1 Introduction

Alkaline phosphatase (ALP) is a glucoproteinase that has the specific catalysis for the dephosphorylation process of lots of biological substances, including small molecules, nucleic acids, and proteins. Previous researches have confirmed that the abnormality of ALP activity or concentration can lead to a great variety of diseases, such as liver dysfunction [1], bone injury [2], and prostate cancer [3]. In clinical practice, it has become an important biomarker for the diagnosis of these diseases. Thus, it is of great significance to develop effective and simple methods for ALP monitoring [4,5,6,7,8].

Up to now, several approaches, including colorimetric [8,9,10,11,12,13,14,15,16,17], fluorometric [6, 7, 18,19,20], electrochemical [21, 22], and Raman [23, 24] means, have been exploited for the analysis of ALP based on its ability of removing the phosphate group from substrates to gain signal readouts via various principles. Among these approaches, colorimetric assays have drawn considerable interest because of their merits of simplicity, easy readout, and low cost [8, 25]. At present, the hydrolysis of colorless p-nitrophenyl phosphate (p-NPP) to yellow p-nitrophenolate (p-NP) is universally used to monitor ALP in clinical practice [26]. Although this type of assay is simple, the poor selectivity significantly limits its practicality. More seriously, the p-NPP substrate is so sensitive to light and heat that it can be hydrolyzed without the participation of ALP, which will produce false-positive results [27]. Therefore, it still needs further effort to explore novel mechanisms and approaches with good practicability and reliability for the determination of ALP.

Since Yan’s group found that magnetic Fe3O4 particles exhibited intrinsic peroxidase-like activity in 2007 [28], nanozymes, defined as nanoscale materials with enzyme-mimicking properties, have attracted increasing attention in the academic community [29,30,31,32,33,34]. With the superiorities of large-scale production, low cost, good robustness, and easy-to-adjust performance compared with natural enzymes, they have been developed for use in catalysis [35, 36], biomedicine [37,38,39], and sensing [40,41,42,43,44,45,46,47,48,49]. For the detection of ALP, several nanozymes or enzyme mimics have found promising applications on the basis of various principles [4, 8, 12, 17, 50,51,52]. For example, Wang’s group used ALP to hydrolyze the o-phosphonoxyphenol substrate to catechol that could stimulate the inert TiO2 nanoparticles to a photoresponsive nanozyme, and with this strategy ultrasensitive monitoring of ALP was gained [12]. Liu et al. developed a fluorescent and colorimetric dual-mode ALP assay based on the destruction of oxidase-like CoOOH nanoflakes by ascorbic acid generated from an ALP-catalyzed dephosphorylation process [50]. Very recently, we reported a sensitive and selective strategy based on the phosphate-quenched oxidase-mimicking activity of Ce(IV) ions for the colorimetric detection of ALP [8]. These successful cases suggest that nanozymes with adjustable activities have great potential in the ALP monitoring application.

Herein, we propose a new colorimetric assay based on the pyrophosphate (PPi)-mediated oxidase-mimicking activity switching of nanosized MnFe2O4 for ALP analysis. As illustrated in Scheme 1, nanosized MnFe2O4 was synthesized and used as an oxidase-like nanozyme with high activity to catalyze the oxidation of colorless 3,3′,5,5′-tetramethylbenzidine (TMB) to blue TMBox, providing a remarkable color reaction rapidly; PPi was found to inhibit the nanozyme activity through the strong interaction of PPi and the Fe(III) species in MnFe2O4, and such that the TMB color reaction was suppressed seriously; when ALP existed, it hydrolyzed the PPi substrate to phosphate (Pi) that showed no notable influence on the MnFe2O4 activity, and as a result the nanozyme-catalyzed TMB color reaction could be observed again. With this principle, sensitive and selective analysis of ALP was obtained. Reliable monitoring of the target in clinical samples was also verified by our developed assay.

