Introduction

Noble metal platinum (Pt) has been recognized as the most efficient biocatalysts to mimic natural enzymes in many fields1,2,3,4,5. The pursuit of biocatalyst development is the desired catalytic activity, high substrate selectivity, and durable stability, which depends sensitively on the rational design of both geometric and electronic structure due to the prominent quantum effect6,7,8,9. A powerfully effective approach to manipulating the electronic structure of Pt-based catalysts, especially shifting the d-band center, is lattice strain engineering10,11,12,13,14,15,16,17. Generally, alloying Pt with transition metals, M (M = Pd, Fe, Ru and so on)18,19,20,21,22, could provide a different dimension to modulate the geometric and electronic properties via the characteristic strain effect and/or ligand effect for high-performance catalysis. Moreover, the use of typical core-shell structures and elastic substrates has also been proposed to master lattice strain property of catalysts due to the lattice mismatch7,10,23,24. Besides, strain effect can also be generated through the vacancy defect25,26,27,28, stacking faults29,30, and in-plane amorphous-crystalline phase boundaries of catalysts with drastically improved catalytic performance31.

However, lattice strain introduced by lattice mismatch or amorphous-crystalline phase boundaries extensively occurs at the interface of two metals, metal-substrate or in-plane phase boundary, and gradually weakens from the interface toward the outmost surface, which severely limits the efficient development of high-performing Pt-based biocatalysts31,32,33. Therefore, the catalytic performance shows strongly dependence on the thickness of shell in core-shell structure or the size of metal in metal-substrate structure. For example, ultrathin Pt shells (~2 nm) deposited on palladium-based nanocubes with compressive-strain or tensile-strain possessed efficient electrocatalytic activity via phosphorization and dephosphorization at high temperature10. Ultrasmall Pt clusters with several atomic layers may be an alternative method to effectively utilize strain effect from the lattice mismatch of interface to boost biocatalytic activity34. Meanwhile, single Pt atoms on noble metal substrate have been proposed to tune the electronic structure and achieve efficient catalytic performance35,36,37,38. To investigate the strain-activity correlations of single Pt atoms or ultrasmall Pt clusters, we should perform strain-distribution analyses at the atomic scale, which is completely distinct from the Geometric phase analysis adopted by conventional lattice strain evolution of nanoparticles (10–100 nm). At the atomic scale, strain values can be accurately calculated by the precise position of the atomic columns39,40. However, it is still challenging to modify the atomic-scale strain for specific enzyme-like reactions.

In this work, we design Pt single atom (Pt1), several Pt atoms (Ptn) and atomically-resolved Pt cluster (Ptc) with polyvinyl pyrrolidone (PVP)-antiaggregating ligand laminated on PdAu nanocrystals (denoted as Ptc-PA) by a simple and mild method at room temperature. Strain analyses at the atomic scale manifest that Ptc possesses reasonable tensile strain along [110] direction (Supplementary Fig. 1). Density Functional Theory (DFT) calculations reveal that tensile strain upshifts the d-band center and exposes high potential surface, signifying more catalytic active sites and stronger adsorption energy to contribute the biocatalytic performances (Fig. 1). Besides, Au core and Pd shell in Ptc-PA would serve as electron donor and electron storage pool, respectively, which further improves the enzyme-like catalytic properties. In particular, Ptc-PA exhibits an excellent peroxidase (POD)-like activity with the catalytic efficiency (Kcat/Km) of 1.50 × 109 mM−1 min−1, about four orders of magnitude higher than natural horseradish peroxidase (HRP), as well as FeN3P SAzyme41, FeN5 SA/CNF42, RhN4 SAzyme43, and conventional nanozymes44,45,46,47. In addition, Ptc-PA possesses 61.5 times higher catalase (CAT)-like capacity than that of natural CAT and comparable superoxide dismutase (SOD)-like activity to that of natural SOD. Owing to remarkable POD-like activity, Ptc-PA enables the detection limit down to 0.7525 pg mL−1 for clinical diagnosis of cancer based on catalyzing enhancement, representing a 132-fold improvement over visual detection methods. Ptc-PA can inhibit acute inflammation of multi-organ damage and mitigate oxidative stress disorders, which results in a significantly improvement of irradiated mice survival rate from 20% to 80%.

Fig. 1: Schematic of the atomic strain engineering to enhance Ptc-PA enzyme-like activities.
figure 1

Firstly, the evolution process from Pt single atom (Pt1), several Pt atoms (Ptn) to Pt cluster (Ptc) on the PdAu is accompanied by a reduction in tensile strain. Strain can prompt the formation of Pt vacancies, and further modulate the electronic structure, especially the upshift of d-band center. Meanwhile, Pd and Au substrates serve as storage layer and generation source of energetic charge carriers, respectively, which contributes to the formation of high potential surface. Thus, Ptc-PA exhibits ~104 times and 61.5 times higher POD-like and CAT-like capacities than natural HRP and CAT, respectively, and comparable SOD-like activity with natural SOD.

Full size image

Results

Structural properties

Surface strain can regulate the electrical properties of Pt catalysts, typically by shifting the d-band center, thereby altering the catalytic activity of metal nanomaterials9,10,31. Generally, surface strain results from lattice mismatch7,13. Since strain is spatially dependent, exposing active strain surfaces is challenging and requires the regulation of geometric structures with surface strain at atomic scale or a few of atomic-layers level. Herein, we synthesized Pt single atom (Pt1), several Pt atoms (Ptn) and atomic resolution Pt clusters (Ptc) laminated on Au nanocrystals with thickness of ~2.5 nm Pd transition layer (labeled as Pt1-PA, Ptn-PA and Ptc-PA, respectively) to investigate the strain-activity correlations (Fig. 2a). In the absence of the nano-transition layer of Pd or AuPd core, Pt clusters would tend to accumulate and stack rather than existing at the several atomic-layers level (Supplementary Figs. 2 and 3).

Fig. 2: Geometric and strain characteristics of the Pt1-PA, Ptn-PA and Ptc-PA.
figure 2

a Schematic evolution of structure and strain at the atomic level. b–d HAADF-STEM images of Pt1-PA (the yellow circle represents isolated Pt atom), Ptn-PA (the yellow circle represents several Pt atoms), and Ptc-PA (the yellow circle represents Pt clusters), respectively. The experiments were repeated independently three times with similar results. e–g HAADF-STEM-EDS mapping images of Pt1-PA, Ptn-PA and Ptc-PA, respectively. h–j High-magnification HAADF-STEM images of Pt1-PA, Ptn-PA and Ptc-PA, respectively. The experiments were repeated independently three times with similar results. k HAADF-STEM image of Pt cluster from Ptc-PA, showing the emergence of Pt atom and vacancies. l Corresponding intensity profile along the green lines from (k). ‘arb. units’ represents arbitrary units. m HAADF-STEM image of Pt cluster from Ptc-PA, showing screw dislocations. Atomic strain mappings (n–p) and the corresponding box-charts of the measured strain values (q–s) along [110] direction for Pt1-PA, Ptn-PA and Ptc-PA, respectively (n = 67, 40 and 51 for (q), (r) and (s), respectively. Boxes represent the median and IQR and the upper and lower whiskers extending to the values that are within 1.5 × IQR). t Intensity profiles of lattice space from Pt cluster in Ptc-PA and pure Pt. ‘arb. units’ represents arbitrary units.

Full size image

To demonstrate geometric and strain characteristics of Pt1-PA, Ptn-PA and Ptc-PA, transmission electron microscope (TEM) and high-angle angular dark-field scanning transmission electron microscopy (HAADF-STEM) analyses were performed. As shown in Fig. 2b and Supplementary Fig. 4a, single Pt atom is dispersed on the surface of Pd shell or doping in the lattice gap at the molar ratio of reactants with 0.1:1:1. As increasing the proportion of Pt precursors, multiple Pt atoms get gathering on the surface with the molar ratio of 0.2:1:1 (Fig. 2c and Supplementary Fig. 4b), and become clustered with the molar ratio of 0.4:1:1, 0.5:1:1 and 1:1:1 (Fig. 2d and Supplementary Fig. 4c–e). The full morphology of Ptc-PA (Pt:Pd:Au=0.4:1:1, 0.5:1:1 and 1:1:1) from TEM image (Supplementary Figs. 4 and 5) displays the dimensional uniformity and Pt clusters with an average diameter of ~2.45 nm, HAADF-STEM image of Pt clusters also reveals the thickness of several atomic-layers (Fig. 2d), and the calculated atom number of Pt clusters is 509 ± 2 according to the formula reported in the literature48. Furthermore, scanning transmission electron microscopy-energy dispersive spectrometer (STEM-EDS) was conducted to analyze the element distribution, visually showing that Pt elements are uniformly dispersed on the material surface in the form of single atom, polyatoms and up to clusters (Fig. 2e–g and Supplementary Fig. 6a, b). Representative surface structures of Pt1-PA and Ptn-PA were presented at the atomic scale by high-magnification HAADF-STEM images (Fig. 2h, i), revealing the crystal spacing of 0.152 nm and 0.1437 nm, respectively. The corresponding fast Fourier transform (FFT) pattern can be seen in Supplementary Fig. 7a,b to represent the Pd face-centered cubic (fcc) structure. Thus, the aforementioned crystal spacing is corresponding to Pd (220) facet, which is larger than pure Pd probably thanks to Pt atom doping. Similarly, atomic-resolution HAADF-STEM image and corresponding FFT pattern of Pt cluster for Ptc-PA with the molar ratio of 0.4:1:1, 0.5:1:1 and 1:1:1 demonstrate Pt fcc structure and the Pt (220) facet of 0.1436 nm, 0.144 nm, 0.138 nm, respectively (Fig. 2j, Supplementary Figs. 6c,d, 7c–e). X-ray diffraction (XRD) pattern of Ptc-PA with the molar ratio of 0.4:1:1 also reveals randomly oriented fcc crystals in which four peaks are assigned to (111), (200), (220) and (311) facets (Supplementary Fig. 8), while the amplifying peaks are still difficult to demonstrate the lattice spacing change due to the approaching peak of Pt and Pd. In addition, it is found from intensity profile of atomic HAADF-STEM image (Fig. 2k–m) that Pt clusters in Ptc-PA contain a few of Pt vacancies, isolated Pt atoms, screw dislocations, as well as a high density of stepped surface atoms.

