Enzyme-Instructed Self-Assembly for Spatiotemporal Profiling of the Activities of Alkaline Phosphatases on Live Cells

Summary

Alkaline phosphatase (ALP), an ectoenzyme, plays important roles in biology. But there is no activity probes for imaging ALPs in live cell environment due to the diffusion and cytotoxicity of current probes. Here we report the profiling of the activities of ALPs on live cells by enzyme-instructed self-assembly (EISA) of a D-peptidic derivative that forms fluorescent, non-diffusive nanofibrils. Our study reveals the significantly higher activities of ALP on cancer cells than on stromal cells in their co-culture and shows an inherent and dynamic difference in ALP activities between drug sensitive and resistant cancer cells or between cancer cells with and without hormonal stimulation. Being complementary to genomic profiling of cells, EISA, as a reaction-diffusion controlled process, achieves high spatiotemporal resolution for profiling activities of ALPs of live cells at single cell level. The activity probes of ALP contribute to understanding the reversible phosphorylation/dephosphorylation in the extracellular domains that is an emerging frontier in biomedicine.

eTOC Blurb

Enzyme-instructed self-assembly (EISA), a multistep process that integrates enzymatic reaction and self-assembly, provide a facile approach for profiling the activities of alkaline phosphatases on live cells, thus revealing details of phosphorylation/dephosphorylation kinetics that are elusive from genomic analysis.

Introduction

Recent progress in the research of extracellular proteins has suggested that reversible phosphorylation/dephosphorylation in the extracellular domains is an emerging frontier for the development of biomarkers and next-generation drugs,¹ where much of the attention has been focused on ectokinases.² Ectophosphatases, however, as the integral part of kinase/phosphatase switch,^{3, 4} also play a critical role for reversible phosphorylation/dephosphorylation in the extracellular domains. For example, mammalian alkaline phosphatases (ALP), as membrane glycosylphosphatidylinositol (GPI)-anchored ectoenzymes, play a critical role from embryogenesis to cancer biology, to bone metabolism, and to neuron functions.^5–9 Thus, the spatiotemporal determination of the activities of ALPs of live cells would complement the genomic analysis and contribute to delineate details of dephosphorylation kinetics in cellular processes. However, there is little development of activity probes for imaging the ALPs of live cells largely because the catalytic domain of ALP locates outside the plasma membrane of cell and the fluorescent probes diffuse away easily. Several agents for detecting or imaging phosphatases still suffer serious caveats. The conventional used assay, ELISA, though being able to detect the activity of phosphatases, is unsuitable for live cell imaging. The fluorescent staining (ELF®97) agent, which bases on restricting bond rotation to generate fluorescent colloids,¹⁰ requires cell fixation,¹¹ or causes irreversible nucleated crystallization¹² and exhibits significant cytotoxicity¹³. The fusion of green fluorescent protein (GFP) to ALP can lead to mislocation ¹⁴ or losing the ALP activities.¹⁵ These facts underscore an unmet need to profile the activities of ALPs for understanding the functions of ALPs at the cellular level and in a spatiotemporal manner.

Our research of enzyme-instructed assembly (EISA) (Fig. 1a)^16–18 serendipitously reveals the selective formation of nanofibrils of a D-peptide around cancer cells due to the overexpressed ALP of cancer cells.¹⁸ The D-peptides, exhibiting low affinity to endogenous proteins and high proteolytic stability confer excellent spatiotemporal control to the EISA process,^19–21 which leads us to develop a phosphorylated and 4-nitro-2,1,3-benzoxadiazole (NBD) conjugated D-peptide (pNDP1) as an imaging probe of ALPs. Being dephosphorylated, pNDP1 forms NDP1 (Fig. 1b), which self-assembles to generate fluorescent, non-diffusive nanofibrils of NDP1 locally for imaging the activities of ALPs on live cells. Using the co-culture of cancer (HeLa) and stromal cells (HS-5), we confirm that EISA can reveal the generic difference between cancer and normal cells in terms of ALP activities (Fig. 1c, d) at single cell level. As a facile method for the globally spatiotemporal profiling the activities of ALPs on live cells, pNDP1 (i.e., the substrate for the EISA of NDP1) also reveals the difference of ALP activities between two pairs of drug sensitive/resistant cancer cell lines (e.g., A2780 and A2780cis, MES-SA and MES-SA/dx5) and a cell line (e.g., MCF-7) under hormonal stimulus. Moreover, the self-assembled nanofibrils of NDP1, being largely innocuous to cells, allow reading out the activities of ALPs in a wide range of cell lines. This work illustrates that EISA, in addition to controlling cell fate,^{18, 22} promises an effective approach for spatiotemporal imaging of the activities of phosphatases on a wide range of live cells for revealing dephosphorylation kinetics that are beyond genomic and proteomic analysis.

Figure 1. EISA forms fluorescent, non-diffusive nanofibrils with high spatiotemporal resolution.

(a) Illustration of the concept of EISA for generating fluorescent nanofibrils. (b) Molecular structures of pNDP1 and NDP1, and the conversion catalyzed by ALP under physiological condition. (c) Illustration of the EISA of NDP1 to form fluorescent pericellular nanofibrils in co-culture. (d) Fluorescent nanofibrils of NDP1 selectively form on HeLa cells (the cancer cell, pointed by white arrows) and in HS-5 cells (the stromal cell, pointed by white arrow heads) in the co-culture of HeLa and HS-5 ([pNDP1] = 500 μM, 0 hour (left), 6 hour (middle) and 24 hour (right) incubation, scale bar = 10 μm, the nuclei stained by Hoechst 33342).

Results

Molecular Design

Based on that the collagen fibril assembly in which enzymes cleave propeptides to convert procollagens to collagen molecules that self-assemble into collagen fibrils²⁴ and that NBD exhibits enhanced fluorescence upon the formation of nanofibrils,^{22, 23, 25} we develop the precursor (pNDP1, Fig. 1b) consisting of two parts—a fluorophore (NBD) and a phosphorylated, self-assembling D-peptide (i.e., D-3-(2-naphthyl)alanine (D-Nal)-D-Phe-D-Phe-D-_pTyr). The aromatic groups of the D-Nal and D-Phe residues provide sufficient hydrophobic interactions to enable the self-assembly of NDP1 to form the nanofibrils in water; _pTyr allows ALP to convert pNDP1 to NDP1. The D-peptide backbone ensures the proteolytic stabilities of pNDP1 and NDP1 in cellular environment. The phosphate (besides as an enzyme responsive trigger) and the terminal carboxylic acid render the conjugate better solubility in water. Based on this design, after coupling β-alanine with NBD-Cl to obtain the NBD-NH(CH₂)₂COOH, we prepare pNDP1 by solid phase peptide synthesis (SPPS)²⁶ (Fig. S1a), and obtain pNDP1 in 60% yield (Fig. S1a). As expected, ALP catalytically removes the phosphate and triggers the self-assembly of the NDP1 to form localized nanofibrils (Fig. 1b, c), which exhibit high fluorescence and grow with time for imaging the activities of ALPs of cells.

