Ca²⁺ is involved in many central physiological activities (Clapham, 2007; Brini et al., 2014). As a universal second messenger, Ca²⁺ regulates cellular activities including cytoskeletal motility, phosphorylation and dephosphorylation-dependent enzymatic activity, and secretion of biomolecules such as neurotransmitters. During propagation of an action potential, voltage-sensitive Ca²⁺ channels on neurons open to allow an influx of Ca²⁺ (Brini et al., 2014). At a synapse, this influx of Ca²⁺ ions leads to neurotransmitter release and signal propagation. Because Ca²⁺ undergoes dramatic translocations during signaling events, being able to measure Ca²⁺ concentration dynamics is of particular importance.

The local concentration of Ca²⁺ varies from nanomolar level to millimolar level in cells. In the resting state, the intracellular concentration is typically maintained at approximately 100 nM in neurons (Khodorov et al., 1993; Bagur and Hajnóczky, 2017). Extracellular Ca²⁺ is typically maintained at 1 mM, 10,000-fold greater than the resting intracellular concentration. Ca²⁺ indicators with different affinities are developed to accommodate the wide span of Ca²⁺ concentrations in various biological contexts.

Small molecule-based Ca²⁺ indicators (Figure 1) typically contain a fluorophore and a Ca²⁺ chelator and often exhibit an increase in fluorescence in response to Ca²⁺. Ca²⁺ indicators are typically designed to operate through a photo-induced electron transfer (PeT) mechanism—a process in which an electron is transferred from the excited state electron donor, typically the Ca²⁺ chelator, to the electron acceptor, typically the fluorophore (Figure 1A) (Aigner et al., 2011). In absence of Ca²⁺, the electron-rich Ca²⁺ chelator quenches the fluorophore via PeT as the electrons from the chelator are transferred to the low-energy molecular orbital of the fluorophore and prevent photo-excited electrons in the high-energy molecular orbital from returning to the low-energy molecule orbital. In presence of Ca²⁺, this PeT mechanism is interrupted, allowing photo-excited electrons in the high-energy molecular orbital to return to the unoccupied low-energy molecular orbital by emitting fluorescence (Carter et al., 2014). The fluorescence of PeT-dependent small molecule-based Ca²⁺ indicators can be modulated by the presence of electron-rich or electron-withdrawing groups nearby.

Genetically encodable Ca²⁺ indicators (GECIs) share a general design that consists of a binding domain and a reporter domain. The reporter domain is typically a fluorescent protein (FP). Common choices for binding domains are the calmodulin (CaM)/CaM-binding peptide pair and troponin-C. Ca²⁺-dependent conformational changes of the binding domain cause changes of the FP chromophore environment, which in turn alters the fluorescence profile, allowing the monitoring of the Ca²⁺ concentration dynamics. A large collection of GECIs is available with various colours and affinities.

Zinc (Zn²⁺) is an essential trace metal element in biology. Zn²⁺ is incorporated in over 300 enzymes that have a broad range of biological functions including: regulation of smell, taste, and appetite; synthesis of DNA and RNA; hormonal regulation; immune functions; and antioxidation (Frassinetti et al., 2006). Zn²⁺ is transported mainly by albumin in the blood and is maintained at a concentration of 1–10 µM in human serum or plasma (Frassinetti et al., 2006). Total cytosolic Zn²⁺ concentration is a few hundreds of micromolar, but most Zn²⁺ is tightly bound, leaving free Zn²⁺ at a picomolar level. Free Zn²⁺ is heterogeneously distributed in cell: 1 pM in the endoplasmic reticulum and the Golgi apparatus, 0.1 pM in the mitochondrial matrix, 180 pM in the cytosol, and 200 pM in the nucleus (Carter et al., 2014). Some important cellular activities are accompanied by Zn²⁺ dynamics. For example, Zn²⁺ secretion is associated with insulin secretion of human pancreatic β-cells (Chen et al., 2019). Zn²⁺ is also suggested to play important roles in the brain for its Ca²⁺-dependent release from synaptic vesicles (Assaf and Chung, 1984).

Analogous to small molecule-based Ca²⁺ indicators, small molecule-based Zn²⁺ indicators also consist of a Zn²⁺ chelator and a fluorophore (Figures 4A–E), function using the principle of PeT, and are suitable for live-cell imaging. Common fluorophores include quinoline, fluorescein, 4-aminoapthalimide, and BODIPY (Carter et al., 2014). The small molecule-based Zn²⁺ indicators have been comprehensively reviewed previously by Carter and coworkers (Carter et al., 2014).

