 ± 0.04 within a few minutes (Figure 7B-D). Despite the weaker affinity of thipPS compared to bithipPS or PSP-2, all three ligands yielded a similar quantitative response. The observed rapid kinetics of Cu(I) removal from crisp-17 is consistent with a mechanism where the ligands can enter the cells and directly interact with the target Cu(I) complexes. Compared to PSP-2, both thipPS and bithipPS removed Cu(I) at a slightly slower rate, which might be attributed to the larger molecular size and thus slower diffusion kinetics across the plasma membrane.

Figure 7.

Ratiometric imaging of labile Cu(I) dynamics in live 3T3 cells with crisp-17 (1.0 μM). (A, Top) Fluorescence intensity images acquired with emission channels of 479 - 536 nm (BP1) and 611 - 750 nm (BP2), two-photon excitation at 880 nm. (A, Bottom) Ratio image with ratio = BP2/BP1. (B, Left) Ratio images before and after addition of 10 μM CuGTSM, and 50 μM thipPS. (Scale bar 50 μm). (B, Middle) Time course of fluorescence ratio change averaged over 10 cells indicated with red circles and gray error bars. (B, Right) Mean fluorescence ratio of cytoplasmic region (P values calculated for n = 10 using a two-tailed test). (C) Experiment as described for B but with 50 μM bithipPS. (D) Experiment as described for B but with 50 μM PSP-2.

CONCLUSIONS

Balancing the conformational flexibility and geometrical preorganization represent an effective concept for tuning the stability and redox properties of Cu(I) complexes with PSP-ligands. Due to the hinged bithiophene backbone, bithipPS adapts well to the coordination requirements of Cu(I) and forms a discrete mono-nuclear complex. With a logK of 20.4, the Cu(I) stability constant of bithipPS is higher compared to PSP-2 (logK = 20.0), which features a more flexible ethylene bridge. At the same time, the Cu(I) affinity of bithipPS does not exceed that of the more rigid phenPS ligand (logK = 21.2), suggesting that the structural preorganization of phenPS matches well the coordination preference of the Cu(I) center. This is supported by the similar Cu-P and Cu-S bond distances in both Cu(I) complexes. By contrast, the wider bite-angle of thipPS leads to buildup of steric strain upon Cu(I) coordination as reflected by the longer Cu-P bond distances and significantly decreased binding affinity (logK = 19.5) compared to bithipPS. Despite the almost 10-fold difference in Cu(I) complex stability constants, the reduction potentials of the Cu(II/I) couple bound to each ligand are strikingly similar. Biological studies with mouse fibroblasts demonstrated that both ligands selectively remove copper without affecting iron, zinc, and manganese levels. As revealed by two-photon excitation microcopy with the Cu(I)-selective ratiometric probe crisp-17, the ligands are also cell permeant and can competitively remove Cu(I), thus rendering both ligands a valuable addition for selectively manipulating copper levels within complex biological systems.