Compared to our previous work that exploits the concept of EISA for imaging intracellular self-assembly²³, the probe pNDP1 used in this manuscript, though utilizing NBD, differs with the previous probe in several aspects. First, they are different molecules with different stereochemistry. While the previous one based on L-peptide, the work employs D-peptide. Secon, the L-peptidic precursor enters cell, undergoes EISA, and forms intracellular fluorescent nanofibrils. But the D-peptidic precursor hardly gets into the cells and mainly undergoes EISA outside the cells. Third, the D-peptidic precursor discussed in this paper is meant to be a probe for ectophosphatases, which are membrane enzymes with catalytic domain outside the membrane.

EISA in a Cell-Free Condition

We first use ALP to verify enzymatic transformation of pNDP1 to NDP1 in a cell-free condition. ALP (2 U/ml), added into the solution of pNDP1 (1.0 wt%, pH 7.4), dephosphorylates pNDP1 to afford NDP1, which self-assembles in water to form a hydrogel (Fig. 2a, b). Transmission electron microscopy (TEM) images reveal that the hydrogel of NDP1 consists of uniform self-assembled nanofibrils with diameter of 9 ± 2 nm, entangling with each other to form a network as the matrices of the hydrogel (Fig. 2b). The rheological test (Fig. 2c) shows the kinetic process of hydrogelation and indicates the hydrogelation point appears at 12 min after the addition of the enzyme (ALP, 2 U/mL). The gelation point appears when only 9.0% of pNDP1 being dephosphorylated at room temperature, which indicates the strong self-assembly ability of the peptide backbone of NDP1. Static light scattering (SLS) reveals that the signal intensity of the solution of pNDP1 (20 μM) increases dramatically after the addition of ALP (2 U/mL, Fig. 2d), suggesting that the conversion of pNDP1 to NDP1 even at a concentration as low as 20 μM results in the self-assembly of NDP1. As a useful and sensitive assay for reporting the formation of the nanofibrils via self-assembly, the strongly fluorescent spots (Fig. 2e) from the resulted NDP1, at 20 μM, also indicate the self-assembly of NDP1 to form nanofibrils, confirmed by TEM showing uniform nanofibrils (Fig. 2f) in the same mixture. TEM of the dried solution of pNDP1 reveals amorphous aggregates (Fig. S1b), suggesting that it is plausible that the pNDP1 likely forms micellar aggregates in aqueous solution.

Figure 2. ALP catalyzed molecular self-assembly to form fluorescent nanofibrils.

(a, b) Transmission electron microscopy (TEM) images of pNDP1, (a) before and (b) after the treatment with ALP (2U/mL) at the concentration of 1.0 wt% and pH of 7.4. Inset is the optical images of the solution and hydrogel, respectively. The scale bar is 100 nm. c) Rheological characterization of the hydrogelation process by treating the solution of pNDP1 with ALP (2U/mL), at the concentration of 1.0 wt% and pH of 7.4. The gelation point (the storage modulus (G′) dominates loss modulus (G″)) appears at 12 minute after the addition of ALP. (d, e, f) Enzyme-catalyzed self-assembly of NDP1 at 20 μM (below its minimum gelation concentration): (d) Static light scattering (SLS) signals of pNDP1 without and with the addition of ALP (2 U/mL) to the solution of pNDP1 (20 μM, pH 7.4). (e) The confocal fluorescent microscope image shows the appearance of bright spots in the solution of pNDP1 (20 μM, pH 7.4) after the treatment of ALP (20 U/mL). The scale bar is 100 μM. (f) TEM image of NDP1 formed after the treatment of pNDP1 (20 μM, pH 7.4) by ALP (2 U/mL). Scale bars are 100 nm, and inset scale bar is 50 nm.

Correlative Light and Transmission Electron Microscopy (CLEM) Imaging of the Nanofibrils on Cells

We use correlative light and transmission electron microscopy (CLEM)²⁷ of high-pressure frozen substituted HeLa cells to confirm the formation of nanofibrils on the cellular ultrastructural features. We choose an incubation concentration of 500 μM to achieve a superior spatial resolution since 20 μM of pNDP1 is insufficient to generate pericellular fluorescent nanofibrils rapidly and adequately on HeLa cells. The growth of the cells on Aclar discs marked with a pattern allows tracking cells of interest throughout the light microscopy and EM sample preparation. For large overviews of the cells, we acquire montages of overlapping images in an automated fashion using the microscope control software SerialEM. Fig. 3a is the merged confocal microscopy image (DIC and fluorescence images) of a HeLa cell of interest, which is treated with pNDP1 (500 μM) for 6 hours and exhibits strong yellow fluorescence on its surface afterwards. Fig. 3b shows montages of the same HeLa cell, revealing the appearance of nanofibrils at the exact location where there is fluorescence in Fig. 3a. Enlarging the boxed area in Fig. 3b makes it easier to identify the nanofibrils (white arrowhead, Fig. 3c). The nanofibrils appear short in length because they are visualized on a section of cell with a thickness of only 70 nm. Tomograph of the same section and 3D reconstruction of the tomograph (in red box of Fig. 3c) confirm the formation of nanofibrils (Fig. 3d, e). The structural comparison between the untreated (as the control, Fig. 3f, g, h) and the treated HeLa cells (Fig. 3a–e) further verifies that the nanofibrils, only presented on treated cells, are formed by EISA of pNDP1.

Figure 3. CLEM confirms pericellular fluorescent nanofibrils on HeLa cells.

(a) Merged image (DIC and fluorescence microscopy images) of treated HeLa cells (being incubated for 6 h with 500 μM of pNDP1 and growing on Aclar plastic film), recorded only a few minutes before the sample was high-pressure frozen; note the highest intensity of fluorescence signal on the cell surface, indicating a high abundance of nanofibrils in that region of the cell. (b) High magnification EM of the whole cell in (a). Note that it is a “montage image” that consist of more than 200 individual high-magnification image tiles. (c) Higher magnification electron micrograph of the region in red box of the same cell in (a) and (b). The region in red box of the cell corresponds to the area of highest fluorescence signal in (a). (d) High magnification electron micrographs of the red boxed area in (c). Note the presence of nanofibrils (white arrowheads), containing NDP1s. (e) 3D-rendered tomograph of the section in (d) with the thickness of around 70 nm. Scale bars: 2000 nm (a, b), 200 nm (c, d). (f) High magnification electron micrograph of the whole cell (wide type, untreated cell); Note that it is a “montage image” that was stitched together from more than 200 individual high-magnification image tiles. (g) Higher magnification electron micrograph of the region in blue box of the same cell in (f). (h) High magnification electron micrographs of the blue boxed area in (g); Note the absence of nanofibrils. Scale bars: 2000 nm (f), 500 nm (g) and 200 nm (h).