The abundance of Zn²⁺-binding proteins in nature provides a number of suitable candidates for engineering GEZIs. They include the CXXC-motif-containing metal-binding domains Atox1/WD4 used in the CALWY family, yeast transcription factor Zap1 in the Zap family, and yeast transcription factor Zif268 in the Zif family.

Most GEZIs are FRET-based. ECFP and mCitrine-based ZifCY2 exhibits a ∆R/R ₀ of 4.0 but is limited by its millimolar affinity (Dittmer et al., 2009). Benefiting from the higher Zn²⁺ affinities of the Atox1/WD4 Zn²⁺-binding pair (Figure 2B), Cerulean and Citrine-based eCALWYs (Figure 4F) exhibit affinities ranging from picomolar to nanomolar (Vinkenborg et al., 2009). Zap indicators also exhibit high Zn²⁺ affinities: ECFP and mCitrine-based ZapCY1 exhibits a ∆R/R ₀ of 2.15 and a K _d of 2.5 pM; ZapCY2 exhibits a smaller ∆R/R ₀ of 1.5 and a larger K _d of 810 pM (Qin Y. et al., 2011).

The ZinCh and eZinCh indicators employ Zn²⁺-binding pockets engineered into FPs. Cerulean and Citrine-based ZinCh-1 was engineered with double mutations of Y39H and S208C on both FPs, leading to two binding sites that each consist of a pair of H39 and C208 (Evers et al., 2007). Further optimization on the binding sites led to eZinCh-2 with a ∆R/R ₀ of 4 and a K _d of 1 nM (Hessels et al., 2015). Although it also binds Pb²⁺ and Cd²⁺, eZinCh-2 is an effective intracellular GEZI for monitoring Zn²⁺ in the cytosol, the ER, the mitochondria, and vesicles.

Chen and coworkers developed single FP-based GEZIs with either Zap1 or a zinc hook as the binding domain (Chen and Ai, 2016; Chen et al., 2019). Zap1 contains two Zn²⁺ finger domains and undergoes a dramatic conformational change upon Zn²⁺ binding. The Pyrococcus furiosus Rad50 zinc hook peptide undergoes homodimerization upon Zn²⁺ binding. ZnGreen1, based on ncpmTFP1 and Zap1, shows superior performance relative to previous FRET-based GEZIs with a ∆F/F ₀ as large as 26.3 and a K _d of 633 pM (Chen and Ai, 2016). ZnGreen2, based on cpmTFP1 and zinc hooks, displays a ∆F/F ₀ of 8.7 and a K _d of 20 µM (Chen and Ai, 2016). One limitation of ZnGreen1 and ZnGreen2 is that they are inverse response indicators with poor photostability. To address this issue, Chen and coworkers replaced mTFP1 with EGFP to create ZIBG1 (Figure 4G) and ZIBG2 (Figure 4H) (Chen et al., 2019). ZIBG1 and ZIBG2 both respond positively to Zn²⁺ binding. ZIBG1 has a ∆F/F ₀ of 2.5 and a K _d of 2.81 µM; ZIBG2 has a ∆F/F ₀ of 7 and a K _d of 282 pM. A red fluorescent GEZI based on mApple and Zap1 was also created with a ∆F/F ₀ of 3.8 and two K _d values of 166 pM and 20 µM according to the biphasic titration curve.

GZnPs are a series of GEZIs also based on Zap1 and cpGFP (Qin et al., 2016; Fudge et al., 2018; Minckley et al., 2019). GZnP1 has a ∆F/F ₀ of 1.2 and a K _d of 34 pM. Engineering efforts on GZnP1 led to GZnP3 with a larger ∆F/F ₀ of 10 and a weaker affinity (K _d = 1.3 nM). The weakened affinity is more appropriate for Zn²⁺ released from granules.

Potassium (K⁺) is one of the most abundant ions in biology. Besides its role as an electrolyte for body fluid balance, K⁺ also plays a pivotal role in excitable cells such as those in the heart and the nervous system (Humphries and Dart, 2015). In the resting state, K⁺ is sequestered into cells by the Na⁺/K⁺ pump and other transporters to maintain a large concentration gradient with the intracellular K⁺ concentration at around 150 mM and the extracellular K⁺ concentration at around 5 mM. When cells are excited, K⁺ effluxes through a variety of voltage-gated K⁺ channels to repolarize cells after depolarization. The dynamics of K⁺ fluxes reflect neuronal and cardiac signal propagation and therefore is an intriguing target for live-cell imaging applications.