Scheme 1
scheme 1

Illustration of the colorimetric detection of ALP based on the PPi-mediated oxidase-like activity switching of nanosized MnFe2O4

Full size image

2 Experimental Section

2.1 Chemicals and Reagents

FeCl3·6H2O, MnSO4·H2O, NH3·H2O, Na4P2O7·10H2O, and Na3PO4 were obtained from Sinopharm Chemical Regent Co., Ltd. TMB was provided by Shanghai Macklin Biochemical Co., Ltd. ALP (30 U mg−1, from Escherichia coli), cysteine, glycine, alanine, glutathione, bovine serum albumin, glucose, dopamine, cholesterol, and cholesterol oxidase (ChOx, 10 U mg−1) were purchased from Shanghai Aladdin Reagent Co., Ltd. Glucose oxidase (GOx, 100 U mg−1, from Aspergillus niger) was obtained from Sigma-Aldrich. Urate oxidase (UOx, 10 U mg−1, from Arthrobacter protophormiae) was provided by Beijing Bailingwei Technology Co., Ltd. All the other chemicals and reagents were of analytical grade and used directly. Deionized water was employed during the whole study.

2.2 Synthesis and Characterization of MnFe2O4

The MnFe2O4 nanozyme was prepared via a simple co-precipitation method similar to our previous report [53]. In detail, 1 mM FeCl3 and 0.5 mM MnSO4 were first dissolved in 30 mL deionized water; then the mixture was bubbled with pure N2 for 30 min to expel the dissolved O2; afterwards, 20 mL NH3‧H2O was rapidly added to the mixture with a continuously vigorous stir; after reaction for 2 h, the formed solids were rinsed with adequate deionized water until the leachate pH is nearly neutral, and the products were collected via centrifugation and dried for use.

The synthesized MnFe2O4 was characterized by X-ray diffraction (XRD) measurements on a 6100 diffractometer (Shimadzu). High-resolution transmission electron microscopy (HRTEM) images were obtained using a JEM-2100 microscope (JEOL). X-ray photoelectron spectroscopy (XPS) measurements were done with an ESCALAB-MKII spectrometer (Thermo-Fisher Scientific Co., Ltd.).

2.3 Evaluation of the Oxidase-Like Activity of Nanosized MnFe2O4

To evaluate the oxidase-like activity of the synthesized MnFe2O4, TMB was employed as a chromogenic substrate. 1 mg mL−1 MnFe2O4 stock solution was prepared with deionized water, and 5 mM TMB stock solution was prepared with ethanol. In detail, 50 μL MnFe2O4 solution was first added to 1.85 mL NaAc-HAc buffer (0.2 M, pH 4.0), and 100 μL TMB solution was further added to the above solution and mixed uniformly for reaction. The UV–Vis spectra of the mixture were monitored by a Cary 8454 spectrometer (Agilent Technologies Co., Ltd.). Steady-state kinetic measurements were carried out by recording the absorbance at 652 nm within 1 min at a 5 s interval. The apparent kinetic parameters were determined according to the Michaelis–Menten equation υ = Vmax[S]/(Km + [S]), where υ represents the reaction velocity, Vmax is the maximum velocity, [S] represents the concentration of TMB, and Km is the Michaelis–Menten constant.

2.4 Detection of ALP Based on the PPi-Mediated Oxidase-Like Activity Switching of Nanosized MnFe2O4

To investigate the influences of PPi and Pi on the oxidase-mimicking activity of the synthesized MnFe2O4, 50 μL MnFe2O4 solution and 50 μL PPi or Pi solutions (10 mM) were first added to 1.8 mL NaAc-HAc buffer (0.2 M, pH 4.0) and mixed uniformly, and then 100 μL TMB solution was added to the above solution for reaction. The mixture was monitored by the UV–Vis spectrometer.

For the colorimetric detection of ALP on the basis of the PPi-mediated oxidase-like activity switching of nanosized MnFe2O4, 50 μL ALP solution (Tris–HCl buffer, 0.1 M, pH 8.5) with a certain activity and 50 μL PPi solution (10 mM) were first incubated at 37 °C for 40 min, and then 1.8 mL NaAc-HAc buffer (0.2 M, pH 4.0) was added to the mixture. After that, 100 μL TMB solution was added to the above solution for further reaction for 10 min. The mixture was also monitored by the UV–Vis spectrometer.