Imaging the Activity of ALPs in Co-culture

After using CLEM to confirm that EISA forms pericellular nanofibrils of NDP1 on a single cell, we examine the use of EISA for profiling the activities of ALP on live cells in a co-culture. As shown in Fig. 1c, d, in the co-culture of cancer (HeLa) and stromal cells (HS-5), yellow fluorescence grows on the surface of the cancer cell (e.g., HeLa) only with extending incubation time, indicating the significant activities of ALPs on HeLa cell surface. In contrast, rather than on the cell surface, only yellow puncta present inside HS-5 cells, likely due to dephosphorylation in endosomes. This result agrees with the overexpression of ALP on HeLa cells,⁸ and reveals low expression level of ALP on the surface of HS-5 cells. The formation of small amounts of fluorescent puncta inside HeLa and HS-5 cells agrees with the presence of phosphatases intracellular organelles²⁸ and the macropinocytosis of the pNDP1 (or NDP1) (Fig. S2a). The fluorescent nanofibrils on HeLa cells grow significantly with increasing incubation time, while the yellow fluorescence in HS-5 cells hardly changes, which is consistent with the higher metabolic rate of HeLa than that of HS-5 cells. Interestingly, there are fluorescent puncta threading between HeLa and HS-5 cells, an observation that coincides with ALP-containing exosomes for intercellular communication.^{29, 30} These results confirm that not only the conversion of pNDP1, as a substrate of phosphatases, is able to profile the activities of ALP on cells, but the self-assembly of NDP1 achieves excellent spatiotemporal resolution to monitor the cellular locations, the communication and activities of phosphatases in mixed population of different types of cells.

Imaging the Activity of ALP on Different Cell Lines and Under Different Conditions

We examine the ALP expression of several pairs of cell lines and under different conditions. Before imaging, we use MTT cell viability/proliferation assay to examine the cell compatibility of pNDP1 within 24 hours at different concentration (Fig. S2b) and confirm pNDP1, at 500 μM, is innocuous to the cells (except Saos-2 and 7F2 cells). Fig. 4a shows the time-dependent accumulation of the fluorescent nanofibrils on the HeLa cells incubated with pNDP1: at 30 min, there is hardly any fluorescence both around and inside the HeLa cells; at 1 h, sporadic bright yellow fluorescence emerges only on the cell surfaces (covering about 10% cell surface); at 3 h, in addition to increased fluorescent nanofibrils outside the cells (covering about 25% cell surface), yellow fluorescent puncta start to present inside cells; at 6 h, fluorescent nanofibrils have covered most area of the cell surfaces (about 75% cell surface). 3D stacked z-scan image of the HeLa cells confirms that fluorescent nanofibrils cover most area of cell surface (Fig. 4b, see Movie 1), the intracellular fluorescent puncta, however, remain almost constant between at 3 h and at 6 h. This selective formation of fluorescent nanofibrils on the HeLa cells, which grow with increasing incubation time, agrees with the result in co-culture (Fig. 1d) and renders an excellent spatiotemporal resolution. After re-seeding the cells (after 6 h incubation with pNDP1) in fresh complete growth medium for another 36 hours, there is hardly any obvious fluorescence on HeLa cells (Fig. 4c), indicating the reversible nature of the nanofibrils of NDP1 formed by non-covalent interaction. This unique reversibility minimizes long-term toxicity of the nanofibrils, thus warranting pNDP1 as a cell compatible activity probe for imaging the activities of ALPs on live cells. Moreover, after pre-treating HeLa cells with (−)-tetramisole (40 μM), an inhibitor of tissue-non-specific ALP (ALPL)³¹, and imaging the cells incubated with pNDP1 together with (−)-tetramisole (40 μM) for 3 hours, we found little difference (Fig. S2c) in the images of HeLa cells with and without the addition of (−)-tetramisole.

Figure 4. EISA for profiling the activities of ALP on live cells.

Confocal microscope images (a-e) show (a) the time course of fluorescence on the HeLa cells incubated with pNDP1 for 30 min, 1 h, 3 h, and 6 h, respectively. Blue pseudocolor = Hoechst 33342, (b) 3D stacked z-scan images of fluorescence on HeLa cells, treated with pNDP1 for 6 h, (c) no fluorescence on HeLa cells after re-incubating the cells (from (a), 6 h) in a fresh medium for 36 hours, (d) the fluorescence emission on two pair of drug-sensitive/resistant cancer cell lines (A2780, and A2780cis, MES-SA, and MES-SA/Dx5, 24 h), incubated with pNDP1, and (e) the fluorescence on cancer cells (MCF-7) incubated with pNDP1, without (upper) and with (bottom) the addition of prednisolone. Scale bar: 20 μm (a/upper row, c), and 5 μm for (a/lower row, d, e). (f) Quantification of ALP on HeLa cells according to the amount of pericellular nanofibrils by measuring the fluorescence intensity of the molecules remaining on the cell surface. (g) The plot of time versus the fluorescence of pericellular nanofibrils on HeLa cells, treated with pNDP1. The initial cell number is 5000 cells/well. (h) The confocal images (scale bar = 10 μm) of different cell lines incubated with pNDP1. The incubation time is 24 hour for HS-5, PC3, U-87 MG, Capan-2, A375, SKOV3, PC-12 Adh, T98G, and 12 hour for Saos-2 and 7F2. The incubation concentration of pNDP1 is 100 μM for Saos-2 and 7F2, and 500 μM for other cells. All the confocal images were taken after removing the medium with pNDP1 and then adding the live cell image solution (Invitrogen Life Technologies A14291DJ).

