Neutrophil elastase-activatable ratiometric photoacoustic nanoprobes for imaging atherosclerotic plaque vulnerability

Abstract

Atherosclerosis is a leading cause of cardiovascular diseases worldwide. Neutrophil elastase, a key protease secreted by neutrophils, contributes significantly to atherosclerosis by degrading the extracellular matrix and destabilizing the fibrous cap of plaques. However, current detection methods for neutrophil elastase lack the ability to monitor its dynamic changes within plaques in a non-invasive and real-time manner. To overcome this limitation, we develop a neutrophil elastase-activated semiconductor polymer nanoprobe with high selectivity, sensitivity, and resistance to interference. We successfully establish a calibration curve correlating the in vivo photoacoustic signal ratio with neutrophil elastase concentration, enabling quantification of intraplaque neutrophil elastase levels. Importantly, we introduce a vulnerability index to assess plaque instability, allowing the nanoprobe to predict the adverse effects of inflammatory responses on the risk of plaque rupture in ApoE^-/- male mice. Additionally, the nanoprobe can visualize dynamic changes in neutrophil elastase during intimal hyperplasia under inflammatory conditions. In conclusion, this nanoprobe provides a tool for elucidating the role of neutrophil elastase in cardiovascular disease progression.

Introduction

Atherosclerosis, a leading cause of cardiovascular disease worldwide^1,2, often culminates in fibrous cap rupture. When the fibrous cap breaks, triggering thrombosis and leading to life-threatening cardiovascular events such as ischemic stroke, transient ischemic attack, myocardial infarction, abdominal aortic aneurysm, intermittent claudication, ulceration, and gangrene^3,4,5,6,7,8. Therefore, identifying plaques at high risk of rupture and assessing plaque vulnerability are of significant clinical importance. Atherosclerosis is now recognized as a chronic inflammatory disorder, with neutrophil-mediated inflammation driving lesion initiation, progression, and complications^{9,10,11,12,13}. Within plaques, activated neutrophils release large amounts of proteases, most notably neutrophil elastase (NE), which accounts for approximately 80% of total proteolytic activity in vivo^14,15. NE-mediated degradation of the extracellular matrix promotes smooth muscle cell migration into the intima and intimal hyperplasia¹⁶; in advanced lesions, excessive NE activity destabilizes the fibrous cap and triggers coagulation cascades, precipitating thrombosis^{15,17,18,19,20}. Accurate quantification of NE within plaques is thus essential for precise lesion characterization, risk prediction, and timely therapeutic intervention.

Existing detection methods—flow cytometry, ELISA, immunofluorescence, and immunohistochemistry—cannot noninvasively monitor NE dynamics in plaques in real time. Although PET permits in vivo NE detection, its spatial resolution is insufficient for plaque imaging²¹. Fluorescence imaging offers high sensitivity but suffers from photon scattering and limited penetration depth^{22,23,24,25,26}; chemiluminescence suffers from substrate-dependent signal variability²⁷. In contrast, photoacoustic (PA) imaging combines deep penetration, high sensitivity and high spatial resolution, making it well suited for visualizing intravascular plaques (Table S1)^{28,29,30,31,32,33,34,35,36}. Therefore, we set out to apply PA imaging for accurate, quantitative measurement of NE and precise assessment of plaque vulnerability. However, two major challenges must be overcome. (1) PA signals in plaques are subject to multiple interferences: probe accumulation and clearance kinetics cause temporal fluctuations, and local reactive oxygen species or other enzymes can affect probe responsiveness. Most NE-responsive PA probes operate at a single wavelength and cannot correct for these confounders, limiting them to qualitative in vivo analysis. (2) Correlating in vivo PA signals with actual NE levels in plaques is difficult. PA tomography can image entire vessels, but pinpointing the exact slice corresponding to a target plaque is nontrivial. Ex vivo dissection permits precise plaque localization and NE assay quantification, yet lacks any in vivo signal matching. Developing a strategy to match in vivo PA data with ex vivo NE measurements is essential for accurate NE quantification.

To overcome these challenges, we develope a NE-activated semiconductor polymer nanoprobe (SPNE) for quantitative in vivo detection of neutrophil elastase (NE) in plaques and evaluation of plaque vulnerability. SPNE comprises an NE-responsive probe (NERP) as the activatable PA module and a benzotriazole-based fusion ring as an NE-inert reference. In its resting state, SPNE produces only weak PA signals; after reacting with NE, a strong PA signal is generated at 695 nm, while the signal at 800 nm remains unchanged. The PA₆₉₅/PA₈₀₀ ratio thus directly reflect NE activity. SPNE shows excellent specificity against reactive oxygen species and other enzymes, and its signal is stable across varying scan depths, probe concentrations, and imaging times. It responds linearly to NE in the 0–2 U/L range (limit of detection: 27.957 mU/L), demonstrating feasibility for in vivo quantification. For spatial validation, we implanted fiducial markers on mice and matched their coordinates with in vivo PA images. After dissection, we mapped PA signals to target plaques, measured NE by ELISA, and established an in vivo calibration curve of PA₆₉₅/PA₈₀₀ to NE content. Then we performed histopathology to quantify the four important components of plaque vulnerability—macrophages, smooth muscle cells, necrotic cores, and collagen fibers, and derived a weighted vulnerability index, which correlated closely with the PA ratio. Using SPNE, we imaged three scenarios—carotid plaque, inflammatory plaque, and inflammatory plaque under combined anti-inflammatory and lipid-lowering therapy. Inflammation sharply increased NE levels and accelerated plaque destabilization, whereas combination therapy significantly reduced NE content and decreased the risk of plaque rupture. This indicates that simultaneous intervention targeting both inflammation and lipids can directly stabilize plaques. Finally, we successfully tracked dynamic NE changes in intimal hyperplasia. Altogether, SPNE offers a non-invasive, sensitive, and quantitative tool for monitoring NE in vivo.

Results

Design, synthesis, and properties of an activatable ratiometric photoacoustic probe for detecting neutrophil elastase (NE)

The hemicyanine molecule exhibits excellent chemical stability, and its absorption spectrum shows a significant blue shift when the ICT effect is inhibited, resulting in significant changes in the PA signal³⁷. So hemicyanine is a promising candidate for a PA response probe. In order to prepare the NE-response probe (NERP), we reacted the NE-cleavable peptide substrate N-acetyl-Ala-Ala-Pro-Val-COOH (Ac-AAVP-COOH) with a hemicyanine fluorophore (Hcy) via dehydration condensation synthesis (Fig. 1a). The chemical structure of Hcy was confirmed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS, Supplementary Fig. 1) and proton nuclear magnetic resonance (¹H NMR, Supplementary Fig. 4), and the chemical structure of NERP was also confirmed by the above characterization methods (Supplementary Figs. 2, 5).

Fig. 1: Design, synthesis, and properties of an activatable ratiometric PA probe for detecting neutrophil elastase (NE).

a The synthetic route of NERP. b Response principle of NERP to NE. c Absorption spectra of NERP incubated with NE (2 U/L) or not. d Fluorescence spectra of NERP incubated with NE (2 U/L) or not. e Synthesis scheme for SPNE via the nano-coprecipitation method. f The average hydrodynamic diameters of SPNE in H₂O. g TEM images of SPNE (n  =  3 independent samples). h Scheme for the ratio of Abs₆₉₅/Abs₈₀₀ SPNE when responsive to NE (2U/L) or not. i The absorption spectra of SPNE upon reacting with various concentrations of NE (0–2 U/L). j Quantification of Abs₆₉₅/Abs₈₀₀ ratios in i (n = 3 independent samples). k Fluorescence spectra of SPNE incubated with NE (2 U/L) or not. l Nonlinear regression analysis of cleavage rate V (mg/mL·s) of SPNE as a function of substrate (SPNE) concentration. m Absorption spectra of SPNE incubated with different concentrations of NE (0–15 U/L). n The Abs₆₉₅/Abs₈₀₀ of SPNE after incubation with different concentrations of NE, calculated from m (n = 3 independent samples). o Absorption spectra of SPNE incubated with NE (2 U/L) for different times (0–20 min). p The dynamic response: the Abs₆₉₅/Abs₈₀₀ of SPNE after incubation with NE for different time (0–60 min) (n = 3 independent samples). q The selective response: the Abs₆₉₅/Abs₈₀₀ ratios of SPNE after incubation with different interfering substance (NQO1: NADH quinone oxidoreductase 1, ALP: alkaline phosphatase, GGT: glutamyltransferase, MMP-2: matrix metalloproteinase 2, PEDs: phosphodiesterase) (n = 3 independent samples). Data are presented as the mean ± SD. Statistical significance was determined using a two-tailed Student’s t-test for pairwise comparisons and a one-way ANOVA analysis of variance for multiple groups.

The principle of the NERP response to NE is illustrated in Fig. 1b. Upon interaction with the peptide, the target molecule is positioned on the donor side of the chromophore, which weakens its electron-donating ability. This action inhibits the intramolecular charge transfer (ICT) effect, resulting in a decrease in the absorbance of NERP at 695 nm and a weakening of the fluorescence signal. After the probe reacts with NE, the target molecule departs from the chromophore, restoring its electron-donating ability, thus restoring the intramolecular charge transfer effect, resulting in increased absorbance at 695 nm and enhanced fluorescence signal.

Therefore, we investigated the spectral properties of NERP before and after the response to NE (Fig. 1c). We dissolved NERP in the Tris buffer (50 mM Tris, 1 M NaCl, 0.05% brij-35, pH = 7.4), then added NE and incubated at 37 °C for 30 min. The solution was subsequently analyzed by absorption and fluorescence spectroscopy. As shown in the absorption spectrum, NERP has an absorption peak near 660 nm, and the absorption at 660 nm decreases with the addition of NE, followed by a significant increase in absorbance at 695 nm. At the same time, we observed the fluorescence spectra before and after the response of NERP (Fig. 1d), and found NERP had almost no fluorescence signal from 660 to 800 nm, however, after incubation with NE, a clear fluorescence signal appeared around 720 nm. These findings indicated that the NE was able to cleave peptides from NERP in the Tris buffer system, resulting in a red shift of absorption and fluorescence turn-on.