For the detection of ALP in real samples, human serum samples obtained from the Affiliated Hospital of Jiangsu University were measured using our assay. In detail, 50 μL PPi solution (10 mM) and 50 μL serum sample were first incubated in a tube at 37 °C for 40 min, and then 1.8 mL NaAc-HAc buffer (0.2 M, pH 4.0) was added to the tube. After that, 100 μL TMB solution was added to the above tube. After reaction for 10 min, the mixture was detected by UV–Vis.

3 Results and Discussion

3.1 Characterization of the Synthesized MnFe2O4

The synthesized material was characterized by several means. Figure 1a presents the XRD pattern of the collected products. Three ambiguous peaks attributed to the (311), (400), and (440) planes of spinel MnFe2O4 (JCPDS No. 10-0319) are observed. These indefinable diffraction signals reveal that the MnFe2O4 material prepared by the co-precipitation method has a poor crystallinity, which is in accordance with the previous study [53]. The HRTEM image (Fig. 1b) does not provide long-range ordered crystal structures. Instead, only some short-range lattice fringes are observed in the HRTEM image. This result also supports the poor crystallinity of the synthesized MnFe2O4. As demonstrated by the full XPS (Fig. 1c), both Fe and Mn are present in the obtained material. In the Mn 2p XPS (Fig. 1d), the four fitting peaks centered at 654.7, 653.5, 643.8, and 642.3 eV are ascribed to Mn4+ 2p1/2, Mn2+ 2p1/2, Mn4+ 2p3/2, and Mn2+ 2p3/2, respectively. This result suggests the co-existence of Mn2+ and Mn4+ in the synthesized MnFe2O4. The Mn4+ species is supposed to originate from the redox reaction between the Mn2+ and Fe3+ precursors during synthesis [54].

Fig. 1
figure 1

a Shows the XRD pattern of the synthesized MnFe2O4. b Displays the HRTEM image of nanosized MnFe2O4. c Shows the full XPS of the synthesized material. d Exhibits the Mn 2p XPS of MnFe2O4

Full size image

3.2 Oxidase-Like Activity of the Synthesized MnFe2O4

Previous studies have revealed that Mn-based materials exhibits oxidase-mimicking activity to trigger some chromogenic reactions [55, 56]. Since our synthesized MnFe2O4 has the Mn4+ and Mn2+ species, as demonstrated by Fig. 1d, it is also expected to provide oxidase-like catalytic activity. To check the guess, we used the synthesized MnFe2O4 to catalyze the oxidation of TMB, a chromogenic substrate that is commonly used to evaluate the enzyme-like characteristics of nanozymes. As compared in Fig. 2a, the MnFe2O4 solution with a concentration of 0.025 mg mL−1 has no obvious color background, and the TMB substrate also exhibits no UV–Vis signal. When the synthesized MnFe2O4 and TMB are incubated together, a color reaction occurs fast, providing a remarkable absorption peak at around 652 nm. The reaction should be attributed to the oxidation of colorless TMB to its blue product TMBox. This result verifies that the synthesized MnFe2O4 has oxidase-mimicking activity.

Fig. 2
figure 2

a Compares the UV–Vis spectra of different systems, and the inset shows the corresponding photograph (a─MnFe2O4, b─TMB, c─MnFe2O4 + TMB). b Presents the photograph of the MnFe2O4 + TMB system occurring in air- or N2-saturated solutions. c Shows the steady-state kinetic plots of the MnFe2O4 nanozyme toward TMB