To further demonstrate the applicability of EISA as a process for profiling the activities of ALP and to gain previously unknown insights on cancer biology, we did imaging of two pairs of drug sensitive/resistant cancer cell lines (e.g., A2780 and A2780cis, MES-SA and MES-SA/dx5) with pNDP1. As shown in Fig. 4d, drug sensitive cancer cell A2780 has almost 100% coverage with fluorescent nanofibrils while much less fluorescent spots present around A2780cis, a platinum-resistant ovarian cancer cell. The EISA results suggest that A2780 and A2780cis express different levels of ALP, which is confirmed by Western blot of ALPL and ALPP of these two cell lines (Fig. S2d). After being incubated with pNDP1, the other pair of drug sensitive/resistant cell lines (i.e., MES-SA and MES-SA/Dx5) exhibit in about 10% cell surface covered by fluorescent nanofibrils, suggesting that they express similar levels of ALPs. The EISA results suggest that MES-SA and MES-SA/Dx5 express almost same levels of ALP, which is confirmed by Western blot of ALPL and ALPP of these two cell lines (Fig. S2e). We also checked the ALP expression of a cancer cell line (MCF-7) under hormonal stimuli (e.g., prednisolone³²). As shown in Fig. 4e, MCF-7 cells generate a relatively moderate amount of fluorescent nanofibrils to cover about 50% of cell surface. After the MCF-7 cells being incubated with prednisolone (0.5 μg/mL) followed by the addition of pNDP1 24 hours later, thicker fluorescent nanofibrils cover almost 100% area of the cell surfaces. Although the mechanism of this increase of enzyme activity is not known, there are several possibilities such as an increased rate of enzyme synthesis, a decreased rate of enzyme degradation, or a conformational change in the enzyme leading to an increased activity of the individual enzyme molecules. An early study³³ favors the last mechanism, which is consistent with the Western blots of MCF-7 showing the expressions of ALPL (or ALPP) (Fig. S2f), with and without the treatment of prednisolone (0.5 μg/mL), to be almost the same. These results unambiguously validate the suitability of EISA, as a process, for imaging the activities of ALPs on live cells.

To verify quantitatively that the amount of fluorescent nanofibrils formed by EISA of NDP1 is proportional to the quantity of ALPs, we simply determine the amount of nanofibrils by measuring the fluorescence intensity of the molecules remaining on the cell surfaces using a plate reader. Fig. 4f shows that the fluorescence intensity, which is proportional to the amount of nanofibrils, has an excellent linear correlation with cell numbers, which corresponds to the overall amounts of ALPs. Besides, incubated with prednisolone, HeLa cells result in two folds increase of fluorescence, agreeing with that prednisolone increases the expression of ALP³². In addition, the measurement the fluorescence emission of pericellular nanofibrils versus time on HeLa cells shows that the amount of nanofibrils increases with the incubation time before leveling off at about 6 hours. This result further confirms the temporal resolution of EISA. The accumulation of nanofibrils on cell surface slows down after incubation for a certain time because the formed nanofibrils likely block the enzyme catalytic sites. This level-off may explain the low cytotoxicity of this probe even after 24 hour. This result also differs from other D-peptide precursors in our previous work, which killing cells^{18, 22}. Such a difference may originate from the subtle difference of the fibrils formed by the different molecules.

As a general approach, EISA allows the evaluation of the activities of ALP on other types of cells at the single cell level (Fig. 4h). A separated incubation of HS-5 cells with pNDP1 reveals that only faint yellow spots present inside the HS-5 cells even after 24 hours of incubation (Fig. 4d), confirming that HS-5 cells have a low level expression of ALP. This result, thus, validates the overexpression of ALP as a generic difference between HeLa (cancer) and HS-5 (stromal) cells. PC-3 (human prostatic adenocarcinoma) only shows yellow spots in cytoplasm with few nanofibrils around the cell, indicating low level of ALP expression and agreeing with the association of acid phosphatases with prostate cancer.³⁴ U-87 MG (human glioma), Capan-2 (human pancreas adenocarcinoma), A375 (human melanoma), and SKOV3 (a drug resistant ovarian carcinoma) exhibit weak fluorescence on their surfaces, indicating low expression level of ALPs. PC-12 Adh (rat pheochromocytoma) has a moderate expression of ALP (about 50% of its surface exhibiting fluorescence). T98G, a human glioblastoma multiforme cell, exhibits strong yellow fluorescence on its surface, indicating overexpression of ALP. Notably, the incubation of pNDP1 (100 μM) with Saos-2, an osteosarcoma cell line known to overexpress ALP, not only results in the fluorescent nanofibrils completely covering the cell surfaces in 12 h, but also need less pNDP1 and take shorter time than those used for the other cells. Similarly, the 12 h of incubation of pNDP1 (100 μM) with 7F2, an osteoblast from bone marrow of mouse, leads to complete coverage of cell surface by the fluorescent nanofibrils. Western blot analysis largely agrees with the imaging result (Fig. S2g).

Discussion

EISA as a Complementary Approach of Existing Techniques

Despite the existence of a range of standard techniques, such as ELISA, Western blots (WB), genomic profiling, and flow cytometry for profiling enzymes or proteins, EISA uniquely addresses the shortcoming of those methods for revealing the activities of enzymes of live cells. For example, ELISA and WB can only show the amount of enzyme, not the enzyme activities. Although one can obtain membrane protein by using membrane protein extraction kit, such an isolation method is far from perfect since it is impossible to completely separate cytosolic proteins and membrane proteins.³⁵ Gene profiling measures the expressions of genes to create a global picture of cellular proteins, but it provides similar information that can be obtained by ELISA and WB, at probably higher cost. As for flow cytometry method, the cell needs to be labeled by fluorescent tag, typically by antibody, before measurement. Flow cytometry is more suitable for a population of cells than for single cells. The result of cytometry is unable to indicate the activities of enzyme as well as enzyme locations. Moreover, these methods are less useful for live cell experiment at single cell or subcellular resolution.

Contrasting to the conventional methods, EISA provides a “spatiotemporal” approach to visualize the enzyme activities on live cells at single cell or subcellular level, a kind of information that was unavailable previously. Because the non-diffusive fluorescent nanofibers are formed locally where EISA occurs, which inherently represents the location of enzymes. Since it is feasible to capture the growth of fluorescence of non-diffusive nanofibers, EISA can easily reveal the activities of the enzymes, even at subcellular location. For example, EISA represents a straight-forward way to visualize (exclusively) the activities of ectophosphatase, even on a single cell and not limit to the same protein amount. Moreover, EISA is applicable for both single cell imaging and a population of cell imaging (Fig. 4a). Single cell imaging is particularly important, since bulk measurements on a population of cells mask single-cell responses and therefore regularly fail to accurately quantify biological processes or identify rare events, especially in tumor microenvironment. For the case of ALP reported here, EISA can quantify ALP activities of a population of cell because it can be read out by using a fluorescent plate reader (Fig. 4f, g). Compared with antibodies, pNDP1 is much cheaper and extremely easy to make (in gram scale). EISA also is compatible with live cells if cell viability is critical for a study. While EISA unlikely intends to replace the well-established techniques (e.g., ELISA, WB, genomic profiling, and flow cytometry), it is a novel method to supplement to the existing techniques for obtaining more comprehensive and dynamic cellular information related to the activity of enzymes.