Next, we aimed to construct ratiometric PA probes. To achieve this, we introduced an NE-nonresponsive molecule, which is a benzotriazol-based fused ring molecule with excellent PA stability (MALDI-TOF-MS and ¹H NMR are shown in the Supplementary Figs. 3, 6). NERP and NE-nonresponsive molecule were converted into nanoparticles by nanoprecipitation using poly (styrene-maleic anhydride) (PSMA) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(poly(ethylene glycol))] (DSPE-PEG₂₀₀₀) as surfactants (Fig. 1e). During the synthesis of nanoparticles, we added an excess of Na₂CO₃ to transform PSMA from a hydrophobic polymer into an amphiphilic polymer. Therefore, the hydrophilic polypeptide chain of NERP, the hydrophilic groups of DSPE-PEG, and PSMA will be jointly exposed on the surface of the nanoparticles, which facilitates the rapid binding of NE to the polypeptide substrate and its rapid removal. NE activatable semiconductor polymer nanoprobes (SPNE) were characterized and confirmed by DLS analysis to have an average diameter of about 40 nm and to remain azure blue transparent liquid in aqueous solution (Fig. 1f). In addition, these nanoparticles presented a spherical morphology as visualized by TEM (Fig. 1g). The colloid stability test revealed that the size of the nanoparticles remained at around 40 nm within 7 days, and the Zeta potential remained at −20 mV. It indicates that the nanoparticles have good colloid stability (Supplementary Fig. 7).

We then investigated the spectral properties of SPNE (Fig. 1h, i). SPNE was dissolved in the Tris buffer, followed by the addition of varying concentrations of NE and incubation at 37 °C for 30 min. In the initial state, SPNE exhibits characteristic absorption peaks of NERP and NE-nonresponsive molecules at 660 nm and 800 nm, respectively. Due to the inhibition of the intramolecular charge transfer (ICT) effect, NERP shows negligible absorbance at 695 nm, resulting in a low ratio signal (Abs₆₉₅/Abs₈₀₀). Following incubation with NE, the absorption spectrum of NERP redshifts, and the absorbance at 695 nm increases sharply. Since NE-nonresponsive molecules remain unaffected by NE, the absorbance at 800 nm does not change, leading to a significant increase in the Abs₆₉₅/Abs₈₀₀. Additionally, we observed that as NE concentration gradually increased, the absorbance at 695 nm correspondingly became larger. We analyzed the correlation between NE concentration and Abs₆₉₅/Abs₈₀₀, revealing a strong linear relationship (R² =  0.995) with a detection limit of 27.957 mU/L, determined by the 3ε/k method (Fig. 1j). Subsequently, fluorescence testing yielded consistent results, showing a significant increase in the fluorescence signal of SPNE at 720 nm following incubation with NE (Fig. 1k).

In order to investigate the affinity of SPNE to NE, we performed a Michaelis-Menten kinetics test. Various concentrations of SPNE solutions were incubated with NE (2 U/L) for 30 min at 37 °C in the Tris buffer. After incubation, the absorbance of the mixture was quantified by UV-vis absorption spectroscopy. Initial reaction velocities were then calculated, plotted against SPNE concentrations, and fitted to a Michaelis-Menten curve. The result showed that Michaelis-Menten constant (K_M) and catalytic rate constant (K_cat) of NE towards SPNE were calculated as 21.46 μM and 39.34 s⁻¹, respectively, resulting in a catalytic efficiency (K_cat/K_M) of 1.833 μM⁻¹s⁻¹ (Fig. 1l), which indicates that SPNE has a high affinity with the NE.

Next, we explored the response of SPNE to different concentrations of NE. The results show that there was a linear relationship between NE concentration and Abs₆₉₅/Abs₈₀₀ in the concentration range from 0 to 2 U/L. With the further increase of NE concentration, the reaction rate of SPNE with NE gradually reached saturation (Fig. 1m, n). Subsequently, we observed the SPNE response to NE at different times, SPNE can reach saturation in about 20 min, indicating that SPNE responds to NE extremely quickly (Fig. 1o, p). This is consistent with the conclusion of Michaelis-Menten kinetics. In order to verify the anti-interference ability of SPNE, we further tested the selectivity of SPNE to various active substances. Notably, when SPNE was incubated with reactive oxygen species and other enzymes, the Abs₆₉₅/Abs₈₀₀ hardly changed significantly, and only NE could greatly increase the Abs₆₉₅/Abs₈₀₀ (Fig. 1q). This indicates that SPNE has excellent selection specificity.

Ratiometric photoacoustic imaging for neutrophil elastase (NE) detection in solution

Next, we investigated the photoacoustic (PA) ratio signal of SPNE in solution for detecting NE concentration (Fig. 2a). When NE was incubated with SPNE, the PA signal of NERP was sharply enhanced at 695 nm. However, the PA signal of NE-nonresponsive molecules did not change at 800 nm, resulting in an increased PA ratio signal (PA₆₉₅/PA₈₀₀). To confirm this phenomenon, SPNE (5 μM) was dissolved in the Tris buffer, then NE (2 U/L) was added, incubated at 37 °C for 30 min, and finally PA spectra before and after the SPNE response to NE were collected using the MSOT system (Fig. 2b). The results showed that the PA intensity signal at 680 to 750 nm was significantly enhanced after SPNE incubation with NE, while the PA intensity at 750 to 900 nm remained essentially unchanged.

Fig. 2: Ratiometric photoacoustic imaging for neutrophil elastase (NE) detection in solution.

a Schematic of the PA spectrum response of SPNE. b PA spectra of SPNE incubated with NE or not. c PA images of SPNE at 695 nm and 800 nm incubated with different concentrations of NE (0–2 U/L). d PA₆₉₅ and PA₈₀₀ of SPNE incubated with different concentrations of NE (0–2 U/L) (n = 3 independent samples). e Linear relationship between concentration of NE and PA₆₉₅/PA₈₀₀ from d (n = 3 independent samples). f The ratios of PA₆₉₅/PA₈₀₀ in the presence of different interfering substance (NQO1: NADH quinone oxidoreductase 1, ALP: alkaline phosphatase, GGT: glutamyltransferase, MMP-2: matrix metalloproteinase 2, PEDs: phosphodiesterase) (n = 3 independent samples). g Scheme of PA testing at different scan sections. h PA₆₉₅ and PA₈₀₀ of SPNE at different scan sections (n = 3 independent samples). i PA₆₉₅ and PA₈₀₀ of SPNE incubated with NE (2 U/L) at different scan sections (n = 3 independent samples). j PA₆₉₅/PA₈₀₀ of SPNE incubated with NE (2 U/L) or not at different scan sections (n = 3 independent samples). k PA₆₉₅/PA₈₀₀ of SPNE under 10 consecutives incubated with NE or not. (n = 3 independent samples). l PA images of various concentrations of SPNE (4–10 μM) incubated with NE at 695 nm and 800 nm. m PA₆₉₅, PA₈₀₀ of SPNE in different concentrations (4–10 μM) incubated with NE (n = 10 independent samples). n PA₆₉₅/PA₈₀₀ of SPNE in different concentrations (4–10 μM) incubated with NE (n = 10 independent samples). Data are presented as the mean ± SD. Statistical significance was determined using a two-tailed Student’s t-test for pairwise comparisons and one-way ANOVA analysis of variance for multiple groups.

To test whether the PA signal ratio accurately detects NE concentration, different concentrations of NE were added to the Tris buffer containing SPNE (5 μM) and incubated at 37 °C for 30 min. Finally, the PA signals at 695 nm and 800 nm were collected by the MSOT system. We can see from Fig. 2c, d that the PA signal at 695 nm increased gradually with the gradual increase of NE concentration, while the signal at 800 nm remained basically unchanged. Then, we analyzed the correlation between the PA₆₉₅/PA₈₀₀ and NE concentration. The results showed that PA₆₉₅/PA₈₀₀ and NE concentration showed a good linear relationship between 0–2 U/L (R² = 0.985), and the detection of limit was 0.134 U/L obtained by 3ε/k (Fig. 2e). We further verified the selection specificity of SPNE, and found that only NE could significantly increase the PA₆₉₅/PA₈₀₀ under the presence of NE, while other reactive oxygen species and enzymes had little effect on the ratio signal, indicating that SPNE had strong PA anti-interference ability (Fig. 2f).

Subsequently, we observed the PA signal changes of SPNE in different sections (Fig. 2g). A 200 μL centrifuge tube was filled with the Tris buffer containing SPNE (5 μM) with or without NE (2 U/L) and incubated for 30 min at 37 °C. The cross-sectional area of the lower end of the centrifuge tube is not consistent due to its different position. We placed the centrifuge tube in the MSOT system and acquired PA images of five cross-section positions at 695 and 800 nm with a spacing of 2 mm. The results showed that before and after the response of SPNE to NE, the PA signals at 695 nm and 800 nm in different sections fluctuated (Fig. 2h, i), but the PA₆₉₅/PA₈₀₀ tended to be stable (Fig. 2j). At the same time, we found that each signal of SPNE (PA₆₉₅, PA₈₀₀, PA₆₉₅/PA₈₀₀) remained stable through 10 consecutive scans of PA signals (Fig. 2k, Supplementary Fig. 8), indicating that SPNE would not change under at least 10 scans. Finally, we observed the changes of PA signals after the incubation of SPNE and NE by changing the probe concentration. The results showed that with the increase of probe concentration, the PA signal gradually increased at 695 and 800 nm (Fig. 2l, m), but the PA₆₉₅/PA₈₀₀ tended to be stable (Fig. 2n). It shows that the PA ratio of SPNE is not affected by its own concentration, and is only related to NE concentration.