Potassium binding benzofuran isophthalate (PBFI) is the earliest small molecule-based K⁺ indicators (Figure 5A) (Minta and Tsien, 1989). It is excited by UV light at 350 nm in the K⁺-unbound state and 344 nm in the K⁺-bound state. Although the indicator is more specific for K⁺ (K _d = 8 mM) than Na⁺ (K _d = 21 mM), better specificity is required for cellular imaging to avoid signal interference by Na⁺, which is typically at an intracellular concentration of 10–40 mM (Minta and Tsien, 1989). PBFI exhibits low brightness primarily due to its low QY of 0.024 in the K⁺-unbound state and 0.072 in the K⁺-bound state. Asante Potassium Green-1 (APG-1) is more red-shifted than PBFI with the peak excitation wavelength at 515 nm and the peak emission wavelength at 540 nm (Rimmele and Chatton, 2014). APG-1 responds to K⁺ with a maximum 4-fold change. Improved APG-2 (Figure 5B) and APG-4 were applied in permeabilized Jurkat and U937 cells to visualize changes in intracellular K⁺ concentration (Rana et al., 2019). Recently, Ratiometric Potassium Sensor-1 (RPS-1) was developed with dual fluorophores: a K⁺-responsive fluorophore Potassium Sensor 525 (PS525) and a non-responsive Coumarin 343 for internal calibration in living tissues (Wang et al., 2021). The dye displays a ∆F/F ₀ of 6 and an apparent K _d of 137 mM.

The identification of K⁺ binding protein (Kbp) enabled the development of GEKIs (Ashraf et al., 2016). Bischof et al. and Shen et al. developed FRET-based GEKIs named GEPII and KIRIN, respectively, (Figure 5C) (Ashraf et al., 2016; Shen et al., 2019; Bischof et al., 2019). The two indicators show comparable ∆R/R ₀ and affinity values: the mCerulean3 and cpVenus-based KIRIN has a ∆R/R ₀ of 1.5 and a K _d of 1.66 mM, and the mseCFP and cpVenus-based GEPII has a ∆R/R ₀ of 2.2 and a K _d of 0.42 mM. Both FRET indicators exhibit outstanding specificity for K⁺, showing no response to Na⁺ at physiologically relevant concentrations. In addition to the FRET-based GEKIs, Shen et al. also described ncpGFP-based GINKO1 (Figure 5D). GINKO1 exhibits a ∆F/F ₀ of 1.5 and higher specificity for K⁺ (K _d = 0.42 mM) than Na⁺ (K _d = 153 mM). Further engineering with site-directed mutagenesis and directed evolution led to GINKO2 with better sensitivity (∆F/F ₀ = 14) and specificity (no response to Na⁺ up to 150 mM) (Wu et al., 2021).

Mg²⁺ plays many pivotal roles in cellular processes and functions. Mg²⁺ serves as an important cofactor for almost every enzyme that needs ATP for catalysis; stabilizes nucleic acids through electrostatic interactions for DNA and RNA synthesis and repair; and regulates mitochondrial Ca²⁺ transport, voltage-gated Ca²⁺ channels, and voltage-gated K⁺ channels as a Ca²⁺ antagonist (Pilchova et al., 2017). Most intracellular Mg²⁺ is bound to biomacromolecules, such that total intracellular concentration of total Mg²⁺ is 17–20 mM but only 0.25–1.5 mM is free Mg²⁺ (Grubbs, 2002; Romani, 2013). Cells maintain a steady-state Mg²⁺ concentration through transporters such as Mrs2p, which transports Mg²⁺ into mitochondria for Mg²⁺ sequestration (Kolisek et al., 2003).

Many small molecule-based Ca²⁺ indicators that exhibit an affinity for Mg²⁺ were modified into Mg²⁺ indicators with o-aminophenol-N, N, O-triacetic acid (APTRA) as the binding group. Some examples include Mag-fura-2 (Figure 6A), Mag-indo-1 (Figure 6B), and Magnesium Green (Supplementary Table S2) (Raju et al., 1989; Morelle et al., 1994; Zhao et al., 1996). Mag-fura-2 is excited at 369 nm and Mag-indo-1 is excited at 349 nm (Raju et al., 1989). Due to the phototoxicity associated with UV excitation, Mg²⁺ indicators with excitation at longer wavelengths were developed. The green indicators Magnesium Green and Mag-fluo-4 (Figure 6C) are excited at 506 and 490 nm, respectively, (Zhao et al., 1996; Shmigol et al., 2001). The above indicators are all subject to interference from Ca²⁺ in live-cell imaging due to their millimolar affinities for Mg²⁺ and micromolar affinities for Ca²⁺.

