Fluorogenic Probes for Imaging Cellular Phosphatase Activity

Abstract

The ability to visualize enzyme activity in a cell, tissue, or living organism can greatly enhance our understanding of the biological roles of that enzyme. While many aspects of cellular signaling are controlled by reversible protein phosphorylation, our understanding of the biological roles of the protein phosphatases involved is limited. Here, we provide an overview of progress towards the development of fluorescent probes that can be used to visualize the activity of protein phosphatases. Significant advances include the development of probes with visible and near-IR excitation and emission profiles, which provides greater tissue and whole-animal imaging capabilities. In addition, the development of peptide-based probes has provided some selectivity for a phosphatase of interest. Key challenges involve the difficulty of achieving sufficient selectivity for an individual member of a phosphatase enzyme family and the necessity of fully validating the best probes before they can be adopted widely.

1. Introduction

Reversible protein phosphorylation is widely used in biology to regulate protein-protein interactions, cellular signaling pathways, etc. [1]. The biological phosphorylation and dephosphorylation of serine, threonine and tyrosine residues has been extensively studied, but phosphorylation also occurs at histidine, lysine, arginine, aspartate, glutamate and cysteine residues [2]. While much less is known about these latter phospho sites, protein phosphorylation is important in both pathological and physiological processes [3–9]. The development of chemical probes to study the biological roles of the enzymes that dephosphorylate phosphoproteins is a topic of intense investigation [10–13].

Substrates that undergo a significant change in fluorescence upon dephosphorylation provide a convenient approach to monitoring phosphatase activity both in vitro and in living systems [14]. Such substrates can provide a highly sensitive readout of phosphatase activity. However, a significant challenge in the field is the development of probes with sufficient selectivity to report on the activity of only one family of phosphatases. Even more challenging is the development of probes that selectively report on the activity of a single phosphatase of interest. An ideal chemical probe for monitoring phosphatase activity in a cellular context would provide a large signal difference between the phosphorylated and dephosphorylated states and react selectively with one family of phosphatases or one specific phosphatase. Excitation and emission wavelengths in the visible region are preferable to UV-excitable probes, and probes with emission in the near-IR region of the spectrum provide additional advantages in tissue and whole animal imaging. In addition, an ideal probe must be thoroughly validated in vitro and in cells/tissues/animals such that its readout can confidently be correlated to activity of the enzyme(s) of interest. In this review, we have highlighted some recent work illustrating progress toward the development of phosphatase-reactive fluorescent probes that can be used to address key questions about phosphatase biology, with a focus on probes that have been used in cellular experiments. Significant contributions have been made in the field in recent years and are discussed here, along with key challenges that remain in the quest to develop an ideal probe.

2. Fluorophores with excitation and emission wavelengths in the visible region of the spectrum

The prototype fluorogenic probe for monitoring phosphatase activity is 6,8-difluoromethylumbelliferyl phosphate (DiFMUP, Figure 1) [15–17]. DiFMUP is an excellent substrate for many phosphatases, including alkaline phosphatases, acid phosphatases, tyrosine phosphatase and serine/threonine phosphatases. Recently, DiFMUP was reported as a sensitive, fluorogenic substrate for assaying histidine phosphatase activity as well [18]. While this substrate is not ideal for cellular applications because it can readily be hydrolyzed by most classes of phosphatases and would thus report on aggregate phosphatase activity levels rather than the activity of an individual family of phosphatases or a single family member, it does provide a facile in vitro assay for phosphatase activity, making it the substrate of choice for many applications [18–25].

Figure 1.

(a) Structures of visible light organic fluorophores. The number below each probe refers to the corresponding reference. (b) Cellular fluorescence resulting from two photon excitation of TP-Phos in HeLa cells. Image adapted from reference 35. (c) Cellular fluorescence from TCF-ALP in the absence (left) or presence (right) of the alkaline phosphatase inhibitor levamisole. Images adapted from reference 31 under a CC BY license. Copyright 2019 Gwynne, Sedgwick, Gardiner, Williams, Kim, Lowe, Maillard, Jenkins, Bull, Sessler, Yoon and James