Importantly, to evaluate the atomic-strain characteristics of representative surface structures in Pt1-PA, Ptn-PA and Ptc-PA, we pinpointed the precise atom-column positions in the atomic-resolution HAADF-STEM images (Supplementary Fig. 9) by means of StatSTEM software49. Note that the StatSTEM software employs a model-based estimation algorithm and takes into account overlap between neighboring columns to quantify the atomic column positions with high precision and accuracy. Inspired by strain analysis methods in the literature40,50, we measured the nearest-neighbor distances in the crystallographic direction [110] and compared them with standard reference values to generate 2D visualized atomic strain maps (Fig. 2n–p and Supplementary Fig. 10a, b). Further, as shown in Fig. 2q–s and Supplementary Fig. 10c, d, the strain values within the representative region are drawn statistically. The statistical results reveal that tensile strain occurs at the surface of Pt1-PA and Ptn-PA and the average strain value is about 10.86%, 5.03% along the [110] direction, respectively. As for Ptc-PA with the molar ratio of 0.4:1:1, 0.5:1:1 and 1:1:1, the average strain value is 3.61%, 2.61% and −1.72% along the [110] direction, respectively. Despite possessing similar structures, the Ptc-PAs with different molar ratios exhibited gradually decreasing strain values with increasing Pt input, which could be ascribed to continuous stress release51. As for Pt1-PA and Ptn-PA, the production of strain effect could be attributed to Pt atom doping, and especially, lattice gap doping in Pt1-PA plays a vital role in tensile strain. As for Ptc-PA, Pt vacancies and isolated Pt atoms could contribute to the opportune strain, which would lead to the increase of atomic arrangement disorder and low-coordination sites and further be beneficial to the improvement of catalytic performance10,52. Furthermore, unlike conventional nanocrystals (10–100 nm), atomically resolved Pt clusters exhibit finite periodic structure and a high density of stepped surface atoms, which could also be responsible for the observed strain. Meanwhile, quantitative analysis of the intensity profile along lattice space (Fig. 2t) shows that there is a larger lattice expansion of Pt cluster in Ptc-PA with the molar ratio of 0.4:1:1 than pure Pt, which is attributed to the tensile surface strain. In addition, Ptc-PA with the molar ratio of 0.4:1:1 exhibits strong stability in PBS and BSA, facilitating the expansion of its application (Supplementary Figs. 11 and 12). Overall, Ptc-PA possesses atomically resolved Pt clusters with the sufficient exposure of active surface strain sites, as well as Pt vacancies.

Since the three types of Ptc-PAs possess similar structures, we focused on the molar ratio of 0.4:1:1. To evaluate the electronic structure of Ptc-PA with the molar ratio of 0.4:1:1, X-ray photoelectron spectroscopy (XPS) and X-ray absorption fine spectrometric (XAFS) measurements were performed. The XPS analyses of Ptc-PA display that the three metals dominantly hold zero-valence states, while Pt 4f is accompanied by a certain amount of Pt (II) (Supplementary Fig. 13), indicating that Pt electronic states are modified by PdAu core or surface strain via electron transfer. From the normalized X-ray absorption near-edge structure (XANES) spectra of the Au-L3 edge (Fig. 3a), the absorption edge in Ptc-PA appears positive shift to higher energy than Au foil, indicating an increase in the average valence state and a decrease in the electron number of Au. It is conjectured that more electrons are transferred from Au to Pd. As shown in the Fourier transformed extended X-ray absorption fine structure (EXAFS) spectra (Fig. 3b), the maximum peak in Ptc-PA occurs at about 2.58 Å exhibits slightly left shift away from the Au foil, revealing that Au-Au bonds are dominant and a small amount of Au-Pd bonds is accompanied. To corroborate this result, the wavelet transform (WT) analysis of Au L3-edge oscillation was carried out (Fig. 3c, d). Except the intensity at around 8.72 Å−1, Ptc-PA shows a secondary intensity at around 5.9 Å−1, which is assigned to the Au-Pd coordination. For the XANES spectra of Pt-L3 edge (Fig. 3e), the absorption edge in Ptc-PA exhibits positive shift compared to that of Pt foil, revealing the electron transfer from Pt to Pd and the average valence state increases. This result is consistent with the XPS analysis of Pt 4f from Supplementary Fig. 13. Meanwhile, the white line intensity of Pt-L3 edge in Ptc-PA significantly increases, suggesting the presence of high Pt-vacancy density, in agreement with the STEM results (Fig. 2k, l). Furthermore, the corresponding Fourier-transformed EXAFS spectra are performed in Fig. 3f, showing that Ptc-PA has a similar peak at ~2.4 Å with Pt foil, which is mainly ascribed to Pt-Pt bond. To further distinguish the local configurations of Ptc-PA, WT maps were provided. As shown in Fig. 3g, h, the maximum intensity at around 9.1 Å−1 in Pt foil is slightly shifted to 9.2 Å−1 in Ptc-PA, which is assigned to the elongated Pt-Pt bond lengths. Ptc-PA also shows a secondary intensity at around 7.8 Å−1, which could be associated with the formation of the Pt-Pd coordination. In addition, curve fitting analysis in k-space of EXAFS was performed to further analyze the coordination configurations of Au and Pt in Ptc-PA (Supplementary Fig. 14). The coordination number of Pt-Pt bond in Ptc-PA is lower than Pt foil (Supplementary Table 1), which could be attributed to the formation of Pt vacancies introduced by lattice strain and also in favor of catalytic performance26,53.

Fig. 3: Synchrotron XAFS measurements of Ptc-PA.
figure 3

a XANES and (b) Fourier-transformed EXAFS spectra of the Au L3-edge in Ptc-PA and Au foil. c, d Wavelet transform spectra obtained at the Au L3-edge in Au foil and Ptc-PA, respectively. e XANES and (f) Fourier-transformed EXAFS spectra of the Pt L3-edge in Ptc-PA and Pt foil. g, h Wavelet transform spectra obtained at the Pt L3-edge in Pt foil and Ptc-PA, respectively. R represents the radial distribution function.

Full size image

Enzyme-like activity

POD-like activity

To investigate the peroxide-like activity of Pt single atom (Pt1), several Pt atoms (Ptn) and atomically-resolved Pt clusters (Ptc) on PdAu biocatalysts, a colorimetric method using 3,3′, 5,5′-tetramethylbenzidine (TMB) in the presence of hydrogen peroxide (H2O2) was performed (Fig. 4a). To explore the effect of different Pt atomic numbers and surface strain on the POD-like activity, Fig. 4b shows that Ptc-PA exhibits the highest value of catalytic efficiency (Kcat/Km) at the molar ratio of reactants with 0.4:1:1, demonstrating that the activity of Ptc-PA increases nonmonotonically with the quantity of Pt atoms and possesses an optimal molar ratio and appropriate strain. In particular, we quantified the POD-like activity of pure Pt NPs and Ptc-PA (Pt:Pd:Au=0.4:1:1) at identical Pt mass concentrations using inductively coupled plasma massspectrometry (ICP-MS), and compared the strain and POD-like activity between Ptc-PA and pure Pt NPs. As shown in Fig. 4c, Ptc-PA (Pt:Pd:Au=0.4:1:1) with 3.61% tensile strain exhibits a POD-like activity ~45.3-fold higher than that of pure Pt NPs without strain, which could be due to atomically resolved Pt clusters in Ptc-PA with the more exposure of active catalytic sites, as well as a larger specific surface area. Above all, these results further emphasize the enormous impact of strain at the atomic level on the POD-like activity of nanomaterials. Furthermore, we performed a systematic comparison of the activities of all species NPs composed of Au, Pd, and Pt elementals at the same mass concentration of the whole sample and synthesized by the same method. It is found that Ptc-PA (Pt:Pd:Au=0.4:1:1) possesses the highest activity, indicating that both Au and Pd elements play a key role in the enhancement of catalytic performance (Supplementary Figs. 15 and 16). Subsequently, the POD-like activity of Ptc-PA (Pt:Pd:Au=0.4:1:1) was explored in detail, showing the concentration-dependence (Fig. 4d and Supplementary Fig. 17). In contrast, there is scarcely any oxidases (OXD)-like activity for Ptc-PA (Pt:Pd:Au=0.4:1:1), indicating the catalytic specificity of POD-like activity (Supplementary Fig. 18). Besides, Ptc-PA (Pt:Pd:Au=0.4:1:1) still maintained ultrahigh POD-like activity in the five-round catalytic reaction (Fig. 4e and Supplementary Fig. 19), suggesting that the enduring catalytic stability.