Other binding motifs were sought to address the lack of Mg²⁺ specificity over Ca²⁺ observed with APTRA. One such motif is the charged β-diketone, which is used in KMG-103 (K _d = 1.8 mM for Mg²⁺, K _d = 6.3 mM for Ca²⁺) and KMG-104 (K _d = 2.1 for Mg²⁺, K _d = 7.5 mM for Ca²⁺) that bind Mg²⁺ preferentially (Komatsu et al., 2004). With an addition of a biarsenical structure to KMG-104, the FlAsH-based KMG-104-AsH (Figure 6D) exhibits an outstanding specificity with K _d values of 1.7 mM for Mg²⁺ and 100 mM for Ca²⁺ (Fujii et al., 2014). The FlAsH-based indicator remain quenched until it covalently links to a tetracysteine tagged protein through its biarsenical motif. This mechanism allows subcellular targeting and decreases background fluorescence. KMG-104-AsH is pH stable in the range of pH 5–6.5 and relatively photostable. The affinity, specificity and stability of KMG-104-AsH make it more suitable for live-cell Mg²⁺ imaging than APTRA-based indicators.

As most Mg²⁺-binding proteins also bind Ca²⁺ with a higher affinity, one of the main challenges for developing GEMIs was to first identify a Mg²⁺-specific binding protein. The first reported GEMIs, the MagFRET series, were based on the Mg²⁺-binding protein human centrin HsCen3 and FPs Cerulean and Citrine (Figure 6E) (Lindenburg et al., 2013). HsCen3 contains two binding sites that bind both Mg²⁺ and Ca²⁺. MagFRET1 has a ∆R/R ₀ of 0.49 for Mg²⁺ and a ∆R/R ₀ of 0.19 for Ca²⁺; MagFRET2 has a ∆R/R ₀ of 0.33 for Mg²⁺ and a ∆R/R ₀ of 0.031 for Ca²⁺. MagFRET1, 2, 7, and 8 have suitable affinities for detecting Mg²⁺ (K _d = 0.15–0.89 mM) in the physiological concentration range of Mg²⁺, but they all have high affinities for Ca²⁺ (K _d = 10–57 µM).

MagIC is a ratiometric GEMI based on mCherry and Mg²⁺/Ca²⁺ sensitive cpVenus (Koldenkova et al., 2015). K _d values of MagIC are 5.1 mM for Mg²⁺ and 4.8 mM for Ca²⁺. Its affinity for Mg²⁺ is too low for detection of intracellular Mg²⁺. The indicator has an in vitro ∆R/R ₀ of 0.5 for Mg²⁺, which is similar to that for Ca²⁺. Just like MagFRETs, the sensitivity of MagIC to Ca²⁺ prevent it from being more widely applied. MARIO1 is another FRET-based GEMI based on a modified cytosolic Mg²⁺-sensing domain of the E. coli Mg²⁺ transporter CorA (CorA-CD) (Maeshima et al., 2018). MARIO1 show improved sensitivity (∆R/R ₀ = 1.53 for Mg²⁺) but still lacks specificity (K _d = 6.2 mM for Ca²⁺ and K _d = 7.2 mM for Mg²⁺).

As another one of the most abundant ions in biology, Na⁺ has many important roles and functions. It regulates osmotic pressure and fluid balance (Reynolds et al., 2006). Na⁺ also plays roles in the immune system by enhancing immune responses for pathogen clearance and affecting the differentiation and functions of immune cells (Wilck et al., 2019). In the nervous system, Na⁺ influxes during the depolarization phase of action potentials for signal propagation (Chahine and O’Leary, 2014). Na⁺ concentration is around 10 mM intracellularly and over 100 mM extracellularly (Strazzullo and Leclercq, 2014). The large gradient across the membrane is maintained by the Na⁺/K⁺ pump, which provides energy for the translocation of Ca²⁺ and other neurotransmitters. There are several small molecule-based Na⁺ indicators, but no genetically encodable indicators have been reported to date due to the lack of an appropriate Na⁺-binding protein.