Given that SPNE exhibits good optical response performance to NE in solution, we observed the response of SPNE to neutrophils under different conditions (Supplementary Fig. 9a). Neutrophils were extracted from the blood of mice, and neutrophils NE levels were regulated with methoxy-methionine-leucyl-phenylalanine (fMLP) and cevestamat. Confocal fluorescence microscopy (Supplementary Fig. 9b) showed that activated neutrophils showed the strongest fluorescence in the SPNE channel, while inhibited neutrophils showed little fluorescence. Quantitative analysis showed that compared with the neutrophil group, the average fluorescence intensity in the neutrophil + activator group increased by 2.7-fold, while that in the neutrophil + inhibitor group decreased by 8.5-fold (Supplementary Fig. 9c).

The capability of SPNE to detect NE in vitro was then examined in NE-positive neutrophils and NE-negative cells, including Mouse breast cancer cells (4T1), mouse melanoma cells (B16), and human monocytes (THP-1). Flow cytometry and quantification results (Supplementary Fig. 9d and e) showed that the fluorescence intensity in 4T1, B16, and THP-1 cells remained low after SPNE incubation due to the lack of intracellular NE, whereas activated neutrophils exhibited the strongest fluorescence intensity. Neutrophils treated with the inhibitor displayed decreased fluorescence intensity, but this remained higher than that of other cell types, attributable to residual NE expression. Importantly, SPNE displayed low cytotoxicity towards neutrophils (Supplementary Fig. 10). Overall, SPNE sensitively detects intracellular NE expression and distinguishes neutrophils activity.

Quantification of neutrophil elastase (NE) content in the aorta by photoacoustic (PA) signal and evaluation of the relationship between photoacoustic (PA) ratio signals and plaque vulnerability

In order to achieve in vivo PA imaging to quantify NE content in plaque and predict the risk of plaque vulnerability. We intend to proceed through the following three steps.

In the initial step, in vivo PA signals were correlated with isolated aortic plaques (Fig. 3a). We utilized ApoE^-/- male mice subjected to a high-fat diet (HFD) for 24 weeks (The feed formula is shown in Table 1), leading to the development of aortic plaques of varying severity. Before PA imaging, four mark points were made on the shoulder and thigh regions of each mouse to facilitate spatial localization. Then, PA images capturing these marks and the entire aorta were obtained through regional tomography. During imaging, strong PA signals emerged at specific locations within particular cross-sections, serving as identifiable marks. The signal intensities significantly surpassed background noise, enabling straightforward localization of the marks.

Fig. 3: Quantification of neutrophil elastase (NE) content in the aorta by photoacoustic (PA) signal and evaluation of the relationship between photoacoustic (PA) ratio signals and plaque relative vulnerability.

a Scheme for matching PA signals with ex vivo plaques. b Physical images of different segments of aortic plaque and corresponding PA images at 695 nm and 800 nm. c Establishing the correlation curves of PA ratio signal (PA₆₉₅/PA₈₀₀) with NE concentration and plaque vulnerability index. d Quantification of PA ratios for four aortic segments (n = 4 independent photoacoustic images). e The NE content of four aortic segments was determined by neutrophil elastase Elisa kit (n = 3 independent samples). f The calibration curve between in vivo PA signal with ex vivo NE concentration, calculated from d and e. g Plaque relative vulnerability characteristics of the 3 aortic segments with α-SMA (smooth muscle cells), F4/80 (macrophages). H&E staining (necrotic cores) and Sirius red staining (collagen fibers). The solid line represents the plaque. The dashed line represents the necrotic core. h Quantitative calculation of macrophages, smooth muscle cells, necrotic cores and collagen fibers proportion in the plaques. i Vulnerability indices of different segments. j Heat map of PA₆₉₅/PA₈₀₀, proportion of necrotic cores, collagen fibers, macrophages, and smooth muscle cells, plaques within NE content of different segments. k The calibration curve between in vivo PA signal with plaque vulnerability index, calculated from d and i (n = 4 independent photoacoustic images). Data are presented as the mean ± SD. Statistical significance was determined using a two-tailed Student’s t-test for pairwise comparisons and a one-way ANOVA analysis of variance for multiple groups.

Table 1 The formula of high-fat feed

Subsequently, we identified the PA images corresponding to each mark and recorded their coordinates as X₀. The mice were then dissected to expose the aorta and marks. Target plaques (plaques 1, 2, and 3) were identified on the aorta, and the actual distances between each target plaque and marks were measured, recorded as D₁, D₂, and D₃, respectively. These measured distances were translated into image coordinates, resulting in the corresponding image positions X₁, X₂, and X₃ for the target plaques, thereby aligning the physical plaques with the in vivo PA images.

Through this matching process, PA data from four distinct fragments were successfully obtained. Comparison of the PA₆₉₅/PA₈₀₀ ratios across these fragments revealed that fragment 1 had a higher PA ratio than fragments 2 and 3, while fragment 4 exhibited the lowest ratio (Fig. 3d). Additionally, ex vivo fluorescence images of the aorta were promptly acquired using the IVIS imaging system (Supplementary Fig. 11a). These fluorescence images displayed four aortic segments with differing intensities: segment 1 showed higher fluorescence signal compared to segments 2 and 3, and segment 4 demonstrated the lowest fluorescence intensity (Supplementary Fig. 11b). This trend was consistent with the in vivo PA₆₉₅/PA₈₀₀ signal variations.

In the second step, we established a quantitative relationship between in vivo PA signals and neutrophil elastase (NE) content within the plaques (Fig. 3b). After aligning the in vivo PA signals with the corresponding ex vivo plaques, we harvested these plaques and quantified their NE content using a neutrophil elastase ELISA kit. The results revealed that NE content in segment 1 was significantly higher than those in segments 2 and 3, while segment 4 exhibited the lowest NE content, correlating with the absence of a prominent plaque (Fig. 3e). Furthermore, we analyzed the relationship between the PA₆₉₅/PA₈₀₀ and NE content in the plaques (Fig. 3f). Remarkably, a strong linear correlation was observed (R² = 0.986), indicating that the PA₆₉₅/PA₈₀₀ ratio reliably mirrors NE content within the plaques.

In the third step, we established a mathematical relationship between PA information and plaque relative vulnerability (Fig. 3c). Currently, most studies on plaque vulnerability are relative, with few efforts dedicated to quantifying plaque vulnerability. In animal studies, researchers often compare other vulnerability-related indicators, such as necrotic core size, lipid content, and foam cell presence, to explore plaque vulnerability³⁸. However, there is a lack of systematic integration among those indicators, and relying on a single indicator provides a one-sided view for linking plaque vulnerability to PA signals. Plaque vulnerability is a complex process influenced by multiple factors, making it difficult for a single indicator to comprehensively reflect the plaque’s true state. Therefore, to assess plaque vulnerability more accurately, there is an urgent need to introduce a standardized index that can generalize and integrate multiple indicators. We attempted to introduce a mathematical equation for the vulnerability formulation. This equation fully reflects the four most representative components of the vulnerable plaque (macrophages, smooth muscle cells, necrotic cores, collagen fibers)^39,40,41,42.

$${{{\rm{Vulnerability}}}}\; {{{\rm{index}}}} \, \left({VI}\right)=\frac{{{{\rm{Macrophages}}}}+{{{\rm{Necrotic}}}}\; {{{\rm{cores}}}}}{{{{\rm{Smooth}}}}\; {{{\rm{muscle}}}}\; {{{\rm{cells}}}}+{{{\rm{Collagen}}}}\; {{{\rm{fibers}}}}}$$

(1)

The index considers multiple factors affecting plaque stability and provides a quantitative criterion for relative vulnerability. The higher the vulnerability index, the more obvious the plaque rupture tendency. This formula is still used today as a key reference for assessing plaque vulnerability.

Subsequently, we conducted vulnerability components analysis on different plaque fragments (Fig. 3g). Macrophages play a central role in the inflammatory response; they originate from monocytes and are recruited in large numbers to atherosclerotic plaques. These cells release various inflammatory mediators and eventually transform into foam cells, thereby exacerbating plaque vulnerability⁴³. In our experiment, cells stained with F4/80 were identified as macrophages³⁹. Immunofluorescence quantitative analysis revealed that fragment 1 had significantly higher macrophages content compared to fragments 2 and 3 (Fig. 3h).

Smooth muscle cells are crucial for maintaining the integrity of the fibrous cap, as well as for the repair and remodeling of plaques⁴⁴. Their quantity and functional status directly influence plaque structural stability. We identified smooth muscle cells using α-smooth muscle actin (α-SMA) staining³⁹. Immunofluorescence quantitative analysis demonstrated that fragment 1 had lower smooth muscle cell content than fragments 2 and 3 (Fig. 3h).

The necrotic core comprises numerous dead endothelial cells, macrophages, debris from smooth muscle cells, lipid granules, and cholesterol crystals⁴⁵. The accumulation of these components can promote secondary cellular necrosis, thereby increasing plaque vulnerability and the risk of thrombotic vascular events⁴⁴. Hematoxylin and eosin (H&E) staining was employed to visualize extracellular vacuoles and lacunae characteristic of necrotic cores³⁹. The analysis indicated that segments 1 and 3 had significantly higher necrotic core content than segment 2 (Fig. 3h).

Collagen fibers contribute to the mechanical stability of the plaque and are involved in its repair and remodeling processes. A reduction in collagen fiber content and quality may elevate the risk of plaque rupture. In this study, collagen fibers were identified using Sirius red staining³⁹. The results showed that fragment 3 had significantly higher collagen fiber content than fragment 2, and both fragments 2 and 3 exhibited higher collagen levels than fragment 1 (Fig. 3h).