Fig. 4: POD-like activity of Ptc-PA.
figure 4

a Schematic diagram for POD-like activity of Ptc-PA. b Strain and Kcat/Km of Au@Pd@Pt with different molar ratio. c POD-like activity of Pt and Ptc-PA (Pt:Pd:Au=0.4:1:1). d The absorbance of POD-like activity of Ptc-PA with different concentrations changes with time (n = 3 independent experiments, data are presented as mean ± SD). e Cyclic stability of TMB catalyzed by Ptc-PA in the presence of H2O2. The Michaelis–Menten at various concentrations of (f) H2O2 and (g) TMB for Ptc-PA (Inset: corresponding Lineweaver–Burk plots, n = 3 independent experiments, data are presented as mean ± SD). h Comparison of strain, Pt atoms from Pt cluster and kinetics for Pt1-PA, Ptn-PA, Ptc-PA and HRP. [E] represents particle concentration of Ptc-PA used in the catalytic reaction. Km is the Michaelis–Menten constant. Vmax is the maximal reaction velocity. Kcat is the catalytic rate constant, where Kcat = Vmax/[E]. The value of Kcat/Km represents catalytic efficiency. i Changes of POD-like activity of Ptc-PA and HRP with time (n = 3 independent experiments, data are presented as mean ± SD). j Temperature stability of POD-like activity of Ptc-PA (n = 3 independent experiments, data are presented as mean ± SD). All ‘arb. units’ on the axis represent arbitrary units.

Full size image

To quantitatively evaluate the POD-like activity of Ptc-PA, enzyme kinetic tests were performed on Ptc-PA and natural HRP (Fig. 4f, g and Supplementary Fig. 20). The particle concentration of Ptc-PA was obtained by conversion from the elemental content measured by ICP-MS. The corresponding Michaelis constant (Km) of Ptc-PA (Pt:Pd:Au=0.4:1:1) for the TMB and H2O2 substrate are 0.044 mM and 2.09 mM, respectively. It is noteworthy that the Km of Ptc-PA (Pt:Pd:Au=0.4:1:1) for TMB substrate was lower than that of natural HRP (Fig. 4h), indicating that Ptc-PA exhibits higher affinity for TMB substrates. Accordingly, the catalytic rate constant Kcat (also defined as the turnover number) of Ptc-PA (Pt:Pd:Au=0.4:1:1) for TMB and H2O2 appears 4.05 × 103 times and 326 times higher than that of natural HRP. Meanwhile, the catalytic Kcat/Km of Ptc-PA (Pt:Pd:Au = 0.4:1:1) for TMB is 1.50 × 109 mM−1 min−1, about 104-fold, 1.07 × 104-fold, 8.02 × 106-fold, 7.98 × 105-fold, 3.4-fold, and 100-fold higher than the natural HRP, FeN3P SAzyme54, FeN5 SA/CNF42, RhN4 SAzyme43, Ptc-PA (Pt:Pd:Au=0.5:1:1) and Ptc-PA (Pt:Pd:Au=1:1:1), respectively (Fig. 4h, Supplementary Fig. 21 and Supplementary Table 2). In addition, the catalytic efficiency of Ptc-PA (Pt:Pd:Au=0.4:1:1) also far superseded that of other conventional nanozymes (Supplementary Table 2). The catalytic efficiency of Ptc-PA (Pt:Pd:Au=0.4:1:1) for H2O2 displays almost 2.43 × 103-fold higher than natural HRP. In addition, the natural HRP shows nearly no catalytic activity at 3 days, while Ptc-PA (Pt:Pd:Au=0.4:1:1) maintains a relatively stable catalytic activity for 50 days (Fig. 4i and Supplementary Fig. 22). Performance stability under different storage conditions, especially temperature, is a very important indicator for the practical application of nanomaterials, and thus the POD-like activity of Ptc-PA (Pt:Pd:Au=0.4:1:1) at different temperatures was evaluated. Figure 4j shows that the catalytic performance of Ptc-PA (Pt:Pd:Au=0.4:1:1) could be efficiently sustained at 0 °C–50 °C, which facilitates the broadening of applications of Ptc-PA (Pt:Pd:Au=0.4:1:1). The above results demonstrate that Ptc-PA (Pt:Pd:Au=0.4:1:1) exhibits robust and stable POD-like capability.

To ascertain whether the main factor contributing to the superior POD-like activity of Ptc-PA stems from atomic-level strain on its surface, we compare the strain and POD-like activity of Ptc-PA (Pt:Pd:Au=0.4:1:1), Ptc-PA (Pt:Pd:Au=0.5:1:1) and Ptc-PA (Pt:Pd:Au=1:1:1). Despite only a 20% difference in Pt composition between Ptc-PA (Pt:Pd:Au=0.4:1:1) and Ptc-PA (Pt:Pd:Au=0.5:1:1), the POD-like catalytic efficiency of Ptc-PA (Pt:Pd:Au=0.4:1:1) with 3.61% tensile strain was 3.4-fold higher than Ptc-PA (Pt:Pd:Au=0.5:1:1) with 2.61% tensile strain, suggesting that a proper strain value facilitates catalytic efficiency10. Moreover, the catalytic efficiency is further reduced when the strain became negative (compressive strain) in Ptc-PA (Pt:Pd:Au=1:1:1) even with the increased exposure of Pt atoms. Therefore, the above results demonstrate that the suitable tensile strain effect can significantly enhance the POD-like catalytic efficiency.

CAT-like activity

Subsequent tests have utilized Ptc-PA with a molar ratio of Pt:Pd:Au=0.4:1:1. For simplicity, the molar ratio will not be reiterated in subsequent writing. As shown in Fig. 5a, b, Ptc-PA exhibits favorable CAT-like activity with concentration dependence, which is also visualized by different degrees of bubbling after decomposition of H2O2. To further investigate and quantify the CAT-like activity, we similarly performed enzyme kinetic tests on Ptc-PA and compared it with natural CAT (Fig. 5c and Supplementary Fig. 23). The Km and Kcat of Ptc-PA (4.49 mM and 1.66 × 107 min−1) is significantly better than that of natural CAT (28.8 mM and 1.73 × 106 min−1), suggesting that Ptc-PA has a stronger affinity and higher catalytic rate for H2O2 substrate (Fig. 5d). In particular, the Kcat/Km of Ptc-PA is 3.69 × 106 mM−1min−1, which is 61.5 times higher than that of natural CAT. In combination, these results indicate that Ptc-PA possesses excellent CAT-like activity beyond natural enzyme and is more efficient in catalyzing the formation of non-toxic O2 and H2O from H2O2.

Fig. 5: Other enzyme-mimetic activity of Ptc-PA.
figure 5

a Time-dependent absorbance at 240 nm of H2O2 treated with different concentrations Ptc-PA (n = 3 independent experiments, data are presented as mean ±SD). b Efficiency of H2O2 scavenging by Ptc-PA in 180 s (n = 3 independent experiments, data are presented as mean ± SD). The inset shows the physical decomposition of H2O2. c The Michaelis–Menten and corresponding Lineweaver–Burk plots at various concentrations of H2O2 for Ptc-PA (n = 3 independent experiments, data are presented as mean ±SD). d Comparison of kinetics for Ptc-PA and CAT. e UV–vis spectra of Ptc-PA catalyzing SOD-like reactions. ‘arb. units’ represents arbitrary units. f Inhibitor rate of SOD-like activity for Ptc-PA and natural SOD. g The radar map of enzymatic activities and schematic illustration of Ptc-PA to mimic enzymatic antioxidant defense system.

Full size image

SOD and GPx-like activity

As an important antioxidant enzyme, SOD-like activity of Ptc-PA was assessed through SOD Assay Kit. The variation of the absorbance of the reaction system at 450 nm shows a strong dependence on the concentration of Ptc-PA (Fig. 5e). Ptc-PA exhibits a high inhibition rate of SOD up to nearly 80% at a concentration of 10 μg/mL. A serial comparison of the SOD-like activity of this Ptc-PA with natural SOD was carried out. As shown in Fig. 5f and Supplementary Fig. 24, SOD-like activity of Ptc-PA is nearly equivalent to that of natural SOD. In addition, Ptc-PA is able to maintain high SOD-like activity after almost 50 days (Supplementary Fig. 24d), suggesting the durable stability of SOD-like activity. Besides, it is found that Ptc-PA also exhibits some degree of GPx-like activity (Supplementary Fig. 25). Taken together, the results indicate that Ptc-PA possesses excellent multispecies enzyme-like activities, especially notable POD-like, CAT-like and SOD-like activity (Fig. 5g). Combing with the structure analyses, the superior catalytic activity of Ptc-PA may be due to the fact that ultrathin Pd shell and Au core serve as storage layer and generation source of energetic charge carriers, respectively55,56,57,58, as well as the atomically resolved Pt clusters with surface strain and Pt vacancies59. In addition, XANES also shows that the coordination number of the active center in Ptc-PA is lower than that of pure Pt, which is also favorable to enhance the overall catalytic properties53,60,61.