Sodium binding benzofuran isophthalate (SBFI) (Figure 7A) was developed at the same time as PBFI (Figure 6). SBFI has a K _d of 7.4 mM for Na⁺ and a K _d of 166 mM for K⁺ (Minta and Tsien, 1989). The indicator provides a ∆F/F ₀ of 2. A drawback of SBFI is its low brightness attributed to its low QY of 0.045 in the Na⁺-unbound state and 0.083 in the Na⁺-bound state. Sodium Green has visible light excitation and brighter fluorescence that enable its usage in flow cytometry (Amorino and Fox, 1995). CoroNa Green has a peak excitation wavelength at 492 nm and a peak emission wavelength at 516 nm (Meier et al., 2006; Iamshanova et al., 2016). NaTRIUM Green 2 (ANG-2) (Figure 7B) is slightly red-shifted with peak excitation and emission wavelengths at 517 and 542 nm, respectively, (Roder and Hille, 2014; Iamshanova et al., 2016). The ∆F/F ₀ value of CoroNa Green is 4 and that of ANG-2 is 20 (Iamshanova et al., 2016). The K _d value of CoroNa Green for Na⁺ is ∼80 mM and that of ANG-2 is 34 mM in presence of K⁺ and 20 mM in absence of K⁺. SBFI, CoroNa Green and ANG-2 have been used to detect changes in intracellular Na⁺ concentrations in prostate cancer cell lines (Iamshanova et al., 2016).

Proton (H⁺) concentration is measured as pH (−log[H⁺]). The pH values vary in different cellular compartments: 7.0–7.4 in the cytoplasm, 7.2 for ER, 6.4 in the Golgi apparatus, 5.0 in lysosomes, 5.4 in secretory granules, 6.2 in early endosomes, and 5.3 in late endosomes (Futai et al., 2000). The pH values of organelles are maintained mainly by vacuolar H⁺-ATPase (V-ATPase), which actively pumps H⁺ against its concentration gradient with the energy from ATP hydrolysis (Nelson et al., 2000). The pH environment is important for the functions of the organelles. For example, the lysosomal low pH is important for the optimal functions of lysosomal enzymes such as lysozyme. In other cases, proton gradients across organelle membrane or plasma membrane are used to provide energy for translocation of other molecules, such as neurotransmitters accumulated into synaptic vesicles (Farsi et al., 2017). In mitochondria, proton gradients provide energy for ATP synthesis (Reid et al., 1966). The capability provided by pH indicators (Supplementary Table S3), to measure pH in subcellular compartments of live cells, enables investigations of pH-related processes such as endocytosis and vesicle secretion.

Tsien and coworkers developed 2', 7'-bis (carboxyethyl)-5,6-carboxyfluorescein (BCECF), one of the first fluorescent small molecule-based pH indicators (Rink et al., 1982). BCECF has a pK _a at 6.98, ideal for intracellular pH measurement. The dye shows two peaks in the excitation spectrum at ∼505 and ∼470 nm. The emission intensity exhibits a 3-fold increase from pH 6.4 to pH 7.5. The isosbestic excitation wavelength of BCECF is at 440 nm, where the excitation would lead to a pH-independent emission. The seminaphtharhodafluor (SNARF) family is an alternative pH indicator series with dual excitation and dual emission. SNARF-1, for example, has two peak excitation wavelengths in an acidic environment at 515 and 544 nm, and a peak excitation wavelength of 573 nm in a basic environment (Whitaker et al., 1991). The peak emission wavelength shifts from 573 to 631 nm as pH increases. With yellow to orange fluorescence, SNARF series can be co-imaged in live cells with blue fluorescent dyes such as fura-2 (Martinez-Zaguilan et al., 1991). SNARF-1 and SNARF-2 both have pK _a values above 7 (7.6 for SNARF-1, 7.5 for SNARF-2). SNARF-4 has a lower pK _a of 6.4 to enable pH detection in slightly acidic environments (Marcotte and Brouwer, 2005). The pH indicators pHrodo green and red (Thermo-Fisher) are ideal indicators to image endocytosis because of their low fluorescence signal at neutral pH and increased fluorescence at acidic pH (Godfrey et al., 2008; Lindner et al., 2020).