Next, we counted the proportion of these four components across different segments (Table S2), weighted their contributions, and calculated a vulnerability index. The results demonstrated that the vulnerability index of fragment 1 was significantly higher than those of fragments 2 and 3 (Fig. 3i). Heatmap analysis of the relationship between SPNE signals and plaque vulnerability components revealed that the PA₆₉₅/PA₈₀₀ was positively correlated with macrophage content and NE activity, and negatively correlated with smooth muscle cells and collagen fibers (Fig. 3j). Importantly, linear regression analysis confirmed a strong linear relationship between PA signal intensity and plaque relative vulnerability (R² = 0.941, P = 0.980, Fig. 3k), indicating that the in vivo PA₆₉₅/PA₈₀₀ of the SPNE nanoprobe was correlated with plaque relative vulnerability. In conclusion, the PA₆₉₅/PA₈₀₀ signal generated by SPNE had a strong correlation with the NE content in plaques and could be used for quantitative analysis of NE levels in vivo.

In vivo photoacoustic imaging of neutrophil elastase (NE) in carotid plaque

With the continuous development of atherosclerosis, it will cause a variety of inflammation-mediated complications, such as pneumonia, periodontitis, colitis, peritonitis, and so on^46,47,48. These complications will further increase the level of inflammation in the plaque through the circulatory system, thereby aggravating the tendency of plaque rupture⁴⁹. To construct a model of atherosclerotic disease with inflammation, we used lipopolysaccharide (LPS)-induced pneumonia in mice as a complication to observe the stability of plaques. At the same time, we further monitored the effect on NE content in mice with pneumonia and atherosclerosis after anti-inflammatory and lipid-lowering treatment.

The specific experimental procedures are shown in Fig. 4a, the left carotid artery (LCA) was isolated from the necks of 15 six-week-old ApoE^-/- male mice using a blunt instrument and subsequently ligated with braided polyester fiber suture. The right carotid artery (RCA) remained untreated. The mice were placed on a high-fat diet for eight weeks and then randomly divided into three equal groups: AS mice, AS mice with inflammation, and AS mice with inflammation and drug. For AS mice, no treatments were administered during the first 7 days, and only PBS was given daily from day 8 to 21. For AS mice with inflammation, mice were given LPS (2 mg/mL, 50 μL) by nasal feeding on days 1, 4, and 7, along with daily PBS administration from day 8 to 21. For AS mice with inflammation and drug, mice were given LPS (2 mg/mL, 50 μL) by nasal feeding on day 1, 4, and 7, while also receiving rosuvastatin (10 mg/kg/d) and dexamethasone (2 mg/kg/d) daily from day 8 to 21. All mice underwent PA imaging to record background signals prior to intravenous injection of SPNE (200 μM, 200 μL) on day 22. PA images were obtained from all mice at wavelengths of 695 nm and 800 nm post-injection at 15, 30, and 60 min. It is worth noting that the carotid artery stenosis model was a classic atherosclerosis model, and the right carotid artery (RCA) of each mouse was a healthy artery and used as a blank control⁵⁰.

Fig. 4: In vivo photoacoustic imaging of neutrophil elastase (NE) in carotid plaque.

a Scheme of the construction of the carotid plaque model. b The coronal plane of mice for PA imaging, the solid line indicates the scan position. c Organ distribution map displayed by the MSOT system during PA imaging of carotid plaques mice, and the arrow and dashed circles indicates the carotid artery. d in vivo PA image for carotid plaques mice. e PA images of each group of mice before and post injection of SPNE. The dashed line represents the carotid arteries, and the arrows refer to the left carotid artery (LCA) and right carotid artery (RCA), respectively. f Quantification of PA₆₉₅ of left carotid artery in each group of mice before and post-injection of SPNE (n = 5 mice per group). g Quantification of PA₈₀₀ of left carotid artery in each group of mice before and post injection of SPNE (n = 5 mice per group). h Quantification of PA₆₉₅/PA₈₀₀ of left carotid artery in each group of mice before and post injection of SPNE (n = 5 mice per group). i Cholesterol content in serum of AS mice and Normal mice (n = 3 mice per group). j H&E staining images of the lung for each group, arrows indicate inflammatory cell infiltration. Data are presented as the mean ± SD. Statistical significance was determined using a two-tailed Student’s t-test for pairwise comparisons, and one-way ANOVA analysis of variance for multiple groups. p values  >  0.05 were considered non-significant, while p values  <  0.05 were considered statistically significant. p values are indicated on the graphs.

During PA imaging of carotid plaques mice, we first selected the scanning range of mouse carotid artery region in the MSOT system (Fig. 4b), and then tomography was performed at a 0.3 mm interval. The distribution map in each organ of the mouse was obtained by the MSOT system (Fig. 4c), and compared with our PA signal image to confirm the exact location of the carotid artery in the PA images (Fig. 4d).

Subsequently, we acquired carotid PA images at different time points before and after injection of SPNE in each group of mice (Fig. 4e), and plotted the time curves of PA signals at 695 nm and 800 nm in the left carotid artery (LCA) of each group of mice (Fig. 4f, g). The variation of the PA signal at 695 nm was correlated with the NE content within the plaque as well as the probe concentration, while the variation at 800 nm was only related to the probe concentration. From the time curve, we can observe that the single detection method of PA₆₉₅ cannot accurately reflect the content of NE in the plaque due to the fluctuation of probe concentration in the plaque.

Then, we further analyzed the PA₆₉₅/PA₈₀₀ signal in each group (Fig. 4h). After 60 min of imaging, the PA₆₉₅/PA₈₀₀ of the AS mice stabilized at a low level, while the PA₆₉₅/PA₈₀₀ of AS mice with inflammation was much higher than that of AS mice, reflecting the rapid increase of NE content in the plaque after stimulation of inflammation. In addition, the PA₆₉₅/PA₈₀₀ was reduced after drug intervention in AS mice with inflammation and drug. The results showed that the combination drug therapy had a significant inhibitory effect on intraplaque NE. Notably, the right carotid artery (RCA) had the lowest signal (PA₆₉₅, PA₈₀₀) because of the absence of plaque (Supplementary Fig. 12). In conclusion, SPNE enables ratiometric PA₆₉₅/PA₈₀₀ imaging of neutrophil elastase in inflammation plaques, effectively distinguishing NE activity among AS mice, AS mice with enhanced inflammation, and AS mice treated with anti-inflammatory drugs.

In order to verify the successful construction of the carotid artery model, serum cholesterol levels were measured in normal and AS mice. The results showed that serum cholesterol content of AS mice was significantly higher than that of normal mice (Fig. 4i). Moreover, it can also be observed that after 8 weeks of high-fat feeding, AS mice showed obvious plaque formation in the left carotid artery, but there was no obvious plaque formation in the right carotid artery (Supplementary Fig. 13). These results confirmed the successful establishment of carotid artery plaque model. Lastly, H&E staining of the lung in each group of mice revealed that lung inflammation increased under LPS stimulation, while combined drug treatment alleviated this inflammation (Fig. 4j).

Analysis of the vulnerability index of carotid plaque

In order to obtain the pathological characteristics of carotid plaques in each group of mice, we euthanized the mice in each group immediately after PA imaging. The carotid arteries of mice in each group were collected, and frozen sections were prepared. Oil red O staining and immunofluorescence staining were subsequently performed (Fig. 5a).

Fig. 5: Analysis of the vulnerability index of carotid plaque.

a Immunofluorescent staining (Ly6G), and Oil red O staining for each group of mice. Dashed line represents the plaque. b Ly6G immunofluorescence quantitative analysis of each group of mice (n = 3 mice per group). c Oil red O staining quantitative analysis of each group of mice (n = 3 mice per group). d Immunofluorescent staining (F4/80 and α-SMA), H&E staining, and Sirius red staining for each group of mice. Dashed line represents the plaque. e–h Quantitative analysis of macrophages, smooth muscle cells, necrotic cores, and collagen fibers proportion for each group (n = 3 mice per group). i Vulnerability index of each group via the vulnerability formula (n = 3 mice per group). Data are presented as the mean ± SD. Statistical significance was determined using a two-tailed Student’s t-test for pairwise comparisons and one-way ANOVA analysis of variance for multiple groups. p values  >  0.05 were considered non-significant, while p values  <  0.05 were considered statistically significant. p values are indicated on the graphs.

We performed Oil red O staining and Ly6G immunofluorescence staining analysis of carotid arteries in each group. The result showed that Ly6G fluorescence signal intensity and lipid content in AS mice with inflammation were significantly higher than those in AS mice (Fig. 5b, c), indicating the ability of lipid uptake and the degree of neutrophils infiltration in plaques increased significantly after stimulation of inflammation. In addition, compared with the AS mice with inflammation, the lipid level and the degree of neutrophils infiltration in the plaque were greatly relieved after the combined anti-inflammatory and lipid-lowering treatment in the AS mice with inflammation and the drug. These phenomena are consistent with the results of lipid and Ly6G fluorescence quantification.

We further analyzed the vulnerable components of the plaque and quantified the proportion of each vulnerable component in the plaque (Fig. 5d). For F4/80 staining, it is clearly observed that the immunofluorescence intensity of AS mice with inflammation was much higher than that of AS mice. This phenomenon was completely reversed after combined treatment. Compared with AS mice with inflammation, the macrophage infiltration in AS mice with inflammation and drug was significantly decreased, and the quantitative results were similar to immunofluorescence images (Fig. 5e).

For α-SMA staining, the fluorescence was significantly lower in AS mice with inflammation than in AS mice, which means the proportion of smooth muscle cells decreased significantly after stimulation of inflammation. Smooth muscle cells are one of the important components to stabilize the plaque structure, and a reduction in the proportion of smooth muscle cells means a decrease in plaque stability⁵¹. Compared with AS mice with inflammation, the proportion of smooth muscle cells in the plaque of AS mice with inflammation and drug treatment was restored after drug treatment, but there was no significant difference (Fig. 5f).

H&E staining showed obvious plaque formation in the carotid arteries of the three groups of mice. Quantitative calculation of necrotic core revealed that the proportion of necrotic core in AS mice with inflammation was significantly higher than that in AS mice. This implied that LPS was able to promote the formation of necrotic cores. After AS mice with inflammation continued to be treated with drugs, we found that the necrotic core of AS mice with inflammation and drugs was significantly reduced (Fig. 5g).