Though the true dynamic monitoring requires better imaging instrumentation, kinetics and spatial distribution are important advantages of EISA. For example, the time-dependence data (Fig. 4a) support the detection of kinetics; the co-culture data confirm the ability to reveal spatial distribution (Fig. 1d).

A Structural Analog of the Probe

We also design molecule pNDP2 as a control of pNDP1 to establish that the enhanced fluorescence reveals effective self-assembly of the probes. Unlike pNDP1, pNDP2 (without the naphthyl residue) is less prone to self-assembly, thus offering a reference to assess the significance of molecular self-assembly for imaging the activities of the ALPs. pNDP2 (Fig. S3a) easily dissolves in water under the same condition (1.0 wt%, pH 7.4) used for pNDP1. However, missing one naphthyl residue compared with NDP1, NDP2 has weaker capability to self-assemble in water and only form a soft hydrogel after the addition of ALP (2 U/mL) in the solution of pNDP2 (Fig. S3b). TEM images reveal that long, flexible and uniform nanofibrils (d = 9 ± 2 nm) interweave with the aggregates in the solution of pNDP1 while only aggregates show in the solution of pNDP2. This difference suggests that enhanced hydrophobic interaction conferred by an additional naphthyl residue favors self-assembly. After enzymatic hydrogelation triggered by ALP (2U/ml), both the hydrogels of NDP1 and NDP2 consist of uniform self-assembled nanofibrils with diameters of 9 ± 2 nm and 7 ± 2, respectively (Fig. S3b). The morphological differences suggested by optical and TEM images indicate that the naphthyl residue increases intermolecular aromatic-aromatic interactions of both precursors and hydrogelators, which is further verified by rheological characterization (Fig. S3c). Both hydrogels exhibit viscoelastic properties of a solid-like materials because their storage moduli (G′) are higher than their loss moduli (G″) and are independent to oscillation frequency. The hydrogel of NDP1 shows stronger mechanical strength than the hydrogel of pNDP2, as evidenced by a significantly higher value of G′ in both strain and frequency dependent oscillations. Most importantly, pNDP2 is unable to form fluorescent nanofibrils on cell surface upon enzymatic transformation (Fig. S3d, also see Movie 2). Interestingly, several fluorescent puncta forms inside the cells (Fig. S3d), which is similar to the case of pNDP1 treated HS-5 cells. This result validates that, in order to form nanofibrils on cell surface, the self-assembly needs to occur quickly, either by more ALPs or by enhanced self-assembly ability of the fluorescent D-peptides. These results not only warrant pNDP1 as an excellent candidate for imaging the activities of ectophosphatases on live cells, but also provide insights for optimizing the activity probes.

Specificity of ALPs over other Phosphatases

ALP (alkaline phosphatase), as implied by name, are most effective in an alkaline environment while pNDP1 only dissolve in slightly basic condition to form oligomers or monomer to be dephosphorylated. The use of D-peptide, in fact, minimizes the promiscuity of pNDP1 because the natural substrates of protein tyrosine phosphatases (PTPs) or protein phosphatases (PPs) are made of L-peptides. Indeed, our result shows ALP dephosphorylates pNDP1 much faster than PTP1B or PP1 does (Fig. S3e), thus establishing the specificity of pNDP1 to ALPs.

Intracellular Puncta

Though pNDP1 mainly acts as an excellent probe for ectophsophatases, yellow fluorescent puncta form inside the cells, especially in HS-5 cells. We thus use inhibitors of different endocytotic processes (filipin III for caveolae-mediated endocytosis, chlorpromazine for clathrin-mediated endocytosis, ethylisopropylamiloride (EIPA) for macropinocytosis) to determine the pathways through which pNDP1 or NDP1 enters cells. The confocal microscopy images of HS-5 cells treated by different inhibitors show that only the addition of EIPA effectively reduces the yellow puncta inside cells, while the presence of both filipin III and chlorpromazine affects little (Fig. S2a). This phenomenon indicates that the live cells likely uptake pNDP1 or oligomers of NDP1 via macropinocytosis.

Imaging with Inhibitor and Antibody

As ectophosphatases, ALPs have different isotypes—one tissue-non-specific alkaline phosphatase (ALPL), and three tissue-specific alkaline phosphatases (i.e., placental (ALPP), germ cell-specific (ALPP2), and intestinal (ALPI)). The aberrant expression and abnormal activity of phosphatases associate with a variety of human diseases, including cancers ^{8, 36–38}. Although the overexpression of ectophosphatases has been implicated in the formation of cancerous cells for about fifty years ⁸, as an inevitable consequence of genomic instability of cancer ³⁹, different cancer cells express different proteins at different levels. Therefore, different cancer cells express different levels of the isozymes of ALPs. Because the ALPL inhibitor, (−)-tetramisole, hardly inhibit the formation of fluorescent nanofibrils on HeLa cells (Fig. S2c), we expect that the isozymes of ALPs on HeLa cells unlikely are ALPL. In addition, Saos-2 cells, a cell known to express a high level of ALPL, result in large amount of pericellular nanofibrils, indicating that pNDP1 is also an excellent substrate of ALPL. To determine the expression of ALPP and ALPL on cancer cells, we use western blot to examine ALPP and ALPL on cell membranes of HeLa and Saos-2 cell line, with the stromal cell line (HS-5) as a control. As shown in Fig. S4a, among the three cell lines, only HeLa cells express high level of ALPP on its cell membrane; the other cells (HS-5 and Saos-2) express low levels of ALPP on their membranes. Besides, HeLa and HS-5 express significantly lower levels of ALPL on their membranes than Saos-2 do. Western blot also indicates that the total membrane ALP expression level is the highest on Saos-2 cells, followed by HeLa, and HS-5 has the lowest ALP levels (Fig. S4a). Besides, the high levels of ALPP on HeLa cell and ALPL on Saos-2 cell, respectively, are also evidenced by heat map of ALPL, ALPP and ALPI of BioGPS cell line gene expression profiles⁴⁰ extracted from database Harmonizome (Fig. S4b). The high level expression of ALPL on Saos-2 cells also leads to a rapid accumulation of nanofibrils formed by EISA of pNDP1 on their surface and causes cell death at high concentration within 24 hours (Fig. S2b). The above results confirm that different cancer cells express different levels of ALPP and ALPL. However, western blot reveals considerable amount of ALPL on HeLa cell surface, indicating the low inhibitory efficiency of (−)-tetramisole.