Full size image

To confirm that the synthesized MnFe2O4 indeed plays an oxidase-mimicking role to trigger the TMB color reaction but not acts as an oxidant, the MnFe2O4 + TMB system occurring in air- or N2-saturated solutions was compared [57]. As depicted in Fig. 2b, the colorless TMB substrate can be catalytically oxidized to blue TMBox fast in the air-saturated buffer, while in the N2-saturated buffer the TMB color reaction is seriously inhibited. This contrast verifies that the obtained MnFe2O4 indeed plays an oxidase-like nanozyme role rather than simply an oxidant [57]. The oxidase-mimicking activity of MnFe2O4 is supposed to originate from the Mn species, because it has been reported to have oxidase-like activity [55, 56] but not the Fe species. To verify this conclusion, we synthesized Fe3O4 nanoparticles using a similar method and tested their enzyme-like characteristics. As demonstrated in Figure S1 (Supplementary Information), Fe3O4 nanoparticles exhibit no ability to trigger the TMB color reaction with the absence of H2O2. The result clarifies that the oxidase-mimicking activity of MnFe2O4 is indeed attributed to the Mn species. In detail, the Mn species exhibits both the Mn4+ and Mn2+ oxidation states (Fig. 1d). It is supposed that an initial oxidation of Mn2+ to Mn4+ first occurs with the reduction of O2 to H2O, and then the oxidation of TMB by Mn4+ leads to the regeneration of Mn2+ [54], thus giving the oxidase-like ability of the MnFe2O4 nanomaterial.

Similar to other nanozymes, the oxidase-mimicking activity of the proposed MnFe2O4 highly depends on the pH value of the buffer used. As presented in Figure S2 (Supplementary Information), with the increasing pH of the buffer in the range of 3–9, the absorbance of the MnFe2O4 + TMB system first increases and then decreases, providing the maximal absorbance at pH 4. The TMB color reaction catalyzed by the MnFe2O4 nanozyme also changes along with the reaction temperature. A volcano-type trend between the oxidase-like activity and the reaction temperature is observed in Figure S3 (Supplementary Information). For the convenience of measurements, the following experiments were carried out at room temperature.

To better assess the oxidase-mimicking activity of the proposed MnFe2O4, steady-state kinetic measurements were carried out under optimized conditions. As presented in Fig. 2c, typical Michaelis–Menten plots are found for the TMB substrate, providing a Km value of 0.23 mM and a Vmax value of 4.17 × 10−7 M s−1 when the concentration of MnFe2O4 utilized is 0.025 mg mL−1. According to the literature [58], the catalytic constant kcat and efficiency kcat/Km are further determined to be 3.86 × 10−3 s−1 and 16.78 M−1 s−1, respectively. Although the catalytic efficiency of the MnFe2O4 nanozyme is still inferior in comparison with natural oxidase, it shows advantages of large-scale preparation, low cost, and good robustness. In addition, the MnFe2O4 nanozyme, similar to other nanozymes, shows excellent stability against harsh conditions. When it is first incubated at different temperatures (Figure S4, Supporting Information) or in buffers with various pH values (Figure S5, Supporting Information) for a certain time and then tested, its enzymatic activity has no remarkable changes, revealing the favorable robustness of the MnFe2O4 nanozyme.

3.3 The Oxidase-Like Activity of MnFe2O4 can be Inhibited by PPi but not Pi

It is interestingly found that the oxidase-like activity of the synthesized MnFe2O4 nanozyme can be inhibited by PPi. As depicted in Fig. 3a, the MnFe2O4 + TMB system triggers a remarkable color reaction fast. When 0.25 mM PPi is added, the color reaction is seriously restrained. Given that the PPi species cannot directly affect the TMB and TMBox substances, the suppression should only be ascribed to the inhibition of the MnFe2O4 nanozyme activity. Previous studies have confirmed that PPi is able to chelate certain metal ions including Fe3+ and Cu2+ [36, 59]. Therefore, it is supposed that the activity inhibition of the MnFe2O4 nanozyme originates from the strong coordination interaction between PPi and the Fe(III) species in MnFe2O4. To back the hypothesis, the MnFe2O4 nanozyme before and after interacting with PPi was characterized by XPS. As shown in Fig. 3b, in the as-synthesized MnFe2O4 no P 2p signal is detected. After incubating with PPi, the collected material offers a weak but recognizable P 2p signal. This result demonstrates that the PPi species indeed combines with the MnFe2O4 nanozyme.