The results of Sirius red staining and quantitative analysis showed that the content of collagen fibers in plaques of AS mice with inflammation was significantly lower than that of AS mice, indicating that inflammatory stimulation led to the decomposition of collagen fibers and the reduction of plaque stability. Further comparison showed that there was no significant difference in collagen fiber content between AS mice with inflammation and AS mice with inflammation and drug, suggesting that the anti-inflammatory and lipid-lowering treatment regimen had no effect on collagen fibers recovery (Fig. 5h).

Then, we summarized the proportions of the above four components (Table S3) in the plaque and calculated the vulnerability index of mice in each group (Fig. 5i), the results showed that the vulnerability index of AS mice with inflammation was much higher than that of AS mice. This suggests that inflammation indeed promotes the development of atherosclerotic plaques. The comparison of the vulnerability index between the AS mice with inflammation and the AS mice with inflammation and drug suggested that the risk of vulnerability caused by inflammation could be alleviated by drug treatment.

Lastly, H&E staining of major organs revealed that no significant damage was noted in other organs (e.g., heart, liver, spleen, and kidney) two hours post-injection of SPNE, which suggested that SPNE exhibited good biosafety (Supplementary Fig. 14).

In vivo photoacoustic imaging of neutrophil elastase (NE) in intimal hyperplasia

Coronary atherosclerotic heart disease is a cardiovascular disease that seriously endangers human health^52,53. Coronary revascularization can be achieved well by percutaneous coronary intervention⁵⁴. It can improve the quality of life of patients and significantly reduce the mortality of patients. However, restenosis caused by vascular intimal injury is a difficult problem of interventional therapy, and it is also an important reason affecting the prognosis of patients and the recurrence of ischemic heart disease⁵⁵. Intimal hyperplasia is a complication associated with various cardiovascular and cerebrovascular surgeries, including arterial bypass grafting, angioplasty, stent implantation, and endarterectomy. With the development of the disease, it causes different degrees of vascular stenosis and may endanger the normal physiological function of the target organ⁵⁶.

During the process of intimal hyperplasia, infiltration of inflammatory cells leads to increased expression of hydrolytic proteases in neutrophils⁵⁷. Proteolytic proteases, mainly NE, promote the remodeling of extracellular matrix, promote the migration of smooth muscle cells to the intima, marking the transition of smooth muscle cells from a contractile phenotype to a synthetic phenotype, and eventually lead to the proliferation of smooth muscle cells in the intima^57,58. Therefore, it is important feasibility to develop probes to image NE at the lesion site to observe the progression of intimal hyperplasia.

To establish a model of intimal hyperplasia⁵⁹. Six-week-old C57BL/6 mice underwent left carotid artery ligation, while the right carotid artery remained untreated (Fig. 6a). This model of intimal hyperplasia was successfully established 12 weeks post-surgery. Acknowledging that varying levels of inflammation can accelerate intimal hyperplasia, we adopted an LPS induction model. Specifically, intimal hyperplasia mice were randomly assigned to two groups: Intimal hyperplasia mice and intimal hyperplasia mice with inflammation. For the intimal hyperplasia mice with inflammation, mice were given LPS (2 mg/mL, 50 μL) by nasal feeding on days 1, 4, and 7, while the intimal hyperplasia mice received no treatment. All mice underwent a pre-photoacoustic imaging session to obtain background signals before intravenous injection of SPNE (200 μM, 200 μL) on day 8. Subsequently, PA images were captured at 695 nm and 800 nm at 5, 15, 30, 45, and 60 min after SPNE injection.

Fig. 6: In vivo photoacoustic imaging of neutrophil elastase (NE) in intimal hyperplasia.

a Scheme of the construction of the intimal hyperplasia model. b Organ distribution map displayed by the MSOT system during PA imaging of intimal hyperplasia mice, and the yellow arrow and dashed circles indicate the carotid artery. c PA images of two groups of mice before and post-injection of SPNE. Dashed line represents the carotid arteries, and the arrows refer to the left carotid artery (LCA) and right carotid artery (RCA), respectively. d PA₆₉₅ and PA₈₀₀ of the left carotid artery in intimal hyperplasia mice before and post-injection of SPNE (n = 4 mice per group). e PA₆₉₅ and PA₈₀₀ of the left carotid artery in intimal hyperplasia mice with inflammation before and post-injection of SPNE (n = 4 mice per group). f Quantification of PA₆₉₅/PA₈₀₀ of the left carotid artery in each group before and post-injection of SPNE (n = 4 mice per group). g Immunofluorescent staining for each group with Ly6G and F4/80. Dashed lines represent the plaque. Data are presented as the mean ± SD.

During PA imaging of intimal hyperplasia mice, the distribution map in each organ of the mouse was obtained by the MSOT system, and compared with our PA signal image to confirm the exact location of the intimal hyperplasia in the PA images (Fig. 6b).

Subsequently, the PA intensity of the carotid arteries of the two groups at different time points was counted and compared. The results (Fig. 6c) showed that the PA signals at 695 nm and 800 nm in the left carotid artery of intimal hyperplasia mice and intimal hyperplasia mice with inflammation gradually increased over time, while the intensity of the right carotid artery was almost the same as that before probe injection (Supplementary Fig. 15). This suggested that SPNE could rapidly enrich to the lesion site and respond to NE, while the right carotid artery without intimal hyperplasia had no PA signal change.

Then we compared and analyzed the trend of the PA signal of the left carotid artery between the two groups. The PA signal increased more slowly at 695 nm or 800 nm in intimal hyperplasia mice. In contrast, PA signals of intimal hyperplasia mice with inflammation rapidly reached a plateau within 15 min (Fig. 6d, e). This phenomenon suggested that inflammation induced increased uptake of inflammatory cells in intimal hyperplasia. At the same time, the up-regulation of the NE activity at the lesion site led to a faster rise of the signal at 695 nm. Next, we focused on analyzing the PA₆₉₅/PA₈₀₀ (Fig. 6f). After 60 min of PA imaging, we found that compared with the intimal hyperplasia mice, the PA₆₉₅/PA₈₀₀ of the intimal hyperplasia mice with inflammation was greater, which implied the contribution of inflammation to the increase of the NE content.

At the end of the imaging, the mice were euthanized, and the carotid arteries of each group were collected and made into frozen sections for further H&E staining and immunofluorescence staining. immunofluorescence analysis showed that the degree of inflammatory cell infiltration (macrophages and neutrophils) in intimal hyperplasia mice with inflammation was significantly higher than that in intimal hyperplasia mice (Fig. 6g), suggesting that induction of inflammation can induce strong recruitment of inflammatory cells to intimal hyperplasia. H&E staining indicated that the structure of the intima was dense in mice with intimal hyperplasia; however, the intima of mice with inflammation has a tendency to rupture in response to inflammatory stimulation (Supplementary Fig. 16). The trend of these pathological sections was consistent with the results of PA imaging, indicating that SPNE can effectively detect NE content in intimal hyperplasia.

Discussion

We have developed a neutrophil elastase (NE)-responsive ratiometric semiconductor polymer nanoprobe (SPNE) comprising two essential components: an NE-responsive hemicyanine molecule acting as a photoacoustic (PA) sensor and a benzotriazole-based fusion ring molecule providing PA reference. SPNE exhibits high sensitivity to NE, with a detection limit of 27.957 mU/L. Unlike single-signal PA probes, SPNE incorporates a unique internal reference, enabling the accurate quantification of NE content through the self-calibration of the ratiometric PA ratio signal (PA₆₉₅/PA₈₀₀). Consequently, SPNE demonstrates robust PA anti-interference capabilities, with its PA₆₉₅/PA₈₀₀ remaining unaffected by external factors such as reactive oxygen species, other enzymes, variations in probe concentration, excitation times, and different scanning cross-sections.

Atherosclerosis is a disease involving numerous complex factors and pathological processes^9,10,11. NE plays a critical role in the development of atherosclerosis by promoting inflammatory cell infiltration, accelerating the inflammatory response within plaques, mediating the degradation of the extracellular matrix and cytokines, and facilitating the rapid progression of atherosclerotic plaques^15,18,19. These actions increase the risk of plaque rupture, making NE a key marker of plaque relative vulnerability. Therefore, quantifying NE content within plaques is essential for understanding the pathological mechanisms of atherosclerosis.

In our study, we establish a correlation between in vivo PA signals and the locations of isolated plaques. NE levels in plaques are accurately quantified using an ELISA kit, enabling the in vivo quantification of NE content—marking an important advancement for precise disease diagnosis. Monitoring NE levels can aid in predicting the progression of atherosclerosis and the risk of cardiovascular events, providing early warning indicators for patients and informing proactive treatment plans. Based on NE levels, personalized drug intervention recommendations can be made, optimizing treatment strategies for precise therapeutic effects.

Additionally, we employ a plaque vulnerability index by weighting key vulnerable components—macrophages, smooth muscle cells, necrotic cores, and collagen fibers^39,40. This approach quantifies plaque relative vulnerability, transforming it from a qualitative concept into a quantitative indicator. Quantifying plaque relative vulnerability not only allows for more accurate prediction of the risk of cardiovascular events but also facilitates personalized medical interventions and preventive strategies to reduce the occurrence of such events.

Atherosclerosis is a chronic inflammatory disease, and inflammation plays a pivotal role in its onset and progression^9,11. The inflammatory response involves various factors and signaling pathways that mediate the initiation of atherosclerosis, plaque growth, local apoptosis, and endothelial neovascularization¹⁰. Additionally, atherosclerosis is often associated with inflammatory complications such as myocarditis, hepatitis, peritonitis, periodontitis, and colitis^46,47,48. This systemic inflammation, initiated by local inflammatory responses, heightens the risk of atherosclerotic plaque vulnerability. In our study, we demonstrate that SPNE can be used to monitor the dynamic changes of NE content in carotid plaque in real time. This suggests our tool provides an important window into inflammatory activity and lays the groundwork for future validation in more rupture-prone models or clinical settings. Moreover, epidemiological data indicates that environmental factors—such as lifestyle choices (e.g., smoking, diet, exercise) and physical environmental factors (e.g., air pollution, industrial activities)—significantly influence the development and progression of atherosclerosis⁶⁰. Leveraging the superior imaging capabilities of SPNE, future studies can explore the effects of different environmental factors on atherosclerotic diseases.