Scavenging free radicals and electrocatalysis

The scavenging activity of •OH and O2•- free radical was investigated using electron spin resonance (ESR), illustrating that Ptc-PA exhibits favorable scavenging ability of free radicals via the migration of energetic charge carriers (Supplementary Fig. 26). Moreover, the electrocatalytic performances for H2O2 reduction and oxygen reduction reaction (ORR) have also been evaluated by Ptc-PA-modified glassy carbon (GC) electrode. In the presence of H2O2, the reduced current of the Ptc-PA-modified GC electrode is significantly higher than that of the unmodified electrode (Supplementary Fig. 27a). The current density of Ptc-PA reaches −1.4 mA/cm2 at a bias of −0.8 V, which is about 7 times higher than that of a pure GC electrode (Supplementary Fig. 27b), indicating that Ptc-PA exhibits efficient catalytic reduction activity against H2O2. The ORR performance of Ptc-PA-modified GC electrodes was evaluated by measuring the cyclic voltammetric (CV) curve in 0.01 M PBS saturated with O2. Compared to the unmodified GC electrode, the current density of Ptc-PA is about 8 times higher than that of pure GC electrode (Supplementary Fig. 27c, d), suggesting the high catalytic performance for O2. Therefore, Ptc-PA exhibits excellent scavenging activity for oxygen free radicals due to their highly catalytic performance, suggesting that the artificial nanozymes can effectively block the oxygen cascade reaction, regulate the redox imbalance and protect the organism from oxidative damage.

Mechanism of catalytic activity

DFT calculations were carried out to uncover the possible catalytic mechanism and the effect of strain, ultrathin Pd shell and Au core on the excellent catalytic activity of Ptc-PA. According to the analysis of geometrical characteristic of Ptc-PA, the simple atomic structure model has been constructed. As shown in the surficial electron density difference analysis (Fig. 6a), the increased electron density exists in the interface of Au-Pd and Pt-Pd due to the electron transfer among the three metal atoms, which could be attributed to the modulation of electronic structure due to the strain effect. To gain insight into the electron transport among the three metal atoms, we researched the Bader charges of different atoms in Ptc-PA (Supplementary Tables 3 and 4). The Pt and Au possess the negative Bader charges of −0.292 and −0.445, respectively, while Pd exhibits a positive Bader charge of 0.737 (Fig. 6b). It suggests that the electrons transfer from Pt and Au to Pd, and ultrathin Pd shell would serve as electronic storage pool and intermediate transfer station, which is in accord with XANES results. Moreover, compared with compression strain, Pt in Ptc-PA exhibits more electron loss with tensile strain (Supplementary Tables 3 and 4), thereby enhancing its ability to capture foreign electrons. However, the absolute value of Bader charge of Pt atom becomes lower without Au, clearly indicating that the introduction of the Au atomic layer contributes to electron transfer from Pt to Pd. Given the electron transfer effect, Pt0 would be transformed into the form of Pt2⁺, which has been demonstrated by XPS results. Pt2⁺ traps electrons more readily from surface external radicals. Based on it, we propose the capable reaction mechanism for the POD-like and CAT-like processes of Ptc-PA:

$${H}_{2}{O}_{2}+{{Pt}}^{2+}\to {O}_{2}+{Pt}+2{H}^{+}$$
(1)
$$2{H}^{+}+{H}_{2}{O}_{2}+{Pt}\to 2{H}_{2}O+{Pt}^{2+}$$
(2)
Fig. 6: DFT calculation.
figure 6

a The deformation charge density analysis of Ptc-PA model (blue and red represent charge depletion and charge accumulation, respectively. Isosurfaces correspond to 0.002 e/Bohr3). b Bader charge of Ptc-PA model. c TDOS and d-band center of Ptc-PA at different strains. The total ESP for the surface layer of (d) Pt1-PA and (e) Ptc-PA. f Schematic illustrating that electron transfer during different enzyme-like reaction and the d-band center upshift of Ptc-PA at different strains. g Variation of Pt-O bond length and adsorption energy for H2O2 in Pt surface for pure Pt, Ptc-P (Pt atomic layer on the Pd atomic layer without Au atomic layer) and Ptc-PA at different strains.

Full size image

The capable reaction mechanism for the SOD-like process of Ptc-PA is as follows:

$$2{{O}_{2}}^{{{\cdot }}-}+{Pt}^{2+}\to 2{O}_{2}+Pt$$
(3)
$$2{{O}_{2}}^{{{\cdot }}-}+4{H}^{+}+{Pt}\to {2H}_{2}{O}_{2}+{Pt}^{2+}$$
(4)

The d-band center serves as a valuable descriptor for evaluating the adsorption potential of metallic materials towards small molecules62. The upshift of the d-band center results in an elevation of the energy level of the antibonding band, leading to a decrease in electron occupancy within the antibonding orbitals and thereby enhancing bond strength. Consequently, this phenomenon facilitates the adsorption of small molecules by metallic materials. Due to the strain effect, the d-band center of Ptc-PA will shift and in turn affect the overall activity12,63. As shown in Fig. 6c, compared with compressive strain, tensile strain leads to a upshift in the d-band center of Ptc-PA. It could enhance the ability to absorb oxygen-containing intermediates64, and in turn, improve the catalytic activity of Ptc-PA. To further investigate the catalytic enhancement mechanism of Ptc-PA compared to Pt1-PA, we conducted and visualized the total Electrostatic Potential (ESP) based on the DFT simulations. Our focus lies on the surface by plotting the ESP on a plane passing through the topmost layer of atoms on the segment. The total ESP which is the summation of nuclei potential and Hartree potential can be used to describe the impact of net charge approximately. The contour plots of the ESP function indicate positive ESP values for all atoms in Pt1-PA and Ptc-PA, suggesting their electrophilic nature (Fig. 6d, e). Moreover, the Pt atoms display higher peak ESP values compared to the Pd atoms, suggesting a stronger electron trapping ability. Notably, the larger peaks on the Pt atoms of Pt1-PA indicate an electric field directed towards the surrounding Pd atoms. Consequently, this electric field may drive small molecules close to the surface of Pt1-PA towards Pt atoms, implying that Pt serves as the active center in catalytic reactions. The higher potential at the surface of Ptc-PA relative to Pt1-PA signifies its enhanced capability for absorbing small molecules. In brief, the schematic of Fig. 6f illustrates that the capable reaction mechanism during different enzyme-like activities is governed by the conversion between Pt0 and Pt2+, and the d-band center upshift of Ptc-PA with tensile strain could make a critical difference for the excellent enzyme-like catalytic activity. To further verify the strain effect on optimizing catalytic activity of Ptc-PA, adsorption energy and Pt-O bond length for H2O2 at different strain were accurately calculated (Fig. 6g). The H2O2 is rapidly decomposed by pure Pt, Ptc-PA and Ptc-P to generate OH* species when it approaches the surface, which is favorable for the downstream reaction. In comparison to the Pt-O bond lengths observed on the Ptc-P surface (1.985 Å) and the pure Pt surface (1.966 Å), the Ptc-PA surface exhibits shorter Pt-O bond lengths (1.945 Å), indicating a stronger bonding between OH* and Ptc-PA. Furthermore, as the strain transitions from negative to positive, the Pt-O bond length decreases about 0.03 Å on the surface of Ptc-PA, suggesting that a certain amount of tensile strain is favorable for Ptc-PA to adsorb H2O2 and catalyze its decomposition. In addition, the H2O2 adsorption energy on surface of Ptc-PA with tensile strain (−2.041 eV) is more optimal than those on pure Pt and Ptc-P (Fig. 6g), demonstrating the vital roles of strain, Au layer and Pd layer in enhancing the catalytic performance of Ptc-PA. Therefore, these findings highlight that the tensile strain can effectively improve the adsorption capacity for H2O2 substrates, in turn boost the catalytic activity of Ptc-PA, which would be used to guide and design of high-performance catalysts.

Ptc-PA-based LFA for cancer patient detection

Development of sensitive detection tools facilitates early clinical monitoring of major disease progression. Taking advantages of Ptc-PA prominent POD-like activity and robust stability, we further introduced the Ptc-PA, as probes in lateral flow assay (LFA). The LFA widely used in testing at home, community and field, and often limited by the detection sensitivity. The Ptc-PA probes may effectively resolve the aforementioned hurdles with two kinds of signal readouts including visual readout and catalytic readout (Fig. 7a). Due to the POD-like activity of Ptc-PA, the catalytic signal was measured through recording the absorbance with a microplate reader, allowing for accurate quantitative detection of low-abundance analytes and different application requirements. (Supplementary Fig. 28).

Fig. 7: Assay performances of PSA and CEA by the Ptc-PA-based LFA.
figure 7

a Schematic illustration of the Ptc-PA-based LFA for sensitive detection of PSA and CEA. Specificity analysis to different nontarget proteins under colorimetric and catalytic modes with Ptc-PA-based LFA for (b) PSA and (c) CEA. Colorimetric and catalytic photographs of visual recognizable color changes in the reaction system for (d) PSA and (e) CEA. Numbers 1−15 represent 6.4, 3.2, 1.6, 0.8, 0.4, 0.2, 0.1, 0.05, 0.025, 0.0125, 0.00625, 0.00313,0.00155, 0.0007525 and 0 ng/mL of PSA or CEA, respectively. Relationships of the T-line intensity and the absorbance value of 652 nm with different concentrations of (f) PSA or (g) CEA (Insert image: The calibration curve of colorimetric and catalytic signal versus concentration of PSA or CEA, n = 3 independent experiments, data are presented as mean ±SD). Recovery rates of (h) PSA and (i) CEA spiked in serum samples under colorimetric and catalytic modes. j Detection results of PSA in 50 clinical serum samples. k Heat map showing the assay results of the 50 clinical serum samples. l Correlation analysis between the Ptc-PA-based LFA and the electrochemical assay for the detection of PSA in serum samples. m Detection results of CEA in 30 clinical serum samples. All ‘arb. units’ on the axis represent arbitrary units.