Fig. 3
figure 3

a Compares the UV–Vis spectra of the MnFe2O4 + TMB system with/without the addition of PPi or Pi (TMB: 0.25 mM; MnFe2O4: 25 μg mL−1), and the inset shows the corresponding photograph (a─MnFe2O4 + TMB, b─MnFe2O4 + PPi + TMB, c─MnFe2O4 + Pi + TMB). b Compares the P 2p XPS of the synthesized MnFe2O4 before and after interacting with PPi. c Indicates that the suppression degree of the MnFe2O4 + TMB system by PPi highly depends on the concentration of PPi added

Full size image

However, when Pi, instead of PPi, is added, the TMB color reaction can be observed as that with no addition of Pi (Fig. 3a). The result reveals that the Pi species barely affects the nanozyme activity, different from the previous study reporting that Pi could suppress the peroxidase-like activity of magnetite nanoparticles (MNPs) [60]. The difference may be related to the two nanozymes used. In the previous study [60], Pi could weakly coordinate with Fe3+ on the surface of MNPs, and the Fe3+ species played the main peroxidase-mimicking role. Therefore, Pi with a high concentration might be able to inhibit the catalytic activity of MNPs. In our work, Pi may also weakly coordinate with Fe3+ in MnFe2O4, but the species with a smaller molecular structure compared with PPi cannot affect the nearby Mn2+/Mn4+ species (which play the main oxidase-mimicking role) as PPi does. Besides, in comparison with the coordination of Fe3+ and PPi, the interaction between Pi and Fe3+ is much weaker. Therefore, no obvious influence of Pi (with a concentration of 0.25 mM) on the catalytic activity of the MnFe2O4 nanozyme is found, while PPi with the same concentration can significantly inhibit the nanozyme-catalyzed TMB color reaction. It is further found that the suppression degree of the MnFe2O4 + TMB system by PPi highly depends on the concentration of the latter. With the increase of the PPi concentration, the MnFe2O4 + TMB system is inhibited more significantly, resulting in the decreasing absorbance at 652 nm, as observed in Fig. 3c.

3.4 Colorimetric Detection of ALP Based on the PPi-Mediated Oxidase-Like Activity Switching of MnFe2O4

With the above finding, a colorimetric assay based on the PPi-mediated oxidase-like activity switching of MnFe2O4 was explored for the monitoring of ALP. As illustrated in Scheme 1, PPi can inhibit the TMB color reaction catalyzed by the MnFe2O4 nanozyme; it is expected that ALP will hydrolyze the PPi substrate to Pi that cannot affect the nanozyme, and such that the TMB color reaction can be observed again. To verify the assay, the UV–Vis spectra of the PPi + MnFe2O4 + TMB and ALP + PPi + MnFe2O4 + TMB systems were compared. As shown in Fig. 4, it is obvious that the presence of ALP can restore the nanozyme-catalyzed TMB color reaction. The phenomenon can be utilized for the colorimetric determination of ALP.

Fig. 4
figure 4

UV–Vis spectra of the PPi + MnFe2O4 + TMB system with the presence of ALP or not (PPi: 5 mM; TMB: 0.25 mM; MnFe2O4: 25 μg mL−1). The inset shows the corresponding photograph (a─PPi + MnFe2O4 + TMB, b─ALP + PPi + MnFe2O4 + TMB)

Full size image

To detect ALP better, the reaction time of ALP and PPi was optimized. Figure S6 (Supplementary Information) depicts the effect of the reaction time between ALP and PPi on the signal readout. When the reaction time extends, the absorbance at 652 nm also increases until a saturation phenomenon appears when the reaction time extends to 40 min. Finally, we selected 40 min as the optimal reaction time of ALP and PPi.