Furthermore, SPNE demonstrats the ability to respond to molecular-level changes within plaques during treatment, allowing for the monitoring of drug efficacy in atherosclerosis-related diseases. This capability facilitates the customization of personalized treatment plans for patients and guides the therapeutic process effectively.

Acute coronary syndromes are a leading cause of morbidity and mortality globally, with thrombotic events resulting from plaque rupture being the primary contributors^52,53. In the context of the carotid artery, plaque rupture can lead to thrombus formation, which may travel through the bloodstream to the middle cerebral artery, causing temporary or permanent occlusion and resulting in stroke—a life-threatening condition. Common clinical interventions include vascular recanalization through stent implantation and endarterectomy. However, these procedures can damage the vascular lining and induce further intimal hyperplasia, leading to restenosis of the blood vessels⁶¹. Restenosis not only results in postoperative recurrence but also causes secondary damage to target organs due to hypoxia and insufficient blood supply following vessel re-narrowing⁶². Thus, intimal hyperplasia is a critical factor in the progression of atherosclerosis. SPNE effectively visualizes changes in NE content associated with intimal hyperplasia, reflecting its severity. This capability is valuable for monitoring the incidence of postoperative complications and minimizing patient discomfort caused by secondary injuries. By providing real-time insights into intimal hyperplasia, SPNE enhances the monitoring of vascular health and the management of atherosclerosis treatment outcomes.

In this study, we develop a neutrophil elastase (NE)-activated semiconductor polymer nanoprobe (SPNE) capable of non-invasively imaging NE in atherosclerotic plaques. The nanoprobe exhibits high sensitivity, strong selectivity, and robust anti-interference capabilities. Notably, we successfully correlate in vivo photoacoustic signals with isolated plaques. We achieve the in vivo quantification of NE content within plaques using the photoacoustic ratio signal (PA₆₉₅/PA₈₀₀). By introducing a vulnerability index formula, we effectively digitize the assessment of plaque relative vulnerability. Utilizing SPNE, we image the impact of pneumonia on carotid atherosclerotic plaques. Inflammatory stimulation contributes to a significant increase in NE content. Conversely, combined anti-inflammatory and lipid-lowering therapy in inflammation mice significantly reduce intra-plaque NE levels. These results demonstrate the potential of SPNE for risk stratification and predicting therapeutic outcomes in atherosclerosis models. Furthermore, SPNE effectively visualizes dynamic changes in NE levels in intimal hyperplasia. Overall, our findings establish SPNE as a powerful tool for the non-invasive quantification of NE in atherosclerotic plaques.

Methods

Materials and characterization

All the chemical materials were purchased from commercial suppliers. m-nitrophenol, Tin (II) chloride were purchased from Aladdin. Phosphorus (V) oxychloride was purchased from Anneji (Shanghai) Pharmaceutical Chemical Co., Ltd. Compound Y11, DFIC were purchased from HWRK CHEM. 2,3,3-trimethylindolenine, Iodoethane, and N,N-Diisopropylethylamine (DIPEA) were purchased from Adamas-beta. 2-(7-Azabenzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate (HATU) was purchased from Bide Pharmatech Ltd. the silica gel used for column chromatography, and 2,4-Dimethylpyrrole were purchased from Leyan.com (China). N-acetyl-Ala-Ala-Pro-Val-COOH was purchased from Genscript. 1,2-dimyristoyl-sn-glycero-3-phospho-ethanolamine-N-[methoxy(poly(ethylene glycol))] (DSPE-PEG₂₀₀₀) was purchased from Shanghai ToYongBio Tech.Inc. Lipopolysaccharide (LPS, Cat. No. L2630), poly (styrene-maleic anhydride) (PSMA, Cat. No. 188050), formyl-methionyl-leucyl-phenylalanine (fMLP, Cat. No. F3506), sivelestat (Cat. No. S7198), dexamethasone (Cat. No. D4902), rosuvastatin (Cat. No. Y0001719), and neutrophil elastase (NE, Cat. No. E8140) were purchased from Sigma-Aldrich.

Transmission electron microscope images were obtained via a JEM-2100F instrument (JEOL). The absorbance was recorded by UV-vis absorption spectrometry (UV-3600, Shimadzu). Fluorescence spectra were recorded by an Edinburgh Instruments F-S5 fluorescence spectrometer. 1H NMR was performed using a Bruker DRX-400 spectrometer. Mass spectrometry was conducted using a matrix-assisted laser desorption/ionization time-of-flight mass spectrometer (ultrafleXtreme). Flow cytometry was measured by the BD FACS Verse system. Immunofluorescence images were measured on a confocal laser scanning microscope (Nikon, Japan). H&E images were obtained by a digital slice scanning system (Pannoramic MIDI). Fluorescent images were obtained by the IVIS spectral imaging system; Imaging parameters: Fluorescence; Excitation wavelength: 640 nm Filter: Cy5.5 channel (690 to 770 nm); Software: Living Image 4.0 software. Photoacoustic images were obtained by MSOT inVision 256-TF (iThera Medical GmbH, Munich, Germany). Wavelengths range: 680 nm to 980 nm; pulse repetition rate: 50.0 kHz ~ 6.5 MHz; Speed of sound: 9; Scan area: 25 × 25 mm²; Resolution preset: 75 μm; Step size: 0.3 mm; Water temperature: 34.0 °C.

Design, synthesis, and properties of an activatable ratiometric photoacoustic probe for detecting neutrophil elastase (NE)

Synthesis of hemicyanine fluorophore (Hcy)

The hemicyanine fluorophore (Hcy) was synthesized following the previous literatures⁶³. The detailed synthesis steps are as follows: 10 mL of dry DMF was added in a flask cooled at 0 °C. Then, 9.5 mL of phosphorus oxychloride was added slowly into the flask, stirred for 30 min at 0 °C. 2.5 mL of cyclohexanone was added into the mixed solution. Then the solution was refluxed for 3 h at 90 °C under N₂ atmosphere. Upon cooling, the solution was poured onto 300 mL of ice water, and precipitation was formed. After vacuum filtration, washing, and drying, the yellow solid (compound 1, yield = 83%) was obtained.

Compound 1 (1.036 g, 6.0 mmol), acetic anhydride (Ac₂O, 30 mL), and 1-ethyl, 2,3,3-trimethyl-3H-indolium iodide (3.62 g, 12 mmol) were added to a three-necked flask. The solution was heated to reflux at 90 °C for 1 hour under N₂ atmosphere. After cooling to room temperature, the product solution was poured into a saturated sodium carbonate aqueous solution to remove Ac₂O. Subsequently, the product was purified by silica gel chromatography with CH₂Cl₂/MeOH (volume/volume = 50/3) as the eluent to obtain a blue-green solid (Compound 2, yield = 65%).

K₂CO₃ (414 mg, 3 mmol), and m-nitrophenol (327 mg, 3 mmol) were dissolved in 20 mL of CH₃CN in a three-necked flask and were stirred at room temperature for 10 min under N₂ atmosphere. Then compound 2 (1.2 g, 2 mmol) was put into the flask, and the reaction mixture was stirred at room temperature for 5 h. The solvent was removed under reduced pressure, the solid was dissolved in CH₂Cl₂, washed with water three times, and dried with Na₂SO₄. CH₂Cl₂ was removed by reduced pressure, then the solid product and SnCl₂ (2.3 g, 10 mmol) were dispersed in 25 mL of MeOH, 3 mL of HCl was added, and the system was stirred at 70 °C under N₂ atmosphere overnight. After reaction, pH of the solution was adjusted to 7 with NaOH (100 mM) aqueous solution, and precipitation was formed in this process. After filtration and washing, the solution became clear. The collected filtrate was dried over Na₂SO₄. CH₂Cl₂ was removed by reduced pressure, the crude product was purified by silica gel chromatography with CH₂Cl₂/MeOH (volume/volume = 25/1) to obtain a green solid (Hcy, yield = 83%).

Synthesis of NE-unresponsive molecule

NE-unresponsive molecule was synthesized following the previous literatures⁶⁴.

Synthesis of neutrophil elastase response probe (NERP)

Peptide N-acetyl-Ala-Ala-Pro-Val-COOH (Ac-AAPV-COOH, 40 mg, 0.1 mmol), 2-(7-Azabenzotriazol-1-yl)-N, N, N’, N’-tetramethyluronium exafluorophosphate (HATU, 50 mg, 0.13 mmol), and N, N-diisopropylethylamine (DIPEA, 20 μL, 0.2 mmol) were dissolved in CH₂Cl₂ (10 mL) with stirring at 0 °C for 60 min. Hcy (5 mg, 0.012 mmol) is dissolved in CH₂Cl₂, then added to the reaction mixture above. The mixed solution was further stirred at room temperature for 48 h. Then the solution was extracted with CH₂Cl₂ (200 mL) and Saturated NaCl solution (200 mL). The collected organic phase was dried over Na₂SO₄, and the organic solvent was evaporated under reduced pressure at 43 °C. The product was purified by silica gel chromatography with CH₂Cl₂/MeOH (volume/volume = 50/3) as eluent to obtain neutrophil elastase response probe (NERP) as a blue solid (yield = 48%).