Full size image

Prostate specific antigen (PSA) and carcinoembryonic antigen (CEA), which were considered crucial biomarkers for the diagnosis of prostate cancer and colorectal cancer, are selected as the model analytes in clinical trials65,66. Sensitive detection of PSA or CEA facilitates early diagnosis of cancer and monitoring of cancer treatment progress. To verify the specificity of our assay system for PSA and CEA antigens, we used two strips to detect a variety of protein markers. Figure 7b, c indicates that this system possessed excellent inclusiveness and exclusiveness for the detection of PSA or CEA. The visual detection limit of the Ptc-PA-based LFA without catalytic enhancement mode had been achieved at an ultralow level of 0.1 ng/mL for PSA or CEA, which was about an order of magnitude lower than the detection limit reported in the literature (Fig. 7d, e). Notably, the visual detection limit of these two biosensors with catalytic enhancement mode (0.7525 pg/mL) was 132-times lower than the naked eye detection limit, approaching the sensitivity of electrochemical detection (Fig. 7d, e). As expected, the T-line color intensities and the catalytic signal enhanced as the concentrations of PSA or CEA increased continuously (Fig. 7f, g). As for PSA, the limit of detection (LOD) with catalytic enhancement system enabled the detection down to 0.33 pg/mL, while the mode without enhancement was 11.08 pg/mL (Supplementary Table 5). As for CEA, the LODs were calculated to be 0.063 ng/mL and 0.46 pg/mL for the Ptc-PA-based LFA and the catalytic enhancement model, respectively (Supplementary Table 6). Different concentrations of PSA or CEA standards were added to four healthy serum samples and calculated of spiked-recovery. The average recoveries of standard addition for PSA were calculated from 89.26% to 113.13%, and the coefficient of variations (CV, n = 3) values were at 0.47%–9.78%, indicating that the developed method was not impacted by biological matrices (Fig. 7h and Supplementary Table 7). Figure 7i and Supplementary Table 8 also demonstrated the specificity of the assay in detecting CEA, and spiked standard samples confirmed its stability against biological matrices interference.

Finally, the clinical application of Ptc-PA-based LFA was evaluated by detecting PSA or CEA in human serum samples. The veracity and practicability of signal-enhanced Ptc-PA-based LFA was further verified by detecting 50 clinical serum samples of PSA (Fig. 7j). The detection results were summarized in Supplementary Fig. 29, which were consistent with the electrochemical detection results, and the correlation coefficient was 0.99 and 0.97 before and after catalytic enhancement, respectively (Fig. 7k, l and Supplementary Fig. 30). Besides, we also validated the accuracy and practicality by testing 30 clinical serum samples with varying levels of CEA (Fig. 7m and Supplementary Fig. 31). Taken together, these results highlight the excellent capability of Ptc-PA-based LFA in cancer marker detection. Notably, with the addition of a catalytic enhancement system based on POD-like enzyme reaction, the sensitivity of this bio-detector can be higher than that of currently known detection systems, which possesses a great potential for immediate cancer diagnosis.

Modulation of radiation oxidation

Artificial enzymes, as a highly stable and inexpensive biocatalyst have been widely applied in the field of biomedicine45,67,68,69,70,71,72. Benefiting from the excellent catalytic activity of Ptc-PA, the inhibition of oxidative damage was further investigated in biological systems. Ptc-PA showed almost no toxicity to several common cell types (CHO, HT22, BV2, MA-c, and NCM460 cells) at concentration up to 300 μg/mL, pointing the favorable biosafety (Fig. 8a and Supplementary Fig. 32). The survival of H2O2 or LPS-treated CHO cells were sharply reduced to 65%–68% (Fig. 8b and Supplementary Fig. 33), while the survival of Ptc-PA-treated cells was remarkably increased to 90%. In particular, the survival rate of cells treated with Ptc-PA after exposure to 4 Gy gamma rays was more than 90%, whereas the viability in the radiation group was significantly reduced to 66% (Fig. 8b). This was attributed to the superior enzyme-like catalytic activity of Ptc-PA73,74. Fluorescence staining at the cellular level and the corresponding flow cytometry results also demonstrated the ability of Ptc-PA to effectively scavenge free radicals (Fig. 8c–g and Supplementary Fig. 34). Exposure to high levels of H2O2 and LPS can result in cytotoxicity and impair the normal function of cells, triggering apoptosis and necrosis. To examine the influence of Ptc-PA on the functionality of damaged cells, we conducted cell cycle and apoptosis assessments. Our findings indicated that Ptc-PA mitigates the accumulation of S-phase arrested cells induced by H2O2 or LPS, suggesting the restoration of DNA replication inhibition (Supplementary Fig. 35). Concurrently, Ptc-PA intervention also ameliorates apoptosis (Fig. 8h, i). The results suggest that Ptc-PA could rescue cell oxidation damage and possess potential applications in the oxidative regulation of biological systems.

Fig. 8: Radioreduction levels in vitro and in vivo before and after treatment of Ptc-PA.
figure 8

a Survival rate of CHO cells after coculture of Ptc-PA with different concentrations (3, 6, 12, 25, 50, 100, 200, 300 μg/mL) for 24 h and 48 h (n = 6 per group). b In vitro protection effect of Ptc-PA with different concentrations (n = 3 per group). Data are presented as mean ± SEM, analyzed by one-way ANOVA with one-sided Tukey’s multiple comparisons test (the p values are shown). Fluorescence microscopic images of intracellular (c) ROS and (d) O2·− levels under different conditions. Experiments were repeated independently five times with similar results. e Fluorescence quantification of CHO cells staining for ROS and O2•− (n = 5 per group). Data are presented as mean ± SEM, analyzed by one-way ANOVA with one-sided Tukey’s multiple comparisons test (the p values are shown). Quantitative analysis of intracellular (f) ROS and (g) O2·− levels under different conditions by flow cytometry. h Cell apoptosis of CHO cells stimulated by 800 μM H2O2 and 1 mg/mL LPS with and without Ptc-PA treatment. i Statistical diagram of early apoptosis and late apoptosis obtained from cell apoptosis test (n = 3 per group). Data are presented as mean ±SEM, analyzed by one-way ANOVA with one-sided Tukey’s multiple comparisons test (the p values are shown). j Schematic diagram of Ptc-PA modulated radiation-induced oxidative damage (Created with BioRender.com). k Survival percentage of irradiated mice with or without treatment of Ptc-PA, as well as Amifostine as a control. l The level of WBC and PLT of mice with or without the treatment of Ptc-PA (n = 3 per group). Data are presented as mean ±SEM, analyzed by one-way ANOVA with one-sided Tukey’s multiple comparisons test (the p values are shown). m The level of DNA and BMNC of mice with or without the treatment of Ptc-PA (n = 5 per group). Data are presented as mean ± SEM, analyzed by one-way ANOVA with one-sided Tukey’s multiple comparisons test (the p values are shown). The level of (n) Lung SOD, Liver SOD, (o) Lung MDA and Liver MDA of mice with or without the treatment of Ptc-PA (n = 3 per group). Data are presented as mean ±SEM, analyzed by one-way ANOVA with one-sided Tukey’s multiple comparisons test (the p values are shown).

Full size image

With the application of nuclear technology, health problems caused by radiation are attracting rising concerns. Therefore, biological effect of Ptc-PA on radiation oxidation was carried out (Fig. 8j). Toxicological assessments are indispensable in the biological applications of nanomaterials, offering valuable insights into their biosafety profiles. The body weight of mice increased steadily within 28 days after intraperitoneal injection of Ptc-PA (Supplementary Fig. 36a). The major organ index and blood routine of the mice were at normal levels (Supplementary Figs. 36b and 37). Pathological studies showed that intraperitoneal injection of Ptc-PA did not cause toxicity to the major organs of mice (Supplementary Fig. 38). In conclusion, Ptc-PA possesses positive biosafety and potential for biomedical application. As a golden index to judge the effect of radiation protection, the survival rate of radiation mice with Ptc-PA treatment increased from 20% to 80% compared with the untreated group, reaching the level of the clinical adjuvant Amifostime (Fig. 8k). This indicates that the radiation protection effect of Ptc-PA is comparable to that of the clinical adjuvants. In addition, the corresponding routine blood indices were significantly restored (Fig. 8l and Supplementary Fig. 39). The changes induced by Ptc-PA in the radiation characteristic markers, total bone marrow DNA, and bone marrow mononuclear cells (BMNC) reflected the exceptional in vivo regulation of radiation oxidation75,76. Ptc-PA effectively protected the hematopoietic system of mice and improved the hematopoietic function of irradiated mice (Fig. 8m). Variations in organismal biomolecules such as SOD and malondialdehyde (MDA) reflect the degree of oxidative damage in the body43,54,74,77. The SOD and MDA contents in the liver and lung of irradiated mice were similarly restored to normal levels, suggesting the protective effect of Ptc-PA in irradiated mice (Fig. 8n, o). In brief, Ptc-PA as an excellent biocatalyst has shown notable oxidative damage modulating effects in the organism, revealing its great potential as a biocatalyst for oxidative stress-induced biological applications.

