Ideally, the half-cell potential of an organic-electrolyte-based reference electrode is determined by equilibrium distribution of the organic electrolyte across the interface between an organic-electrolyte-doped membrane and the aqueous sample. The limit of applicability (LOA) of these reference electrodes is reached when sample ions transfer into the reference membrane at high concentrations. Recent insights highlight the need to evaluate the lipophilicity of the organic electrolyte's anions and cations separately; however, accurate predictions of the LOAs also require consideration of ion fluxes. LOAs are influenced not only by the concentration of interfering sample ions but also by the mobilities of the organic electrolyte and interfering ions in the sample and membrane. This was recently shown with numerical simulations and described with expressions for limiting cases in which the interfering ion enters the membrane by either ion exchange with the organic electrolyte ion of the same charge sign or co-extraction with the organic electrolyte ion of opposite charge sign. More general expressions describing LOAs that take diffusion into account and apply when ion exchange and co-extraction occur simultaneously have been missing. Here, a quasi-steady-state model is presented that describes the LOAs based on organic electrolyte partitioning, ion exchange, and diffusive mass transfer limitations. Depending on the lipophilicity of the organic electrolyte and the rate of diffusion in the membrane, four limiting cases can be identified. Above the LOA, these reference electrodes exhibit characteristic sub-Nernstian, Nernstian, or super-Nernstian responses to the sample ions. The respective LOAs can be quantitatively predicted as a function of sample composition using two straightforward equations.
Simultaneous monitoring of multiple protein-protein interactions in live cells remains a key challenge in biology and drug discovery. While multiplexed FRET enables parallel molecular readouts, existing approaches are often constrained by spectral overlap, complex instrumentation, or incompatibility with live-cell models. To overcome these limitations and increase accessibility to the broader biological community, we present multiplexed dark FRET (MDF), a genetically encoded platform that uses spectrally distinct donors (mNeonGreen, mScarlet-I3) paired with nonemissive acceptors (ShadowY, ShadowR). We first establish that MDF fluorophores exhibit minimal background FRET under co-expression, enabling clean separation of donor lifetimes under multiplexed conditions. Using fluorescence lifetime (FLT) detection, we demonstrate MDF's versatility through three biologically and translationally relevant examples: (1) cell-type-specific biosensing in organoids, as exemplified in 3D neuro-glial spheroids; (2) target specificity for drug discovery through discrimination of TNFR1 versus TNFR2 receptor conformations and selective FLT modulation by receptor-specific small molecules; and (3) protein misfolding, as exemplified through simultaneous monitoring of alpha-synuclein oligomerization and misfolding. We further show that MDF can be applied within a single cellular environment, demonstrating the feasibility of same-cell multiplexing under optimized transient transfection conditions. MDF provides a scalable framework for real-time, live-cell biosensing across high-throughput, target-specific, and tissue-level applications in complex biological systems.
The enzyme γ-glutamyl transferase (GGT) plays an important role in redox homeostasis and is overexpressed in many different cancers. The ability to assess GGT activity provides critical insights into cancer prognosis and pathology. Here, we present a novel imaging agent, (15N)2-DT-GG-PA, for in vivo assessment of GGT activity by hyperpolarized nitrogen-15 magnetic resonance imaging (HP 15N MRI). This 15N-labeled HP MRI probe is developed by introducing a long-lasting HP (15N)2-diazirine tag to deliver essential properties for in vivo detection of GGT activity. Our studies show a large chemical shift difference (Δδ = 6.3 ppm) between (15N)2-DT-GG-PA and its product, (15N)2-DT-PA, and long HP lifetimes of both compounds (T1 = 196 and 81 s, respectively, at 1 T). We also demonstrate the in vivo feasibility of detecting GGT activity using HP (15N)2-DT-GG-PA in rat kidneys on a clinical 3 T MRI scanner. Finally, the success of (15N)2-DT-GG-PA highlights the transformative potential of the 15N-tagging approach in designing novel HP 15N MRI probes beyond structural restrictions for 15N-isotope labeling.