Synthesis of NE-activatable semiconductor polymer nanoprobes (SPNE)

For synthesis of NE-activatable semiconductor polymer nanoprobes (SPNE), DSPE-PEG₂₀₀₀ (2.5 mg), PSMA (2.5 mg), NERP (17 μg), and NE-unresponsive molecule (40 μg) were dissolved in a centrifuge tube containing of THF (1 mL), and then placed into an ultrasonic cleaning instrument and sonicated for 1 minute. The above solution was rapidly injected into a serum vial containing 9 mL of ultrapure water under continuous sonication, and the mixture was further sonicated for 10 min. Na₂CO₃ solution (50 mg/mL, 100 μL) was added into the mixture during the ultrasound procedure. Excess THF was removed by rotary evaporation under reduced pressure at 47 °C. Finally, the resulting SPNE was purified using ultrafiltration (100,000 MWCO, 1600 × g, 5 min) and concentrated to 200 μL.

Measurement of NERP response to NE

A stock solution of NERP was prepared at a concentration of 1 mM in methanol. Subsequently, 1 μL of stock solution was added to the Tris buffer (391 μL, 50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35), and then 8 μL of neutrophil elastase (NE) stock solution (100 U/L) was added to this solution. The mixture was then incubated at 37 °C for 30 min. The absorption and fluorescence spectra were recorded for NERP with NE or without NE.

Measurement of SPNE response to NE

A stock solution of SPNE was prepared at a concentration of 1 mM in deionized water (The concentration of SPNE was calibrated with the concentration of NERP). Subsequently, 1 μL stock solution was added to the Tris buffer (195 μL, 50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35), and then 4 μL neutrophil elastase (NE) stock solution (100 U/L) was added to this solution. The mixture was then incubated for 30 min at 37 °C. The fluorescence spectra were recorded for SPNE with NE or not.

The absorption ratio signal for assessing the sensitivity of SPNE to NE

A stock solution of SPNE was prepared at a concentration of 1 mM in deionized water. Then 1 μL of stock solution was added to the Tris buffer (50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35). Subsequently, different volumes of NE stock solution (10 U/L) were added into this solution. The final volume of the solution was 200 μL, the final SPNE concentration was 5 μM, and NE concentration ranged from 0 to 2 U/L. Subsequently, this mixture was incubated at 37 °C for 30 min. Absorption spectra of these solutions were then recorded, and the absorption ratio (Abs₆₉₅/Abs₈₀₀) were calculated according to Abs₆₉₅, Abs₈₀₀. Finally, a linear relationship was established based on the Abs₆₉₅/Abs₈₀₀ and NE concentration.

Enzyme kinetics studies

Various concentrations of SPNE (2.3, 4.6, 6.9, 9.2, 11.5, 13.8 μg/mL, the corresponding NERP concentration was 1.25, 2.5, 3.75, 5, 6.25, 7.5 μM) were incubated with NE (5 mU/L) at 37 °C for 30 min in the Tris buffer (200 μL, 50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35). Subsequently, the solutions were taken for absorption analysis. The initial reaction velocity (mg·mL^-1s^-1) was calculated, plotted against the concentration of SPNE, and fitted to a Michaelis-Menten curve. The kinetic parameters were calculated by using Michaelis-Menten equation²⁶:

$$V=\frac{{{{{\rm{V}}}}}_{\max }\times [{{{\rm{S}}}}]}{{{{{\rm{K}}}}}_{{{{\rm{M}}}}}+[{{{\rm{S}}}}]}$$

(2)

Where V is initial reaction velocity, [S] is substrate concentration (SPNE), V_max represents the reaction rate at which the enzyme is saturated with the substrate, and the K_M value, called Michaelis constant, is the substrate concentration at which the enzymatic reaction rate V is half the maximum enzymatic reaction rate value.

Response of SPNE to different NE concentrations

1 μL SPNE stock solution (1 mM) was added to the Tris buffer (50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35). Subsequently, different volumes of NE stock solution (100 U/L) were added to the above mixture. The final volume of the solution was 400 μL, the final SPNE concentration was 2.5 μM and NE concentration ranged from 0 to 15 U/L. Then these solutions were incubated at 37 °C for 30 min. The absorption spectra of these solutions were subsequently recorded.

Response of SPNE to different times

1 μL SPNE stock solution (1 mM) was added to the Tris buffer (391 μL, 50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35). Subsequently, 8 μL NE stock solution (100 U/L) was added to the above mixture. This solution was incubated at 37 °C for different times (0 ~ 70 min, two-minute intervals). Absorption spectra of these solutions were subsequently recorded.

The absorbance ratio signal for detecting the selectivity of SPNE

Various species were incubated with SPNE (5 μM) in Tris buffer (200 μL, 50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35) for 30 min at 37 °C, as following: ¹O₂ (100 μM), H₂O₂ (100 μM), HClO (100 μM), O₂^•- (100 μM), •OH (100 μM), ONOO^- (100 μM), granzyme B (GB, 50 ng/mL), NADH quinone oxidoreductase1 (NQO1, 50 ng/mL), caspase 3 (50 ng/mL), alkaline phosphatase (ALP, 50 ng/mL), β-galactosidase (50 ng/mL), glutamyltransferase (GGT, 50 ng/mL), matrix metalloproteinase 2 (MMP-2, 50 ng/mL), phosphodiesterase (PDEs, 50 ng/mL), neutrophil elastase (NE, 2 U/L). Then, the absorption of each sample was measured.

Ratiometric photoacoustic imaging for neutrophil elastase (NE) detect in solution

Measurement of SPNE Responsiveness to NE

A stock solution of SPNE was prepared at a concentration of 1 mM in deionized water. Subsequently, 1 μL of stock solution was added to the Tris buffer (195 μL, 50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35), and then 4 μL of neutrophil elastase (NE) stock solution (100 U/L) was added to this solution. The mixture was then incubated at 37 °C for 30 min. The PA spectra were recorded for SPNE with NE or without NE.

The PA ratio signal for detecting the sensitivity of SPNE to NE

A stock solution of SPNE was prepared at a concentration of 1 mM in deionized water. Then 1 μL of stock solution was added to the Tris buffer (50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35). Subsequently, different volumes of NE stock solution (10 U/L) were added to this solution. The final volume of the solution was 200 μL, the final SPNE concentration of 5 μM, and NE concentrations ranging from 0 to 2 U/L. This solution was placed in a 200 μL centrifuge tube and incubated at 37 °C for 30 min. PA spectra of these solutions were subsequently recorded by the MSOT system. The PA ratio (PA₆₉₅/PA₈₀₀) was calculated according to PA₆₉₅, PA₈₀₀. Finally, a linear relationship between the PA₆₉₅/PA₈₀₀ and NE concentration was established.

The PA ratio signal for detecting the selectivity of SPNE

Various species were incubated with SPNE (5 μM) in the Tris buffer (200 μL, 50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35) for 30 min at 37 °C, as following: ¹O₂ (100 μM), H₂O₂ (100 μM), HClO (100 μM), O₂^•- (100 μM), •OH (100 μM), ONOO^- (100 μM), granzyme B (GB, 50 ng/mL), NADH quinone oxidoreductase1 (NQO1, 50 ng/mL), caspase 3 (50 ng/mL), alkaline phosphatase (ALP, 50 ng/mL), β-galactosidase (50 ng/mL), glutamyltransferase (GGT, 50 ng/mL), matrix metalloproteinase 2 (MMP-2, 50 ng/mL), phosphodiesterase (PDEs, 50 ng/mL), neutrophil elastase (NE, 2 U/L). Then, this solution was placed in a 200 μL centrifuge tube, and the PA₆₉₅/PA₈₀₀ of each sample was measured by the MSOT system.

PA anti-interference capability of SPNE under different scan sections

A 200 μL centrifuge tube was filled with the Tris buffer (50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35) containing SPNE (5 μM) with or without NE (2 U/L) for 30 min at 37 °C. 5 cross sections were selected at a spacing of 2 mm, the 5 cross sections of PA₆₉₅ and PA₈₀₀ were recorded by using the MSOT system. Finally, PA₆₉₅/PA₈₀₀ of SPNE with NE or not was calculated according to PA₆₉₅, PA₈₀₀.

PA anti-interference capability of SPNE under multiple excitations

A 200 μL centrifuge tube was filled with the Tris buffer (50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35) containing SPNE (5 μM) with or without NE (2 U/L) for 30 min at 37 °C. PA₆₉₅ and PA₈₀₀ were continuously recorded 10 times through the MSOT system, and the PA₆₉₅/PA₈₀₀ of SPNE with NE or not was calculated according to PA₆₉₅, PA₈₀₀.

PA anti-interference capability of SPNE under different probe concentrations

A 200 μL centrifuge tube was filled with the Tris buffer (50 mM Tris, 1 M NaCl, pH 7.4, containing 0.05% Brij-35) containing different concentrations of SPNE (4, 6, 8, 10 μM). Then NE (2 U/L) was added into this solution, and incubated for 30 min at 37 °C. PA₆₉₅ and PA₈₀₀ were recorded through the MSOT system, and the PA₆₉₅/PA₈₀₀ of SPNE with NE was calculated according to PA₆₉₅, PA₈₀₀.

Cellular culture

Mouse breast cell line 4T1 cells, Human monocyte line THP-1 cells, Mouse melanoma cell line B16 cells were purchased from the Punuosai biology science and technology Co., Ltd. 4T1, THP-1, and B16 were cultured in DMEM supplemented with 1% antibiotics penicillin/streptomycin (100 U/mL) and 10% fetal bovine serum. THP-1 was cultured in 1640 supplemented with 1% antibiotics penicillin/streptomycin (100 U/mL) and 10% fetal bovine serum. Primary mouse neutrophils were extracted from living mice's blood by using the peripheral blood neutrophil isolation kit (P9201, Solarbio Science & Technology Co., Ltd.), and cultured in Free 1640 medium. All cells were cultured in a cell incubator containing 5% CO₂ at 37 °C.

Cell imaging

Neutrophils were seeded into confocal cell culture dishes (5 × 10⁴ cell/dish) and incubated for 24 h. Then cells were pre-incubated with fMLP (10 μM, 1 hour) or NE inhibitor (Silvelestat, 20 μM, 1 hour) or nothing. The treated cells were then incubated with SPNE at concentrations of 5 μM for 2 h. After incubation, the cells were stained with Hoechst 33342 for nuclei and directly taken for real-time imaging. Cell fluorescence images were acquired by a Confocal Laser Scanning Microscope. Hoechst used an excitation wavelength of 405 nm and emission wavelength of 425–475 nm; SPNE used an excitation wavelength of 639 nm and emission wavelength of 663–738 nm.