Discussion

In summary, we have systematically investigated the atomic strain-activity correlation during the atomic structure evolution from Pt1, Ptn to Ptc, demonstrating the efficient enzyme-like catalytic mechanism. Surface atomic precision Pt clusters maximize the exposition of catalytic active sites. Combining structural analysis and quantum mechanical calculations accurately of Ptc-PA suggests that tensile strain effectively promotes the emergence of Pt vacancies and active low coordination sites. Besides, the upshift of the d-band center and the generation of high-energy charge carriers also play an equally important role in regulating the enzyme-like catalytic activity. The Ptc-PA exhibits ~104-fold and 61.5-fold higher POD-like and CAT-like activities than the natural POD and CAT, respectively. In addition, the SOD-like activity of Ptc-PA is equivalent to natural SOD enzymes. Biological results reveal that the catalytic colorimetric assay for Ptc-PA achieves a LOD 132 times lower than the naked eye due to the notable POD-like activity of Ptc-PA, resulting in a sensitive clinical diagnosis of cancer in patients. Moreover, Ptc-PA significantly modulates radiation-mediated oxidative damage and survival rate. Our work provides a reliable idea for designing biocatalysts with ultrahigh catalytic activity.

Methods

Materials

All chemicals are commercially available with the highest purity and used without further treatment. Gold chloride (HAuCl4·3H2O, ≥99.9%), Sodium tetrachloroplatinate (K2PtCl4, ≥99.9%), Potassium tetrachloropalladate (Na2PdCl4, ≥99.99%), glutathione (GSH, ≥98%), cysteine (Cys, ≥99%), Ascorbic acid (AA, ≥99.99%) and PVP (K30, Mw = 40,000, ≥ 99.99%) were purchased from Aladdin. Phosphate-buffered saline (PBS) buffer (0.01 M, pH = 7) and bovine serum albumin (BSA, ≥98%) was purchased from Solarbio. Adjust the PBS or BSA solution to a pH = 5 solution using 0.1 M HCl (≥99.99%) to prepare for use. Ultrapure water (18.2 MΩ*cm) was used for all the experiments. Nitrocellulose (NC) membrane, glass fiber membrane, polyvinyl chloride (PVC) backing card, and absorbent pad were obtained from Shanghai Kinbio Technology Co., Ltd (Shanghai, China). TMB (99%) single-component substrate solution and 30% H2O2 (99.8%) was purchased from Sigma-Aldrich Chemical Co, Ltd (St. Louis, USA). Goat anti-rabbit IgG antibody (bs-0295G, 0.5 mg/mL), Goat anti-mouse IgG antibody (bs-0296G, 0.5 mg/mL), PSA, CEA were all obtained from Beijing Bioss Biotechnology Ltd. PSA capture antibody (Ab1, SDT-195-50, 1 mg/mL) and PSA labeled antibody (Ab2, SDT-195-69, 1 mg/mL,) were obtained from Starter Bioscience Co.Lt. CEA capture antibody (Ab1, 3CEA-23, 0.5 mg/mL) and CEA labeled antibody (Ab2, CEA-100, 1 mg/mL) were obtained from Fapon Biotech. The human serum samples with different PSA or CEA concentrations were obtained with the patients’ permission and collected by the Tianjin Medical University Affiliated General Hospital.

Materials preparation

The Ptc-PA was synthesized according to the previous literature78. In a typical synthesis, aqueous solutions of HAuCl4 (20 mM, 2.5 mL), K2PtCl4 (20 mM, 2.5 mL), Na2PdCl4 (20 mM, 2.5 mL) and PVP (0.1 g) were homogenized by ultrasound, followed by the addition of AA solution (0.4 M, 1 mL) under sonication. The difference in the synthesis of Pt single atom (Pt1), several Pt atoms (Ptn) and atomically resolved Pt clusters (Ptc) on (Au core) @ (Pd thin shell) is that the input of K2PtCl4 is 10%, 20% and 40% of the original input molar amount, respectively, all other conditions are the same. Then the whole solution was placed for 6 h. The product was collected by centrifugation at 13,400 × g for 20 min, and the final product was redispersed in PBS. The synthesis of Pt, Au, Pd, AuPt, AuPd and PtPd were also based on the same method.

Materials characterization

TEM and high-resolution TEM (HRTEM) characterizations were carried out using a JEOL JEM-2100F operated at 200 kV equipped with energy-dispersive spectrometry analyses. The aberration-corrected scanning transmission electron microscopy (AC-STEM, ARM200F, JEOL, Japan) was adopted to observed the atomic and cluster information of NPs. The atomic number of Pt is calculated from the TEM measurements and the following equation:

$$N=\frac{\frac{4}{3}\pi {r}^{3}}{{V}_{2}}{N}_{A}$$
(5)

where the N is the atomic number of Pt in Ptc-PA, r is the radius of Pt cluster in Ptc-PA. For Ptc-PA (Pt: Pd: Au=0.4:1:1), r = 1.225 nm. For Ptc-PA (Pt: Pd: Au=0.5:1:1), r = 1.224 nm. For Ptc-PA (Pt: Pd: Au=1:1:1), r = 1.228 nm. V2 is the volume of Pt atom, V2 = 9.1 cm3/mol. NA is Avogadro constant, NA = 6.02214076 × 1023.

The XAFS spectra of Au L3-edge and Pt L3-edge in Ptc-PA were tested and provided by Beijing Synchrotron Radiation Facility. The XAFS results were analyzed by the ATHENA and ARTEMIS modules of IFEFFIT software packages. XRD pattern was recorded with a Rigaku Rint 2500 diffractometer with monochromated Cu Kα radiation. UV–vis absorption spectra were recorded on Shimadzu 3600 UV–Vis–NIR spectrophotometer. XPS analyses were carried out using a ThermoFisher Scientific spectrometer with a monochromatic Al Kα X-ray source with 300 W operating power. The content of metallic elements in NPs was tested by the 7900 ICP-MS (Agilent, US).

Enzyme-like activity test

POD-like activity test

POD-like activity of Ptc-PA was determined by colorimetric method. First, the working solution for TMB substrate color development kit (Sbjbio, Nanjing, China) was prepared according to the provided instructions. Subsequently, Ptc-PA NPs (20 μL) and working solution (180 μL) were mixed in a 96-well plate, and the absorbance at 652 nm (characteristic absorption peak of TMB) was tested at different NPs concentration and reaction time. The testing procedure for HRP followed the same methodology. For the kinetic tests, the reaction rates of samples were tested using UV–vis spectroscopy at different concentrations of TMB or H2O2. In brief, Ptc-PA (2 μL, 2.5 mg/mL), TMB (2 μL, 0–800 mM), H2O2 (2 μL, 50 mM), and acetic acid buffer solution (198 μL, pH = 4.5) were added to a 96-well plate. The absorbance of the reaction mixture was measured immediately at 652 nm using UV–vis spectroscopy and recorded over time. To determine the reaction rate of the sample towards different concentrations H2O2 (0–25 mM), TMB concentration was set as 800 mM. The initial reaction rate was calculated by Microsoft Excel. The maximum reaction velocity (Vmax) and Michaelis–Menten constant (Km) were calculated using the Michaelis–Menten equation:

$$V={V}_{\max }\frac{[E]}{{K}_{m}+[E]}$$
(6)

The kinetic constant (Kcat) was calculated based on the equation:

$${K}_{{cat}}=\frac{{V}_{\max }}{\left[E\right]}$$
(7)

where V represents reaction rate, Vmax represents the rate of reaction when the catalyst is saturated with substrate, Km value is termed as Michaelis–Menten constant and is the concentration of the substrate at which the rate of reaction is the maximum rate of reaction typically. [E] represents particle concentration of Ptc-PA used in the catalytic reaction.

The [E] was calculated based on the equation:

$$[E]=\frac{{{{{\rm{Au}}}}}_{{{{\rm{ICP}}}}-{{{\rm{MS}}}}}}{{m}_{{{{\rm{A}}}}{{{\rm{u}}}}}\times {N}_{A}}$$
(8)

where the [E] is particle concentration of Ptc-PA used in the catalytic reaction, AuICP-MS is the mass concentration of Au in the reaction system measured by ICP-MS. NA is Avogadro constant, NA = 6.02214076 × 1023. mAu is the mass of individual gold particles. The mAu was calculated based on the equation:

$${m}_{{{{\rm{A}}}}{{{\rm{u}}}}}=\frac{4}{3}\times \pi {r}^{3}\times \rho$$
(9)

where the r is radius of individual gold particles. The density of gold ρ = 19.3 g/cm3.

CAT-like activity test

Ptc-PA with different concentrations (0–10 ng/μL) were added to the H2O2 (40 mM), and the resulting absorbance changes at 240 nm were measured using a UV–vis spectrometer to evaluate its CAT-like activity. For kinetic tests, the reaction rates of Ptc-PA for the decomposition of H2O2 with different concentrations (0–50 mM) was tested using the kinetic mode of UV–vis spectrophotometer. The initial reaction rate was calculated using Microsoft Excel, then Vmax and Km were determined by the curve simulated from Origin software.