The development of many of the genetically encodable pH indicators was inspired and enabled by the intrinsic pH sensitivity of FPs themselves. The environmental pH can influence the protonated state of the chromophore, leading to fluorescence changes. EGFP shows dual excitation at approximately 390 and 480 nm and the ratio of the two peaks in the excitation spectra change in response to changes in pH (Kneen et al., 1998). When excited at 480 nm, EGFP shows an emission increases with increasing pH. EYFP (pK _a = 7.1) is more suitable for intracellular pH sensing than EGFP (pK _a = 6.15) due to its neutral pK _a (Llopis et al., 1998).

Fusing a pH-sensitive FP to a FRET donor or acceptor leads to a FRET-based pH indicator. For example, the FluBpH series is based on a fusion of EcFbFP, a flavin mononucleotide (FMN)-binding FP excited at 380 nm, and YFPs with different pK _a values, to generate FRET indicators covering different ranges of pH (Rupprecht et al., 2017). The YFPs include: Citrine (pK _a at 5.7), EYFP (pK _a at 6.1), and EYFP-H148G (pK _a at 7.5). The ratio of 570 nm (YFP fluorescence emission) to 495 nm (EcFbFP fluorescence emission) reflects the pH of the environment. Alternatively, a CFP and YFP-based pH indicator has a peak excitation wavelength at 410 nm and peak emission wavelengths at 476 nm (CFP) and 535 nm (YFP) (Hellwig et al., 2004). pHusion is a tandem concatenation of mRFP1 and the pH-sensitive EGFP (Gjetting et al., 2012). pH-Lemon, based on mTurquoise2 and EYFP, exhibits a highly sensitive ratio change in the range of pH 4.0–7.0 (Burgstaller et al., 2019).

Engineering of GFP for increased pH sensitivity has resulted in a variety of green pH indicators. One example is the ratiometric pH-sensitive variant E²GFP (GFP with F64L/S65T/T203Y/L231H) with a pK _a around 7 (Bizzarri et al., 2006). E²GFP shows ratiometric excitation and emission: the peak emission wavelength is at 510 nm at low pH and red shifts to 523 nm at high pH. In addition, the peak excitation wavelength is at 424 nm at low pH and becomes two peaks (401 and 515 nm) at high pH. Other engineering efforts resulted in the deGFPs that have a variety of pK _a values: deGFP1 variant (S65T/H148G/T203C, pK _a∼8.0), deGFP2 (S65T/C48S/H148C, pK _a∼7.2), deGFP3 (S65T/T203C, pK _a∼6.9), and deGFP4 (S65T/C48S/H148C/T203C, pK _a∼7.3) (Hanson et al., 2002).

Miesenböck and coworkers developed ratiometric pHluorin and ecliptic pHluorin by mutating key residues (Q94, R96, H148, I167, T203, S205, and E222) that affect Y66 protonation or excitation spectrum (Miesenböck et al., 1998). With a fixed emission wavelength at 508 nm, ratiometric pHluorin exhibits an excitation ratio (395–475 nm) change greater than 3-fold in the pH range from 5.5 to 7.5. Ecliptic pHluorin has a ∼5-fold fluorescence emission change in the pH range from 5.5 to 7.5. Superecliptic pHluorin introduced two mutations (F64L and S65T) on ecliptic pHluorin that led to better protein folding and improved sensitivity (∼50-fold) required for imaging of neuronal synaptic vesicle fusion and neurotransmitter release (Sankaranarayanan et al., 2000).

In addition to EGFP, other FPs have been utilized to generate a palette of pH indicators with different colours. Introducing H148G to YFP led to pHVenus with a pK _a at 7.3 (Wachter et al., 1998). pHRed is a red pH indicator based on mKeima with a pK _a of 6.5 and a ratiometric change (585 nm versus 440 nm) over 10-fold (Tantama et al., 2011). pHTomato is another red pH indicator with peak excitation and emission wavelengths at 550 and 580 nm, respectively, (Li and Tsien, 2012). It has a higher pK _a at 7.8 and a ∆F/F ₀ over 1 in the range of pH 7.5–9.8. pHuji is an mApple-based pH indicator with a ∆F/F ₀ of 21 in the range of pH 5.5–7.5 (Shen et al., 2014). pHmScarlet, the latest addition to red pH indicators, is 6-fold brighter than pHuji (Liu et al., 2021). As described by Shen et al., a general limitation of red fluorescent pH indicators, with respect to their use for imaging of vesicle fusion, is their lower Hill coefficients (n_H) relative to green fluorescent pH indicators (Shen et al., 2014).