Based on the Western blot results that HeLa cells mainly express ALPP and Saos-2 cells express high level of ALPL, we use inhibitors of ALPP and ALPL (L-phenylalanine,⁴¹ L-Phe (3 mM) and 2,5-dimethoxy-N-(quinolin-3-yl)benzenesulfonamide,⁴² DQB (2 μM)), respectively, in the cell culture of HeLa and Saos-2 to determine the effects of the phosphatase inhibitors on the imaging. The addition of ALPP inhibitor (L-Phe) barely reduces the formation of fluorescent nanofibrils on HeLa cell surface, compared with control cells treated by pNPD1. This result agrees with that L-Phe is an ineffective uncompetitive inhibitor of ALPP (with the IC₅₀ of 20 mM). Contrary to the case of ALPP, ALPL inhibitor (DQB), indeed, decreases the formation of the nanofibrils of NPD1 on Saos-2 cells, indicating that it is the EISA of pNPD1 to result in fluorescence. Since HeLa cell also has considerable amount of ALPL on its surface, the treatment of DQB (2 μM) reduces the fluorescence on cell surface (Fig. S4c). This observation agrees with that DQB is an efficient ALPL/TNAP inhibitor with an IC₅₀ of 0.19 μM. We also test several other inhibitors of phosphatases. Homo-arginine results in little change of the formation of fluorescent nanofibrils. While causing cell death at high concentration (>100 nM), okadaic acid⁴³, at 10 nM, hardly change the formation of fluorescent nanofibrils on cancer cells, agreeing with that okadaic acid is a specific inhibitor of serine/threonine phosphatase.⁴⁴ In addition, we find that the addition of the antibody of ALPP or ALPL hardly affects the formation of the fluorescent nanofibrils, indicating that these antibodies have little influence on enzyme activities of ALPs. These results indicate that these activity probes may help screening the inhibitors of ALP by cell based assays.

Merits of D-peptide

In our previous study^{45, 46}, we unexpectedly found that enzymatic dephosphorylation of a tyrosine phosphate by ALP depends little on the chirality of the precursors. As the enantiomers of the naturally occurring L-peptides, D-peptides usually resist endogenous proteases, are less prone to cellular uptake, and interact minimally with endogenous cellular components. These features not only guarantee pNDP1 as an excellent ectophosphatase probe, but also make EISA of D-peptides suitable for long-term biomedical applications. Moreover, pNDP1 acts as a general probe for imaging the activity of ALPs. Although the expression of certain isozymes of ALPs is a specific feature of certain cancer cells, the expression level and/or the subtypes of ALPs likely would be dynamic. That is, they may change upon the biological cues. Therefore, a probe for revealing all the isozymes of ALPs should be an essential step to general a more comprehensive activity profiles of ALPs. Such information would be essential for developing specific activity probes for an isozyme of ALPs. The general probes of ALPs also may serve as a starting point to design specific probes for isozymes of ALPs. In addition, this research will contribute to design activity probes that have other modality (e.g., radioactive) for selectively imaging of ALPs, thus improve tumor detection and precision medicine.

In conclusion, this work illustrates a novel probe based on EISA for profiling the activities of ALPs of cells. The activity profiling of ALPs may contributes to the understanding of the reactome of cells, especially the cell signaling process controlled by ectokinase/ectophosphatases, an emerging research frontier in translational medicine¹ and cell biology². Moreover, one noteworthy observation of this work is that the cancer cell lines (e.g., A2780 and T98G) exhibiting high activity ALP coincide with tumors having low immunohistochemical staining of PD-L1 (e.g., ovarian carcinoma and glioma),⁴⁷ implicating that ALPs may help tumor evade immune response via purinergic signaling.⁴⁸ In addition, this work provides an imaging probe for screening ALPs on other cancer cells, which may lead to useful clues for determining cancer stem cells, which remain controversial due to inadequate criteria.⁴⁹

Experimental Procedures

Materials and Instruments

ALP was purchased from Biomatik, NBD-Cl from TCI and amino acids from GL Biochem, PP1 from Sigma-Aldrich, PTP1B from Abcam. All the solvents and chemical reagents were used directly as received from the commercial sources without further purification. All the products were purified with Water Delta600 HPLC system, equipped with an XTerra C18 RP column and an in-line diode array UV detector. Rheological data were obtained on TA ARES G2 rheometer with 25 mm cone plate, confocal microscopy images on Leica TCS SP2 spectral confocal microscope. Electron microscopy imaging was performed on a FEI Morgagni 268 TEM with a 1k CCD camera (GATAN, Inc., Pleasanton, CA, USA) or a 300 keV Tecnai F30 intermediate voltage TEM (FEI, Inc., Hillsboro, OR, USA) with a 4k CCD camera (GATAN).

Synthesis

Both the precursors pNDP1/2 and the hydrogelators NDP1/2 were prepared by the standard solid-phase peptide synthesis (SPPS),²⁶ which used 2-chlorotriyl chloride resin (100–200 mesh and 0.3–0.8 mmol/g) and N-Fmoc-protected amino acids with side chains properly protected. Before SPPS, Fmoc-D-Tyr(PO₃H₂)-OH, Fmoc-D-(3-(2-Naphthyl))-Ala-OH and NBD-NH(CH₂)₂COOH, which were directly used in SPPS, were prepared from D-Tyr-OH, D-(3-(2-Naphthyl))-Ala-OH and NBD-Cl based on previous work⁵⁰ and literature⁵¹. Fig. S1a illustrates the synthetic procedure of pNDP1.

Light Scattering

The static light scattering (SLS) experiments were performed using an ALV (Langen, Germany) goniometer and correlator system with a 22 mW HeNe (λ = 633 nm) laser and an avalanche photodiode detector. Before testing, all samples were filtered with 0.45 μm PTFE filters after sonicating and heating. The SLS tests were carried out at room temperature, and the angles of light scattering detection were 30, 60, 90, and 120°. The resulting intensity ratios are proportional to the amount of aggregates in the samples.

Transmission Electron Microscopy (TEM)

Negative staining technique was used in TEM imaging. The 400 mesh copper grids (#1200211, Spi Supplies) coated with continuous thick carbon film (~ 35 nm) were glowed prior to use in order to increase the hydrophilicity. After being loaded on the grid, samples (4 μL) were rinsed by dd-water for twice or three times. Immediately after rinsing, the grids containing sample were stained with 2.0 % w/v uranyl acetate for three times. The grids were allowed to dry in air prior to imaging.