A number of other fluorophores have been investigated for utility as fluorogenic phosphatase substrates in addition to the coumarin scaffold. The recently developed phosphorylated resorufin (pRES, Figure 1), for example, has several advantages over coumarin-based probes [26]. Specifically, the excitation and emission wavelengths are significantly red-shifted (λ_ex = 560 nm, λ_em = 585 nm as compared to λ_ex = 360 nm and λ_em = 455 nm for DiFMUP), resulting in lower background from cellular autofluorescence and less toxicity arising from high energy excitation. Notably, in addition to the large fluorescence increase observed upon dephosphorylation of pRES, a significant shift in absorbance results in a color change from orange to bright pink. The substrate pRES provides a facile colorimetric and fluorogenic assay for monitoring tyrosine and alkaline phosphatase activity. The fluorinated derivative, F₂pRES, can also be used to monitor acid phosphatase activity. In addition to its utility for monitoring in vitro phosphatase activity, pRES was also shown to provide a sensitive readout of intracellular phosphatase activity [26]. There was no cellular autofluorescence observed at the wavelengths used in this experiment. Incubation of HeLa cells with 50 µM pRES for 10 min resulted in a significant accumulation of intracellular red fluorescence as a result of hydrolysis of pRES by intracellular phosphatases. It was proposed that the fluorescence observed arose primarily from tyrosine phosphatase catalyzed pRES hydrolysis, as pretreatment of the cells with pervanadate, a known tyrosine phosphatase inhibitor [27], significantly reduced the fluorescence observed. However, HeLa cells also express significant levels of alkaline phosphatase activity [28]. While pervanadate does not appreciably inhibit alkaline phosphatase, it does degrade to the potent alkaline phosphatase inhibitor vanadate [29], so it is difficult to rule out the possibility that alkaline phosphatase activity contributes to hydrolysis of pRES in HeLa cells.

Several other scaffolds have recently been investigated as fluorogenic substrates for protein phosphatase activity (Figure 1) [30–36]. These fluorophores all have red-shifted excitation and emission profiles as compared to DiFMUP, and one (TP-Phos [35]), can be excited using two-photon excitation at long wavelength, resulting in negligible cellular autofluorescence and minimal tissue damage. The probe TP-Phos also takes advantage of an auto-immolative linker that spontaneously releases the fluorophore upon dephosphorylation of an ortho-alkyl aryl phosphate. The addition of an ortho substituent adjacent to the aryl phosphate results in significant selectivity of the probe for alkaline phosphatase – tyrosine phosphatases, dual-specificity phosphatases and serine/threonine phosphatases were unable to hydrolyze this substrate [35]. All of the probes are excellent substrates for alkaline phosphatase in vitro and are not appreciably hydrolyzed by acid phosphatase, but the ability of tyrosine and serine/threonine phosphatases to hydrolyze these substrates was not reported (with the exception of TP-Phos, as described above). In cellular experiments, two of these probes (LP1 and Lyso-Phos) were suggested to localize in lysosomes [32,33], while the subcellular localization of the others was not investigated in detail. All of the probes are clearly processed by intracellular phosphatases and cell lines known to express high levels of alkaline phosphatase activity show the largest fluorescence increases when incubated with the probes. Nonetheless, there is a possibility that tyrosine phosphatase activity may contribute to the signal seen in several cell types. Most of the experiments used vanadate to inhibit intracellular phosphatase activity, which can inhibit both alkaline phosphatase and tyrosine phosphatases (although it is significantly more potent against alkaline phosphatase). In cellular experiments with TCF-ALP, levamisole, an alkaline phosphatase selective inhibitor, was also used and minimal fluorescence was observed in C2C12 cells pretreated with levamisole prior to incubation with TCF-ALP [31]. These data support the assertion that this probe reports primarily on alkaline phosphatase activity in this cell line.

3. Peptide-based fluorophores

To develop probes which show increased selectivity towards a variety of different phosphatase families, peptide-based probes can be used. The sequences of these peptides are often obtained from known physiological substrates of the phosphatase of interest or discovered de novo through the use of peptide library screening methods, and thus can provide significantly more selectivity as compared to small molecule fluorescent probes. Frequently, the fluorophore is incorporated within the peptide sequence, although examples where the fluorophore is placed on a peptide terminus have been reported [37,38].

The diversity of this strategy in developing probes for different phosphatase families is best exemplified by the C-Sox (cysteine sulfonamide-oxine) fluorophore, which is typically placed into the peptide sequence two residues away from the phosphoresidue (Figure 2a) [39]. Upon coordination of a Mg²⁺ ion between the phosphate hydroxyl groups and the C-Sox fluorophore, a strong fluorescent signal is observed which decreases as the probe is dephosphorylated due to loss of Mg²⁺ coordination. Using this method, probes to measure serine/threonine, tyrosine, histidine, and arginine phosphatase activity have been reported [40–43]. While the serine/threonine phosphatase targeted probe was hydrolyzed well by PP2A, significant background signal was observed in HeLa cells depleted of PP2A, indicating that some other cellular activity contributes to a loss of signal from this probe. Likewise, both the tyrosine and histidine phosphatase targeted probes showed some selectivity towards PTP1B and PHPT1 respectively in vitro, but a significant decrease in background fluorescence was observed in cells lacking PTP1B or PHPT1. These data suggest that the C-Sox probes could be processed by off-target cellular phosphatases or some other cellular process.