Chloride (Cl⁻) is an important anion in biology. It regulates proteins and genes including kinases (e.g., Cl⁻-dependent GTP-utilizing protein kinase, nucleoside diphosphate kinase (NDPK), and with-no-lysine kinase (WNK)), channels (e.g., Na⁺-K⁺-2Cl⁻ cotransporter), and receptors (e.g., glutamate ionotropic receptor kainate (GRIK)) (Valdivieso and Santa-Coloma, 2019). Cl⁻ in mammalian cells is maintained at concentrations ranging from 5 to 100 mM by transporters such as potassium chloride cotransporter 2 (KCC2). The serum Cl⁻ concentration is maintained at 100 mM by the kidney. The gradient is used by chloride channels for cellular processes. For example, GABA type A receptor (GABA_A) conducts Cl⁻ upon GABA binding to allow signal propagations (Goetz et al., 2007).

Small molecule-based Cl⁻ indicators include 6-methoxy-N-(3-sulfopropyl) quinolinium (SPQ), N-(ethoxycarbonylmethyl)-6-methoxyquinolinium bromide(MQAE),6-methoxy-N-ethylquinolinium iodide (MEQ), and lucigenin. These Cl⁻ indicators are quenched by Cl⁻ through collisional quenching. The rate of fluorescence quenching is accelerated by higher concentrations of Cl⁻. All of them are long Stoke shift dyes that are excited by UV lights and emit photons at wavelengths approximately 100 nm longer than their peak excitation wavelengths (Verkman et al., 1989; Biwersi et al., 1994). When applied in cells, SPQ and MQAE show severe photobleaching (Krapf et al., 1988; Verkman et al., 1989). MEQ is much less prone to photobleaching than MQAE in Swiss 3T3 fibroblasts (Biwersi and Verkman, 1991). Lucigenin is incompatible with intracellular imaging because its electron-deficient acridine ring is prone to nucleophilic attack in biological samples (Biwersi et al., 1994).

Copper is an essential nutrient to human health. It has implications in cardiovascular (Saari, 2000), immune (Percival, 1998), and nervous systems (Opazo et al., 2014). The element exists predominantly in two forms in biology: the monovalent Cu⁺ and the divalent Cu²⁺. The two different oxidation states of copper are exploited by enzymes that catalyze redox reactions (e.g., cytochrome c oxidase and NADH dehydrogenase) (Festa and Thiele, 2011). Many of these enzymes are located on the membranes of compartmented organelles in eukaryotic cells. Mitochondria, in particular, contain many copper-dependent enzymes for energy production (Baker et al., 2017). Most copper ions are in the cofactors of enzymes, leaving only 10⁻¹⁸—10⁻¹³ M free Cu⁺ (Tapiero et al., 2003). Excess copper is toxic because free Cu⁺ can generate free radicals in cells and destabilize iron-sulfur clusters. Therefore, regulated trafficking is required to prevent cell damages (Stohs and Bagchi, 1995; Macomber and Imlay, 2009; Kaplan and Maryon, 2016). Fluorescent copper indicators enable monitoring of free Cu⁺ concentration, thus leading to a better understanding of Cu⁺ dynamics and localizations under stimuli or stress. To the best of our knowledge, there are no fluorescent indicators designed to be specific for Cu²⁺.

CTAP-1 is the first small molecule-based fluorescent indicator for Cu⁺ with a ∆F/F ₀ of 4.6 (Yang et al., 2005). The indicator consists of a tetrathiaza crown ether Cu⁺-binding motif and a pyrazoline fluorophore. The tetrathiaza crown ether allows specific binding to Cu⁺ (K _d = 40 nM) over other ions including Cu²⁺, and monovalent ions K⁺ and Na⁺. The pyrazoline fluorophore has a peak emission wavelength at 480 nm with UV excitation. CTAP-1 is membrane permeable and allows monitoring of Cu⁺ localization in NIH 3T3 cells grown in a medium supplemented with Cu⁺. One drawback of CTAP-1 is that it aggregates in cells. This issue was largely addressed with CTAP-2, which also exhibits an improved ∆F/F ₀ of 65 (Morgan et al., 2011).

Coppersensor-1 (CS1) is another small molecule-based Cu⁺ indicator based on BODIPY fluorophore and an azatetrathia binding domain (Zeng et al., 2006). The indicator displays a ∆F/F ₀ of 10 and peak excitation and emission wavelengths at 540 and 566 nm, respectively. CS3 is a more sensitive and brighter version with a ∆F/F ₀ of 75 (Dodani et al., 2011a, 20), enabling visualization of intracellular Cu⁺ concentration at basal and depleted levels and identification of Ca²⁺-dependent Cu⁺ redistribution.