Cell Culture

All cell lines were purchased from the American Type Culture Collection (ATCC) or Sigma-Aldrich and were propagated in a fully humidified incubator containing 5% CO₂ at 37°C. The HeLa and T98G cells were propagated in Minimum Essential Media (MEM, #11095080, Gibco™ Life Technology) supplemented with 10% fetal bovine serum (FBS, #10438026, Gibco™ Life Technology) and 1% antibiotics (Penicillin Streptomycin, #15140122, Gibco™ Life Technology). The HS-5 and A375 cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, #11965092, Gibco™ Life Technology), supplemented with FBS to a final concentration of 10% and 1% antibiotics. The A2780 cells were propagated in RPMI-1640 Medium (ATCC® 30–2001™), supplemented with 10% FBS, 2 mM glutamine. The A2780cis cells were propagated in RPMI-1640 medium, supplemented with 10% FBS, 2 mM glutamine and 1 μM cisplatin was necessary every 2–3 passages. The MES-SA, MES-SA/dx5, Capan-2, and SKOV3 cells were propagated in McCoy’s 5A (#16600108, Gibco™ Life Technology) supplemented with 10% FBS, and 1% antibiotics. PC-3 cells were propagated in F-12K Medium (ATCC®30–2004™), supplemented with 10% FBS. MCF-7 cells were propagated in MEM, supplemented with 10% FBS, and 0.01 mg/mL human recombinant insulin. PC-12 Adh cells were propagated in F-12K Medium, supplemented with FBS to a final concentration of 2.5%, horse serum to a final concentration of 15%. The Saos-2 cells were propagated in McCoy’s 5A supplemented with 15% FBS, and 1% antibiotics. The 7F2 cells were cultured in Alpha minimum essential medium (α-MEM, #12561049, Gibco™ Life Technology) with 2 mM l-glutamine and 1 mM sodium pyruvate without ribonucleosides and deoxyribonucleosides, 90%, FBS, 10%.

MTT Cell Viability/Proliferation Assay

Cells in exponential growth phase were seeded in a 96 well plate (#087722C, Fisher HealthCare) at a concentration of 1×10⁴ cell/well. The cells were allowed to attach to the wells for 24 h at 37 ºC, 5% CO₂. The culture medium was removed and 100 μL culture medium containing compounds (immediately diluted from fresh prepared stock solution of 10 mM) at gradient concentrations (0 μM as the control) was placed into each well. After culturing at 37 ºC, 5% CO₂ for 24 h, each well was added with 10 μL of 5 mg/mL MTT ((3-(4, 5-DimethylthiazoL-2-yl)-2, 5-diphenyltetrazolium bromide, #M5655, Sigma-Aldrich), and the plated cells were incubated at dark for 4 h, 37 ºC, 5% CO₂. 100 μL 10% SDS (# BP166-500, Fisher HealthCare) with 0.01 M HCl was added to each well to stop the reduction reaction and to dissolve the purple precipitates. After incubation of the cells at 37 ºC for overnight, the OD value at 595 nm of the solution was measured in a microplate reader (Beckman Coulter DTX 880 Multimode Detector). Data represent the mean ± standard deviation of three independent experiments.

Confocal Microscopy Imaging

Cells in exponential growth phase were seeded in glass bottomed culture chamber/dish in 1 × 10⁵ cells/well. The cells were allowed for attachment for 12–24 hours at 37 ºC, 5% CO₂. The culture medium was removed, and new culture medium containing pNDP1/2 at certain concentration was added. After incubation for certain time, cells were stained with 0.001 mg/mL Hoechst 33342 for 10 min at 37 ºC in dark. After that, cells were rinsed three times by PBS buffer/live cell imaging solution for 3 times, and kept in the live cell imaging solution (Invitrogen Life Technologies A14291DJ) for imaging.

Correlative Light and Transmission Electron Microscopy (CLEM) Imaging

Aclar discs with a 1.5 mm diameter were punched from Aclar film (EMS, #50426-10; Fort Washington, PA, USA) and marked with a finder-grid pattern to permit tracking specific cells of interest throughout light microscopy and EM sample preparation. Then they were mounted on a Lab-Tek II chambered coverglass (#155379 Nalge Nunc International) using Matrigel (Matrigel Basement Membrane Matrix Growth Factor Reduced: #356230: BD Biosciences, Bedford, MA, USA). HeLa cells were then seeded on the cover glass including the Aclar discs and grown to less than confluent density in MEM supplemented with 10% FBS. Cells treated with pNDP1 (500 μM in complete medium) were incubated for 6 h before confocal microscopy and EM preparation. Cells were imaged by phase contrast and/or fluorescent light microscopy using an inverted Leica SP2 confocal laser scanning microscopy. After confocal microscopy imaging, the Aclar discs with the cells were transferred into half of an aluminum planchette (Wohlwend, Switzerland) covered with a drop of medium containing 150 mM sucrose as cryo-protectant for high-pressure freezing, and the second half of the planchette was added to enclose the cells in a cavity with a 0.1 mm height. The samples were rapidly frozen using a Leica HPM-100 high-pressure freezer (Leica Microsystem, Vienna, Austria). The frozen cells were freeze-substituted at low temperatures over 3 days in a solution containing 1% osmium tetroxide (EMS), 0.5% anhydrous glutaraldehyde (EMS), and 2% water in anhydrous acetone (AC32680-0010 Fisher Scientific) using a Leica AFS-2 device. After the temperature was raised to 4 °C, the cells were infiltrated and embedded in EMbed 812-Resin (EMS). Ultrathin sections (~70 nm) were collected on slot grids covered with Formvar support film and post-stained with uranyl acetate (supersaturated solution) and 0.2% lead citrate, before being inspected using a FEI Morgagni 268 TEM with a 1k CCD camera (GATAN) or a 300 keV Tecnai F30 intermediate voltage TEM (FEI) with a 4k CCD camera (GATAN). For large overviews of the cells at medium magnification we acquired montages of overlapping images in an automated fashion using the microscope control software SerialEM. For CLEM, we grew cells on Aclar discs that were marked with a pattern that allowed tracking of cells of interest throughout the light microscopy and EM sample preparation. After locating cells of interest, e.g., those containing fluorescently labeled hydrogelator, by light microscopy, the cells were rapidly frozen, fixed, and resin-embedded as described above. For TEM analysis, the block was trimmed, guided by the pattern on the Aclar disc, so that only the quadrant containing the cells of interest remained for ultrathin sectioning. After post-staining of the sections, we recorded overview maps of the sections at low magnification in the TEM to localize again the cell(s) of interest, before recording images at higher magnification for the ultrastructural investigation.

Readout Fluorescence on Cells by Plate Reader

1) cell number dependent: After incubating the serial dilutions of HeLa cells (ranging from 2.5 × 10³ to 1 × 10⁴ cells/well (of a 96-well plate)) with pNDP1 (500 μM) in complete cell growth medium for 24 hours, the medium was removed and cells were rinsed with PBS buffer for twice. The fluorescent nanofibrils were dissolved by 100 μL DMSO in each well for a 6-hour incubation, followed by fluorescence detection using a DX880 multimode detector. 2) The plot of time versus the fluorescence emission of pericellular nanofibrils on HeLa cells: HeLa cells were treated with pNDP1 (500 μM) in complete growth medium. The initial cell number is 5000 cells/well.