Figure 2.

(a) Structures of C-Sox probes used to measure phosphatase activity. Upon dephosphorylation, Mg²⁺ binding is significantly reduced, leading to a decrease in fluorescence. (b) Structure of the CD45 selective probe SP1. Fluorescence resulting from probe dephosphorylation in RAW 264.7 cells in the absence (top) or presence (bottom) of the PTP inhibitor orthovanadate. Green fluorescence indicates cell nuclei. Number below each probe is the corresponding reference. Images adapted from reference 45. Copyright 2012 National Academy of Sciences

An alternative approach to the development of peptide-based phosphatase probes involves the use of a fluorogenic phosphotyrosine mimetic moiety in place of the native phosphoresidue. An example of this is the use of pCAP, an unnatural amino acid variant of the fluorogenic phosphatase substrate 4-methylumbelliferyl phosphate [44]. This fluorophore was used to produce a CD45 selective probe (SP1) based on the amino acid sequence surrounding tyrosine 394 of the kinase Lck, a known biological substrate of CD45 (Figure 2b) [45]. Screening of SP1 against a 7-member panel of cytosolic PTPs revealed some in vitro selectivity towards CD45. Monitoring SP1 hydrolysis in Jurkat T cells (CD45-expressing) and J45.01 T cells (CD45-null) revealed higher levels of fluorescence in the Jurkat cells, and partial knockdown of CD45 or introduction of a CD45 selective inhibitor showed a significant reduction in fluorescence, indicating that the probe is primarily hydrolyzed by CD45 in T cells. The probe was used to identify cell-permeable CD45 inhibitors via an in-cell screen [45]. In addition, SP1 was used to identify differences in CD45 activity in B cells isolated from patients with systemic lupus erythematosus as compared to healthy controls for the first time [46], providing a nice demonstration of the utility of a well-validated protein phosphatase targeted probe.

4. Near-IR fluorophores

In order to move beyond in vitro analyses and cellular imaging, several groups have been developing fluorogenic phosphatase probes with excitation and emission wavelengths in the near-IR (NIR) region of the spectrum. With appropriate excitation and emission profiles and sensitivity, NIR probes have the potential to be used in tissue and whole animal imaging experiments. Two main scaffolds have been used for this purpose in recent work; a dihydroxanthene-hemicyanine scaffold and a cyanine dye scaffold (Figure 3a & 3b). The dihydroxanthene-based probes in Figure 3a have fairly similar excitation and emission profiles and serve as sensitive reporters of alkaline phosphatase activity in vitro [47–52]. These probes report on phosphatase activity in cells, although the selectivity for alkaline phosphatase over other phosphatase activity was not explored in detail. Generally, alkaline phosphatase rich HeLa cells showed significantly higher fluorescence when incubated with the probes than alkaline phosphatase poor HEK 293 cells [47,49,53]. Probes based on a cyanine dye scaffold localized in the mitochondria, consistent with similar reports in the literature [53,54]. In addition, the NIR probes can be used in animal imaging experiments. For example, the probes NALP, LET-3 and QcyP were used to visualize alkaline phosphatase overexpressing tumor xenograft in nude mice [49,50,53]. Upon intratumoral injection of probe, significant fluorescence was observed in the tumor region, while mice that received intratumoral injections of vanadate prior to probe displayed significantly lower fluorescence. These studies indicate that NIR probes could ultimately be useful for imaging phosphatase-rich tissues.

Figure 3.

Structures of the (a) dihydroxanthene-hemicyanine probes and (b) cyanine dye probes. (c) Fluorescence imaging of HepG2 cells with QcyP (left), a mitochondria tracking dye (middle), and the merged images (right) demonstrating its mitochondria localization. Overlap between the two fluorescent probes is indicated by the yellow color. Images adapted from reference 53. (d) Near IR imagining of NALP up to 2 hours after injection into BALB/c mice with a tumor xenograph. Number below each probe is the corresponding reference. Images adapted from reference 49.

5. Conclusions

Several fluorogenic probes for monitoring intracellular phosphatase activity have been reported recently. Significant advances have been made in the development of probes with excitation and emission wavelengths in the visible and near-IR regions of the spectrum, providing greater compatibility with tissue and whole animal imaging. Peptide-based probes have been optimized to provide some selectivity for individual families of protein phosphatases and even for an individual member of an enzyme family. Despite these advances, significant challenges remain. The development and validation of selective substrates is particularly challenging. The selectivity of many of the probes described here for their intended phosphatase target(s) has not been established unambiguously. In addition, the peptide-based probes, which have the potential to be optimized for selectivity, utilize fluorophores that require UV excitation. Efforts to develop a well-validated, selective probe that can be excited with visible or near-IR irradiation are in progress and would be expected to have significant impact on the field.