Some small molecule-based Cu⁺ indicators were developed for more specific purposes. Mito-CS1 can be targeted to mitochondria via a triphenylphosphonium moiety and therefore enables Cu⁺ imaging in mitochondria (Dodani et al., 2011b). RCS1, a ratiometric indicator, has a ∆R/R ₀ of 20 and peak emission wavelengths at 505 and 570 nm in vitro and is effective as demonstrated in live C6 rat glioma cells (Domaille et al., 2010). ACu1 can be excited by two-photon at 750 nm and was used to image Cu⁺ in hippocampal slices at depths ranging from 90 to 220 µm (Su Lim et al., 2011). Cao Cu-3 is based on a tricarbocyanine scaffold and exhibits a ∆F/F ₀ of 9.6 with near-infrared fluorescence and was applied in live MG63 cells (Cao et al., 2012).

The apparent Cu⁺ affinities of the small molecule-based Cu⁺ indicators range from 10⁻⁸ to 10⁻¹⁴ M, which is typically too weak for the basal concentration of Cu⁺ in the cell. Genetically encodable Cu⁺ indicators address this issue with improved Cu⁺ affinities.

FRET-based genetically encodable Cu⁺ indicators have been developed using several Cu⁺-binding proteins. Wegner and coworkers reported Ace1-FRET, Mac1-FRET, and Amt1-FRET (Wegner et al., 2010, 2011). Amt1-FRET is based on a Cu⁺-responsive transcriptional regulator Amt1 and the CFP/YFP FRET pair (Wegner et al., 2010). The indicator has a high affinity (K _d = 2.5 × 10^–18 M) that enables visualization of the tightly regulated Cu⁺. Ace1-FRET and Mac1-FRET are based on yeast copper regulators Ace1 and Mac1, respectively (Wegner et al., 2011). The two indicators show opposing responses to Cu⁺: Mac1-FRET decreases in FRET efficiency upon Cu⁺ binding and Ace1-FRET increases in FRET efficiency upon Cu⁺ binding. Ace1-FRET has a K _d of 4.7 × 10^–18 ± 8.8 × 10^–19 M and Mac1-FRET has a K _d of 9.7 × 10^–20 ± 1.3 × 10^–20 M for Cu⁺. The Zn²⁺ indicator eCALWY was also modified for Cu⁺ binding (Vinkenborg et al., 2009; Koay et al., 2013). The Zn²⁺-binding domains used in eCALWY are Cu⁺ chaperones ATOX1 and WD4, which contains CXXC Zn²⁺-binding motifs (Vinkenborg et al., 2009). Mutating several cysteines to methionines in the CXXC motifs abolished Zn²⁺ binding while retaining Cu⁺ binding (Koay et al., 2013). The resulting eCALWY-C2M/C3M shows specific binding to Cu⁺ (K _d = 10⁻¹⁵–10⁻¹⁶ M) over Zn²⁺ (K _d = 1.4 × 10⁻⁶ M) and no binding to Co²⁺, Cu²⁺, or Ni²⁺.

The single FP-based Cu⁺ indicators include EGFP-145Amt1, YFP-Ace1 and YAGn series. EGFP-145Amt1is based on Amt1 and ncpEGFP (Liang et al., 2012). YFP-Ace1 is based on Ace1 and YFP (Liu et al., 2013). YFP-Ace1 was optimized by adding GGS linkers of different lengths before N146 of YFP. The resulting variants were named YAGn where n denotes the number of GGS repeats in the linker. YFP-Ace1 and YAG1 are excitation ratiometric at 440 and 496 nm. The longer linkers abolish the ratiometric change. The fluorescent response is 38, 35, 30, 25, and 25% for YFP-Ace1, YAG1, YAG2, YAG3, and YAG4, respectively. The Cu⁺ affinity is increased with increasing linker lengths [K _d values for YFP-Ace1, YAG1, YAG2, YAG3, and YAG4 are (8.2 ± 1.2) × 10⁻¹⁸ M, (2.0 ± 0.8) × 10⁻¹⁸ M, (1.2 ± 1.0) × 10⁻¹⁸ M, (4.6 ± 1.2) × 10⁻¹⁹ M, and (3.3 ± 0.9) × 10⁻¹⁹ M].