Under optimized conditions, we utilized the developed assay to detect the ALP target with different activities. As expected, the UV–Vis responses of the ALP + PPi + MnFe2O4 + TMB system increase along with the increasing activities of ALP (Fig. 5a). It is further found that the absorbance at 652 nm is linearly increased with the ALP activity ranging from 0.6 to 55 U L−1. Based on the signal-to-noise of three (S/N = 3) rule, the limit of detection (LOD) for ALP determination is determined to be as low as 0.27 U L−1. The detection range and LOD can fully meet the requirements for ALP monitoring in clinical practice. In comparison with previously reported assays for ALP sensing (Table S1, Supplementary Information), our assay is also comparable or even better in terms of linear range and LOD. More importantly, our assay exhibits the following merits over other methods: (1) compared with fluorescence measurements that usually need some labels [5, 18, 61], our colorimetric assay can offer comparable analytical performance with simpler sensing elements, and smartphone or even naked eyes can be utilized for signal readout [48]; (2) different from a majority of previously reported colorimetric approaches that depend on the optical changes of elements themselves [25, 62, 63], our assay is on the basis of the PPi-mediated activity switching of an oxidase-like nanozyme to catalytically amplify the colorimetric signal, and thus excellent sensitivity and low LOD are obtained. These merits will endow it with great promise in practical applications.

Fig. 5
figure 5

a Shows the UV–Vis spectra of the ALP + PPi + MnFe2O4 + TMB system with various activities of ALP (PPi: 5 mM; TMB: 0.25 mM; MnFe2O4: 25 μg mL−1). b Depicts the linear relationship between the absorbance at 652 nm and the ALP activity. c Compares the UV–Vis responses of the developed assay toward different biological species (1─50 U L−1 ALP, 2─10 mM KCl, 3─10 mM Na3PO4, 4─1 mM cysteine, 5─1 mM glycine, 6─1 mM alanine, 7─1 mM glutathione, 8─5 mM bovine serum albumin, 9─5 mM glucose, 10─0.1 mM dopamine, 11─0.1 mM cholesterol, 12─50 U L−1 GOx, 13─10 U L−1 UOx, 14─10 U L−1 ChOx)

Full size image

In addition to high sensitivity, the developed assay is also specific for ALP determination. We assessed the possible effects of common substances co-existing in human blood on the selective detection of ALP. As demonstrated in Fig. 5c, only ALP provides a remarkable absorbance at 652 nm, while other species, including inorganic salts (KCl and Na3PO4), amino acids (cysteine, glycine, and alanine), polypeptides (GSH), proteins (bovine serum albumin), small biological molecules (glucose, dopamine, and cholesterol), and natural enzymes (GOx, UOx, and ChOx), cannot trigger the restoration of the nanozyme-catalyzed TMB color reaction inhibited by PPi. The result confirms that our assay has good selectivity for the sensing of ALP. The excellent selectivity should be attributed to not only the exclusive inhibition of the MnFe2O4-catalyzed TMB color reaction by PPi but also the specific hydrolysis of PPi by ALP.

To evaluate the potential of our assay in practical applications, it was utilized to monitor the ALP levels in clinical samples. Four human serum samples available from the Affiliated Hospital of Jiangsu University were tested by our assay and compared with the clinical data. Since the normal level of ALP in adult serum ranges from 46 to 190 U L−1, our assay is sensitive enough to detect the ALP levels in these samples directly. The detailed processes for real sample measurements have been described in the experimental section. The detection results are listed in Table 1. These results are found to be in agreement with the clinical data, providing the relative errors ranging from − 4.6% to + 5.9%. This indicates that our colorimetric assay can be used for the reliable determination of ALP in clinical practice.

Table 1 Results of our assay for ALP determination in human serum samples
Full size table

4 Conclusions

In summary, we have developed a facile but effective colorimetric assay based on the PPi-mediated oxidase-mimicking activity switching of nanoscale MnFe2O4 for the determination of ALP. Our assay can provide enough sensitivity and specificity to sense ALP, and excellent reliability and practicability have also been verified by using it to monitor the target in practical samples. Taken together, our ALP assay shows great potential as a convenient, low-cost, and efficient tool in clinical practice, and the proposed strategy can also inspire the development of new ALP-linked immunosorbent assays.