SOD-like activity test

According to the protocol provided in the SOD Assay Kit (Dojindo, Japan), appropriate volumes of assay working solution and enzyme working solution were prepared. For the experimental group, Ptc-PA with different concentrations (0–0.5 mg/mL, 20 μL), working solution (200 μL) and enzyme working solution (20 μL) were added sequentially to a 96-well plate. After incubation at 37 °C for 20 min, the absorbance at 450 nm (A450) was measured using microplate spectrophotometer (CMax Plus, Molecular Devices). The elimination rate of the superoxide was calculated by quantifying the decrease of A450.

GPx-like activity test

The GPx-like activity of Ptc-PA was determined following the previous literature. Briefly, a reaction mixture containing 200 μM H2O2, 2 mM GSH, 200 μM NADPH, 1.7 units mL–1 GR, and Ptc-PA (0–10 ng/μL) was added to a volume of 200 μL PBS neutral buffer. The changes of absorbance at 340 nm were recorded by the UV–vis spectrometer, which represents the concentration of NADPH.

•OH scavenging test

The •OH scavenging process was investigated using an ESR spectrometer (Bruker EMX plus, Germany). Initially, a mixture containing 5 mM H2O2 and 20 mM FeSO4 was employed to generate •OH radicals, which were then captured by 50 mM 5-tert-butoxycarbonyl 5-methyl-1-pyrroline N-oxide (BMPO) to produce spin adduct (BMPO/•OH) with four peaks in the ESR spectrometry. The ability to scavenge •OH was determined by detecting the change in peak intensity before and after the addition of Ptc-PA.

O2 •– scavenging test

The O2•– scavenging process was also examined by an ESR spectrometer (Bruker EMX plus, Germany). We first used 2.5 mM KO2 and 3.5 mM 18-crown-6 to generate stable O2•–, which were then captured by 25 mM 5-(diethoxyphosphoryl)−5-methyl-1-pyrroline-N-oxide (DEPMPO) to produce spin adduct (DEPMPO/O2•–) with six peaks under ESR spectrometry. The ability to scavenge O2•– was determined by detecting the change in peak intensity before and after the addition of Ptc-PA.

Electrochemical test

The electrochemical properties of the materials were characterized mainly by cyclic voltammetry on the setup produced by Shanghai Chenhua, employing a three-electrode system (GC electrode as the working electrode, Pt wire as the counter electrode, and saturated calomel electrode as the reference electrode). Ptc-PA (0.5 mg/mL, 20 μL) was dropped onto the surface of the GC electrode and dried at room temperature. After that, 3 μL of perfluorinated sulfonic acid (5%w/w) was added to serve as a waterproof protective layer and dried further at room temperature. During the drying period, N2 was uniformly injected into the electrolyte (0.1 M PBS) for 30 min. For the H2O2 reduction reaction, 50 µL of 30% H2O2 stock solution was added to the N2-saturated electrolytic cell. Subsequently, the electrode device was connected and CV scans were performed at room temperature using a scanning rate at 50 mV/s. The electrochemical test for ORR requires the same sample preparation operation procedure, while only O2-saturated electrolyte without H2O2 replaced N2-saturated electrolyte. The CV scans were performed directly, while the control group was tested using empty platinum-carbon electrodes.

DFT calculations

The DFT calculations were performed using the Vienna ab initio Simulation Package (VASP 6.1.2)79,80. The projector augmented wave method was employed to represent the elemental core and valence electrons81. To estimate the exchange-correlation potential energy, the Perdew–Burke–Ernzerhof generalized gradient approximation (GGA-PBE) functional was utilized and the cutoff energy was set to 520 eV for the plane-wave basis82. The electron-ion interaction was described using a norm-conserving pseudopotential, where the valence-electron configuration for Au, Pt and Pd atoms consisted of 5d106s1, 5d96s1 and 4d10, respectively. The simulation employed a seven-layer supercell arranged specifically along the (111) crystal plane, comprising of three layers of Au, two layers of Pd, and two layers of Pt. The supercell had dimensions of 8.55 Å × 5.70 Å, accommodating a total of 42 atoms. To avoid the interaction between layers caused by periodicity and to enhance the reliability and accuracy of calculations, a vacuum distance of 35 Å was implemented along the z-direction, which is perpendicular to the surface. In addition, the Monkhorst Pack scheme was employed for all Brillouin zone integrations83, and a 5 × 7 × 1 k-mesh was utilized for relaxation calculations. For structural relaxation, the convergence criteria were set to 1 × 10−6 eV for energy and −0.02 eV/Å for force. Dipole corrections were applied in all the DFT calculations. The calculation, with or without strain, allowed for uniform expansion (contraction) of the (111) lattice in all three Cartesian directions while excluding corrections to the interlayer distance. Some post-process wave function analyses including ESP were performed using Multiwfn software84. In this work, the binding energies of different structures were calculated by the following equation:

$${E}_{{f}_{a@b}}={E}_{A@B}-{E}_{A}-{E}_{B}$$
(10)

Where A is Ptc-PA or Pt, B is H2O2. EA and EB are the total energies of A species and B species, respectively.

The d-band center of different structures was calculated by the following equation:

$${\varepsilon }_{d}=\frac{{\int }_{\!\!-\infty }^{\infty }x\rho (x)dx}{{\int }_{\!\!-\infty }^{\infty }\rho (x)dx}$$
(11)

Where ρ(x) represents the distribution of electronic density of states as a function of energy (x).

The study conducted by Tianou He et al. demonstrates that Pt clusters exhibit enhanced catalytic activity when subjected to structural strains within the range of ±(2–4)%10, which is consistent with our experimental findings. Considering that different structures and strains can significantly influence the electronic structure and catalytic performance of materials, we have chosen to simplify the variables by setting both tensile and compressive strains at 3%. All the structural strains are applied in the [110] direction. Notably, strain is additionally applied to the optimized structure after the original modeling in order to investigate the impact of strain on the electronic structure and catalytic activity of these clusters.

PSA and CEA detection

Conjugation of Ab2 to Ptc-PA

The conjugation of Ab2 to the surface of Ptc-PA was achieved via a simple physical absorption procedure. First, 10 μL Ab2 (1 mg/mL) was added to 1 mL of the prepared Ptc-PA solution (0.4 mg/mL) and incubated for 1 h at room temperature. Subsequently, BSA solution (100 μL, 10 mg/mL) was added to the mixture as a blocking agent. After another 1 h, centrifugation (10621 × g, 10 min) was performed to recover the final product (Ptc-PA-Ab2 probe), which was then redispersed in PBS (100 μL, pH 7.4). Finally, the Ptc-PA-Ab2 probe was stored at 4 °C for further experiments.

Fabrication of the Ptc-PA-based LFA test strip

The Ptc-PA-based LFA test strip (3.5 mm width) contains five main parts: NC membrane with a test line (T-line) and a control line (C-line), sample pad, conjugate pad, PVC backing card, and absorbent pad. The conjugate pad was pre-blocked with 4 mL of immune buffer (5 mM PVP-10000, 0.15 M sucrose, 0.45 mM BSA, 2% Tween-20 in 0.01 M PBS) and dried at 37 °C for 8 h. The Ab1 (0.5 mg/mL) and goat anti-rabbit IgG antibody/anti-mouse IgG antibody (0.5 mg/mL) were dispensed onto the porous NC membrane at a jetting rate of 1.0 μL/cm, respectively. The interval between the two lines was 4 mm apart before drying completely at ambient temperature. Finally, different pads were constructed on a PVC backing card with an overlap of ~2 mm before cutting into 3 mm by guillotine cutter. These fabricated strips were stored in a refrigerator at 4 °C for the subsequent assays.

Protocols for the Ptc-PA-based LFA

A series of PSA or CEA standard solutions with different concentrations (6.4, 3.2, 1.6, 0.8, 0.4, 0.2, 0.1, 0.05, 0.025, 0.0125, 0.00625, 0.00313, 0.00155, 0.0007525 and 0 ng/mL) were prepared in an assay buffer (0.01 M PBS, pH 7.4). Briefly, 3 μL of the Ptc-PA-Ab2 probe was incubated with 60 μL of PSA or CEA solutions and dispensed onto the sample pad to migrate under capillary action for 20 min. Then, optical images of the strips were directly recorded using iPhone 13 pro smartphone-based imaging device, and the pixel intensity within the T-line regions was calculated using the ImageJ software. In the catalytic enhancement mode, the T-lines in the strips were clipped and transferred to substrate solution (a mixture containing TMB single-component substrate solution and 3% H2O2). Due to its POD-like catalytic activities, Ptc-PA facilitated obtaining a catalytic signal by recording the absorbance value of 652 nm with a multimode plate reader after incubation at room temperature for 5 min. The specificities experiment of Ptc-PA-LFA involved studying PSA or CEA (3.4 ng/mL) along with other proteins. All reactions were repeated three times.

Detection of the clinical samples

The human serum samples with different concentrations of PSA or CEA were obtained with the patients’ permission and collected by the Tianjin Medical University Affiliated General Hospital. Serum samples from 20 healthy men were included in this project. 40 male patients with prostate cancer and 20 with colorectal cancer, aged between 30 and 65 years old. This study was approved by the Tianjin Medical University Affiliated General Hospital, and all blood donors signed informed consent (IRB2020-KY-097). All assays were conducted and adhered to legal requirements and ethical guidelines. The standard addition method was employed to introduce different concentrations of PSA or CEA standards (3.2 ng/mL, 1.6 ng/mL, and 0.8 ng/mL) to four healthy serum samples by the standard addition method. The serum samples obtained from patients with varying concentrations were analyzed using a specially designed biosensor and compared to the results obtained from a standard electrochemical sensor.