Membrane Protein Extraction and Western Blot analysis

The extraction of membrane protein was assisted by the Mem-PER™ Plus Membrane Protein Extraction Kit from ThermoFisher Scientific (#89842). Briefly, when reaching confluency in a 10-cm petri dish, cells were washed with PBS buffer for 3 times and resuspended in PBS buffer after being scrapped off from the surface of the dish with a cell scraper (#08100241, Fisher HealthCare). Cell suspension was centrifuged (1,000 × g) and the supernatant was carefully removed and discarded. The cell pellet was then resuspended in 0.75 mL Permeabilization Buffer and vortexed briefly to make a homogeneous cell suspension. After being kept at 4°C with constant mixing for 10 minutes, the cell suspension was centrifuged for 15 minutes at 16,000 × g. The supernatant containing cytosolic proteins was carefully removed and transferred to a new tube, followed by the addition of protease inhibitors (# 78440, ThermoFisher Scientific). The pellet remaining in the old tube was again resuspended in 0.5mL of Solubilization Buffer and incubated at 4°C for 30 minutes with constant mixing. Afterwards, the suspension was centrifuged at 16,000 × g for 15 minutes at 4°C and the supernatant containing solubilized membrane and membrane-associated proteins to a new tube followed by the addition of protease inhibitor. The cytosolic and membrane aliquots were stored at −80°C for future use. Only membrane proteins are used for western blot analysis in this manuscript.

Protein quantification was performed using BSA Coomassie Protein Assay (Thermo Scientific). Equal amounts of protein were run in Mini-Protean®TGX™ gels (Bio-Rad, #456-1093). Western blotting was performed according to standard protocol. The following primary antibodies were used for western blot analyses: Anti-ALPP (Abcam, ab133602), Anti-ALPL (Abcam, ab108337), anti-β-actin (Cell Signaling, # 8457L), anti-rabbit IgG (cell Signaling, #7074S).

Supplementary Material

Supporting information

Figure S1. a) The synthetic route of pNDP1/NDP1. b) TEM images of the amorphous aggregates in the solution of pNDP1 at the concentration as low as 20 μM. The scale bar is 100 nm.

Figure S2. a) Confocal microscopy images show cellular uptake of pNDP1 and oligomers of NDP1 in HS-5 cells treated by endocytosis inhibitors (30 μM chlorpromazine; 5 μg/ml filipin III; 100 μM ethyl-isopropyl-amiloride (EIPA)). Only the addition of EIPA obviously inhibits the cellular uptake, suggesting the uptake pathway is micropinocytosis. The scale bar is 10 μm. b) 24-hour cell viabilities of different cell lines treated by pNDP1 at the concentrations of 500, 400, 300, 200, 100 μM, determined by MTT cell viability/proliferation assay. 500 μM of pNDP1 is biocompatible in the culture of HeLa, HS-5, A2780, A2780cis, MES-SA, MES-SA/Dx5, MCF-7, PC-3, U-87 MG, Capan-2, A375, SKOV3, PC-12 Adh, and T98G. As for the culture of Saos-2 and 7F2, 200 μM of pNDP1 is nontoxic. c) Fluorescent confocal microscopy images show the fluorescence emission in the HeLa cell culture incubated with pNDP1 at the concentration of 500 μM in culture medium, with and without (−)-tetramisole (40 μM) for 6 hours. The scale bar is 10 μm. d) Western blot analysis shows relative amount of ALPL and ALPP on the membrane of A2780 and A2780cis cells. e) Western blot analysis shows relative amount of ALPL and ALPP on the membrane of MES-SA MES-SA/Dx5 cells. f) Western blot analysis shows relative amount of ALPL and ALPP on the membrane of MCF-7 cells with and without the treatment of prednisolone (0.5 μg/mL). g) Western blot analysis shows relative amount of ALPL and ALPP on the membrane of different cell lines.

Figure S3. a) Molecular structures of pNDP2 and NDP2, and the conversion catalyzed by ALP under physiological condition. b) Transmission electron microscopy (TEM) images of pNDP1, (top) before and (bottom) after the treatment with ALP (2U/mL) at the concentration of 1.0 wt% and pH of 7.4. Inset is the optical images of the solution and hydrogel, respectively. The scale bar is 100 nm. c) Rheological characterization of hydrogel formed by treating the solution of pNDP1 and pNDP2 with ALP (2U/mL), at the concentration of 1.0 wt% and pH of 7.4. (left) The strain dependence of the dynamic storage (G′) and loss storage (G″) is taken at a frequency equal to 6.28 rad/s, and (right) the frequency dependence is taken at a strain equal to 0.78 %. d) Fluorescent confocal microscopy images show the time course of fluorescence emission in the HeLa cell culture incubated with pNDP2 at the concentration of 500 μM in culture medium. The scale bar is 50 μm. e) Time-dependent curves show the dephosphorylation process of pNDP1 (5 mL, pH 7.4, 500 μM) treated by ALP (1 μg), PTP1b (1 μg), or PP1 (1 μg) 37°C in PBS buffer.

Figure S4. a) Western blot of ALPP and ALPL on cell membrane of cancer cells (HeLa and Saos-2), and a normal cell (HS-5). b) Heat map of ALPL, ALPP and ALPI of BioGPS cell line gene expression profiles. Extracted from database Harmonizome. c) The treatment of ALPL/TNAP inhibitor DQB (2 μM) reduces the fluorescence on HeLa cell surface (3 hour incubation). Scale bar = 20 μM.

Bigger Picture.

Reversible phosphorylation/dephosphorylation in the extracellular domains, as an essential cellular process, is an emerging field for translational medicine. For example, recent advance in cancer research reveals that many types of cancer cells overexpress certain phosphatases. However, current phosphatase probes are unable to serve as activity probes for imaging tumor specific phosphatases on cancer cells for diagnosis and monitoring of cancer. This work proves the feasibility of using a small molecule—D-tetrapeptide derivatives—for EISA as the selective molecular probes for imaging the activity of certain phosphatases overexpressed by cancer cells. The biostable D-peptidic probes for imaging the activities of tumor specific phosphatases provide a new class of functional imaging agents for biologists and medical scientists to validate the functions and inhibitors of cancer specific phosphatases at cellular level. The work contributes a molecular tool for precision medicine by measuring, probing, or targeting of phosphatases on and in live cells.