In vitro experiments

Chinese hamster ovary CHO cells and mouse hippocampal neuronal HT22 cells were obtained from the Institute of Radiation Medicine, Chinese Academy of Medical Sciences and Peking Union Medical College. Mouse microglia BV2 cells and mouse astrocytes-cerebellar MA-c cells were obtained from Tianjin Huanhu Hospital. Human colonic epithelium NCM460 cells were obtained from Tianjin Medical University General Hospital. NCM460 cells were cultured in McCoy’s 5A (Gibco) with 10% fetal bovine serum (FBS, BI) at 37 °C with 5% CO2. Other cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Gibco) with 10% fetal bovine serum (FBS, BI) at 37 °C with 5% CO2. Penicillin (100 μ/mL) and streptomycin sulfate (100 mg/mL) were added as required for cell growth.

Cytotoxicity assay

The above cells in the logarithmic growth phase were seeded into sterile 96-well plates at a density of 4 × 103 cells per well. Sterile PBS was added around the perimeter of the plates to replenish evaporated water during culture. After seeding, the cells were incubated overnight in a sterile incubator (37 °C, 5% CO2) until fully attached to the wall. When the cell density reached 60%, co-incubated with Ptc-PA at different concentrations for another 24 h or 48 h. Cells were washed with sterile PBS and added to serum-free medium. 3-(4,5)-dimethylthiahiazo (-z-y1)-3,5-di- phenytetrazoliumromide (MTT) with the concentration of 5 mg/mL was added to each well and incubated for 2.5 h. Dimethyl sulfoxide was added to dissolve the precipitate after discarding the supernatant of the medium and the absorbance at 490 nm was tested in each well.

Cell viability

CHO cells (4 × 103), HT22 cells (4 × 103), BV2 cells (4 × 103) and MA-c cells (4 × 103) were cultured in the 96-well plate. 1 mg/mL LPS or 500 µM H2O2 was added to cells at ~60% confluence and stimulated for 4 h. The original medium was discarded and replaced with medium containing Ptc-PA with different concentrations for overnight incubation. The wells were washed with PBS to remove serum and subsequently co-incubated with the cells for 2.5 h using 5 mg/mL MTT. Cell viability was compared between the different treatment groups by analyzing the absorbance at 490 nm. To assess the effect of Ptc-PA on cell viability in response to radiation, CHO cells were inoculated into 96-well plates at a density of 4 × 10³ per well and cultured overnight. The cells were co-incubated with Ptc-PA at various concentrations (6–200 μg/mL) for 1 h, after which the plates were exposed to 4 Gy gamma rays and incubated for 24 h at 37 °C. The medium was then removed and the cells were washed twice with PBS. Cell viability was assayed according to the instructions in the Cell Counting Kit-8 (Beyotime, C0037).

Measurement of intracellular oxidative stress

CHO cells (2 × 105) were seeded into 6-well plates and cultured overnight in an incubator until fully attached to the surface. 0.5 mg/mL LPS and 200 µm H2O2 were then used to stimulate the cells for 6 h, followed by the intervention of Ptc-PA (25 μg/mL). The cells were then placed in an incubator for 24 h and cleaned 3 times with PBS. The culture medium was replaced with 5 µM DCFH (Beyotime, S0033S) or 25 µM DHE (Beyotime, S0063) probe solution, and incubated for 25 min at 37 °C in the dark to determine the total ROS level or O2•– level, respectively. The probe solution was removed, and the cell was washed again 3 times with PBS. Fluorescence images of cells from different treatment groups were captured using a fluorescence microscope (EVOS, AMG), and quantitative analysis was conducted by the flow cytometer (BD AccuriTM C6).

Cell apoptosis

The measurement of apoptosis was conducted using the FITC Annexin V Apoptosis Detection Kit I (BD, No. 556547). CHO cells were cultured overnight in 6-well plates. When growth reached 60% confluence, cells were stimulated with 800 μM H2O2 or 1 mg/mL LPS overnight and then cultured with medium containing 25 µg/mL Ptc-PA for another 24 h. Cells were detached by trypsin and collected into centrifuge tubes, centrifuged at 1000 × g for 5 min and the supernatant discarded. The cells were washed twice with PBS, centrifuged at 1000 × g for 5 min, and the supernatant was removed. The cells were then resuspended in an appropriate volume of 1× loading buffer to adjust the cell count to 106 cells/ml. 5 μL of FITC Annexin V was added to 100 μL of cell suspension and stained in the dark for 5 min. Afterward, 5 μL of propidium iodide (PI) was added, and the cells were stained for 15 min. The final assay was performed using flow cytometry (FACSAria III).

Cell cycle

The cell cycle was determined by quantitative DNA content assay (Solarbio, CA1510). CHO cells were cultured overnight in 6-well plates. When growth reached 60% confluence, cells were stimulated with 800 μM H2O2 or 1 mg/mL LPS overnight and cultured with medium containing 25 µg/mL Ptc-PA for another 24 h. Cells were washed with PBS and collected to prepare single cell suspensions. Cells were fixed with 70% ethanol overnight and washed once with PBS. The cells were centrifuged at 1000 × g for 5 min, and the supernatant was removed. After that, the cells were resuspended in RNase solution and incubated at 37 °C for 30 min. After cooling, PI dye solution (400 μL) was added to the sample, and the mixture was incubated at 37 °C in the dark for another 30 min. Flow cytometry analysis (BD AccuriTM C6) was performed on the sample.

In vivo treatment

All animal procedures were approved by the Institute of Radiation Medicine, Chinese Academy of Medical Sciences and Peking Union Medical College (IRM-DWLI-2021107). Efforts were made to reduce the number of animals used and minimize their suffering.

Toxicological studies

SPF Male C57BL/6J mice (6–8 weeks) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. The mice were housed under controlled conditions with a constant temperature range of 21–23 °C, relative humidity maintained between 45 and 60%, and a 12-h light-dark cycle. Food and water were available ad libitum. The mice were randomly divided into 2 groups (control group and Ptc-PA group, n = 3 per group). Each mouse was injected with 200 µL of saline or Ptc-PA solution (5 mg/mL), and body weights were recorded daily post-injection. On day 28, blood was collected for hematological analysis. Major organs (heart, liver, spleen, lung, and kidney) were collected, weighed, fixed in 10% neutral buffered formalin for 24–48 h, embedded in paraffin, sliced into thin sections 4-μm thickness and subjected to pathological Hematoxylin and Eosin (H&E) staining.

Animal models

SPF Male C57BL/6J mice (7–9 weeks, 21–23 g). were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. The mice were assigned to four groups (n = 25 per group) randomly. Both the Con group and the Rad group were intraperitoneally injected with 200 µL saline, while the Rad + Ptc-PA groups were intraperitoneally injected with 50 mg/kg Ptc-PA. Amifostine was injected intravenously into mice as a control. Thirty minutes post-injection, mice were anesthetized with sodium pentobarbital (1%, 50 mg/kg) by intraperitoneal injection (i.p.). All the groups of mice were then exposed to 7.5 Gy gamma rays covering the entire body. The irradiated mice were housed in a standard feeding environment with clean feed and water. Survival rates and weight changes were recorded daily. Blood samples were collected for routine blood tests on day 30, and serum was separated by centrifugation for blood biochemistry analysis.

BMNC and bone marrow DNA measurements

The mice were euthanized at the time points of 7 days after being irradiated and all organs were taken out (n = 5 per group). Both femurs were resected and cleaned of connective tissue. The bone marrow cells were flushed into PBS by the injector. To prevent interference from tissue debris and bone fragments, the samples were filtered through a 200-mesh nylon filter and then counted with a blood cell counter to compare the amount of BMNC in different groups of mice. For bone marrow DNA measurements, the bone marrow cells were rinsed with 5 mM calcium chloride solution, and the mixture was gently stirred to obtain a single-cell suspension. The suspension was then refrigerated at 4 °C for 2 h. The samples were then centrifuged at 700 × g for 15 min. After centrifugation, the supernatant was removed and the pellet was resuspended in 5 mL of 0.2 M perchloric acid solution and heated in a 90 °C water bath for 15 min. The samples were cooled to room temperature and filtered. Finally, the absorbance of filtrate was measured at 268 nm using a UV–vis spectrophotometer.

Oxidative stress

Liver and lung samples were collected from mice on day 7 after radiation exposure. Tissue homogenates were prepared by adding PBS to the samples and grinding until no tissue mass remained. The homogenate was centrifuged at 10,000 × g for 10 min. The supernatant was removed and stored at −80 °C. The protein concentration of the supernatant was measured using the Enhanced BCA Protein Assay Kit (Beyotime, P0010) according to the experimental procedure in the instructions provided by the manufacturer. The levels of MDA and SOD in the tissue samples were measured using the Lipid Peroxidation MDA Assay Kit (Beyotime, S0131S) and Total SOD Activity Assay Kit (WST-8, Beyotime, S0101M), respectively.

Statistic methods

Data are presented as mean ± standard deviation (SD) or standard error of the mean (SEM). For multiple comparison, one-way analyses of variance (ANOVA) with one-sided Tukey’s multiple comparisons test were used to assess difference in means among groups.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.