ImageJ software was utilized to remove the signal background and quantify cellular fluorescence intensity.

Flow cytometry analysis

For flow cytometry analysis of NE by SPNE, neutrophils, 4T1, THP-1, and B16 were seeded into 12-well plates (5 × 10⁴ cells/well), respectively. After 24 h incubation, neutrophils were treated with fMLP (10 μM, 1 h) or NE inhibitor (20 μM, 1 h), respectively. Subsequently, all cells were treated with SPNE (5 μM) for 2 h. Then all the cell was collected, transferred to a centrifuge tube, washed three times, and replaced with fresh medium. Suspensions obtained were analyzed in a flow cytometer.

Cytotoxicity assay

For cytotoxicity assay, neutrophils were seeded in 96-well plates (1 × 10⁴ cells/well) and cultured for 24 h. The cell was treated with different concentrations of SPNE (0, 10, 20, and 50 μg/mL) and cultured for 24 h. After incubation, the cell was collected, centrifuged, and washed three times with DPBS and replaced with free 1640 medium containing CCK-8 (C0038, Beyotime Biotechnology Co., Ltd) for another 4 h. The absorbance of CCK-8 was measured using a microplate reader at 405 nm. Cell viability was calculated by the ratio of the absorbance of the cells incubated with SPNE to that of the cells incubated with cell culture medium only.

Quantification of neutrophil elastase (NE) content in the aorta by photoacoustic (PA) signal and evaluation of the relationship between photoacoustic (PA) ratio signals and plaque vulnerability

Establishment of an atherosclerosis mouse model

All animal procedures were conducted in strict accordance with relevant laws and ethical guidelines, receiving approval from the Institutional Animal Care and Use Committee of Hunan University (protocol number: SYXK (Xiang) 2022-0007). Mice were group-housed in ventilated clear plastic cages under appropriate ambient temperature (~25 °C) and humidity (50%), and standard 12 h:12 h light/dark conditions.

For the establishment of an aortic atherosclerosis model, ApoE^-/- (Apolipoprotein E knockout) mice (male, 8 weeks old, 18-20 g) were procured from GemPharmatech Co., Ltd (Nanjing, China). Then, all mice fed a high-fat diet (HFD, Medicience Biomedical Technology Co., Ltd.) for 24 weeks. The composition of high-fat feed is shown in the following table.

Photoacoustic signal localization experiments and in vivo photoacoustic imaging in aortic plaques

Before injection of SPNE into the 24-week HFD ApoE^-/- male mice (AS mice), marks were made on the lateral thigh and near the scapula of AS mice for spatial localization by black mark pen. Then the mice were anesthetized, and SPNE (200 μL, 200 μM) was injected intravenously into AS mice. 30 min later, PA images of marks and the whole aorta at 695 nm and 800 nm were acquired by tomography. In the PA imaging process, the marks on the mice produced a very strong photoacoustic signal, and its signal strength is much stronger than the background noise, allowing us to identify them easily. Then we identify the photoacoustic image corresponding to the mark, record the coordinates of this image, and set it as the origin X₀. Subsequently, we dissected the mice and exposed the aorta and the marker. We identify the target plaques (plaque 1, 2, 3) and measure the actual distance between the plaque and the mark, and record it as D₁, D₂, D₃. We bring the measured distance into the image coordinates to obtain the coordinates of the plaque in the photoacoustic image. At this point, we successfully matched the isolated plaques to the photoacoustic images.

Ex vivo fluorescence imaging of the aorta

After PA imaging of AS mice, the aorta was harvested from the mice for fluorescence imaging by the IVIS spectral imaging system, and bright field images.

Imaging parameters: Fluorescence; Excitation wavelength: 640 nm; Filter: Cy5.5 channel (690 to 770 nm).

Determination of Neutrophil Elastase content in aortic plaques

To examine the NE content in different aortic segments, four aortic segments with different fluorescence intensities were collected according to the result of ex vivo fluorescence imaging. The plaques were thoroughly ground on an ice bath by adding PBS (mass/mass = 1/9), the tissue homogenate was centrifuged, and the supernatant was taken for NE content determination. The corresponding NE contents of the four fragments were determined by NE enzyme-linked immunosorbent assay kit (JL12179, Jianglai Biological Co., LTD., Shanghai, China). The specific methods are as follows:

• Step 1, 1 mL of universal diluent was added to the lyophilized standard and then diluted at the following concentrations: 1000 ng/mL, 500 ng/mL, 250 ng/mL, 125 ng/mL, 62.5 ng/mL, 31.25 ng/mL, 15.62 ng/mL, 0 ng/mL.

• Step 2, The samples or standards with different concentrations were added into the corresponding Wells according to 100 μl per well, and 100 μL universal diluent was added into the blank Wells. The plates were covered with the sealing membrane and incubated at 37 °C for 60 min.

• Step 3, The microplate was removed, and the liquid was discarded without washing. 100 μL of biotinylated antibody working solution was added directly to each well, and the plates were covered and incubated at 37 °C for 60 min.

• Step 4, Discard the liquid, add 300 μL 1x washing solution to each well, let stand for 1 minute, shake off the washing solution, pat dry on absorbent paper, and repeat washing plate 3 times.

• Step 5, 100 μL of enzyme conjugate working solution was added to each well, and the plate sealing membrane was covered and incubated at 37 °C for 30 min.

• Step 6, Wash the plate 5 times according to the step 4 washing method.

• Step 7, 90 μL of substrate (TMB) was added to each well, capped with plate sealing membrane, and incubated at 37 °C in the dark for 15 min.

• Step 8, The microplate was removed, 50 μL of termination solution was added directly to each well, and the OD value of each well was immediately measured at a wavelength of 450 nm.

Analysis of vulnerable components in different segments of the aorta

After mice were euthanized, three segments of the aorta with plaques were collected and immediately embedded in an optimal cutting temperature compound (OCT). Samples were frozen at −80 °C for 30 min and sectioned at 10 μm thickness using a Leica CM1950 freezing microtome.

For H&E staining, the frozen sections were placed in the paraformaldehyde fixative solution (G1101, Wuhan Service Bio Technology Co., Ltd.) for 30 min and washed with PBS. Then the tissue sections of blood vessels were stained with hematoxylin (CR2112128, Wuhan Service Bio Technology Co., Ltd.) at room temperature for 5 min. Subsequently, the tissue sections were washed with acidic ethanol at room temperature for 10 s, followed by reverse blue with bluing solution at room temperature for 3 min. Next, the tissue sections were washed three times with deionized water and stained with eosin (CR2202010, Wuhan Service Bio Technology Co., Ltd.) at room temperature for 10 s. Subsequently, the tissue sections were washed three times with 10 % ethanol, and then the sections were mounted with a drop of neutral balsam and sealed with a coverslip.

For Sirius red staining, the frozen sections were placed in the paraformaldehyde fixative solution (G1101, Wuhan Service Bio Technology Co., Ltd.) for 30 min and washed with PBS. Then the tissue sections were stained with hematoxylin at room temperature for 5 min. Subsequently, the tissue sections were washed with deionized water, and stained with Sirius red (G1472, Solarbio Science & Technology Co., Ltd.) staining solution at room temperature for 8 min. Finally, the tissue sections were washed three times with deionized water, and then the sections were mounted with a drop of neutral balsam and sealed with a coverslip.

For macrophage staining, the frozen sections were placed in the paraformaldehyde fixative solution (G1101, Wuhan Service Bio Technology Co., Ltd.) for 30 min and washed with PBS. Then the tissue sections were blocked with 5% BSA at room temperature for 30 min. Next, the tissue sections were incubated with rat anti-mouse F4/80 antibody (AbD Serotec, MCA497, 1:200) for 2 h at 37 °C. After washing with PBS, the tissue sections were incubated with an Alexa Fluor 488-labeled goat anti-rat IgG (CHAMOT, CM008-0.1A2F, 1:200) for 1 hour at 37 °C. Subsequently, the tissue sections were washed three times with deionized water and dried at 37 °C. At last, the tissue sections were stained with DAPI for 15 min at 37 °C and sealed with a coverslip.

For smooth muscle cell staining, the frozen sections were placed in the paraformaldehyde fixative solution (G1101, Wuhan Service Bio Technology Co., Ltd.) for 30 min and washed with PBS. Then the tissue sections were blocked with 5% BSA at room temperature for 30 min. Next, the tissue sections were incubated with rabbit anti-mouse α-SMA antibody (GeneTex, GTX629702, 1:200) for 2 h at 37 °C. Then the sections were incubated with an Alexa Fluor 594-labeled goat anti-rabbit IgG (CHAMOT, CM003-0.1A2F, 1:200) for 1 hour at 37 °C. Subsequently, the sections were rinsed three times with deionized water and dried at 37 °C. At last, sections were stained with DAPI for 15 min at 37 °C and sealed with a coverslip.

Finally, the digital section scanning system was used to collect images of H&E staining sections and Sirius red-stained sections, and the immunofluorescence images were obtained using a Laser Confocal Microscope (Nikon, Japan). DAPI used an excitation wavelength of 405 nm and emission wavelength of 425–475 nm; F4/80 used an excitation wavelength of 488 nm and emission wavelength of 500–550 nm. α-SMA used an excitation wavelength of 560 nm and an emission wavelength of 570–620 nm.

ImageJ was used to calculate the area of green fluorescence in F4/80 immunofluorescence staining, the area of red fluorescence in α-SMA immunofluorescence staining, the extracellular vacuoles and lacunae area of the plaque stained by H&E staining sections, and the red area of the plaque stained by Sirius red staining sections, respectively. The proportion of macrophages, smooth muscle cells, necrotic cells, and collagen fibers in the plaque was calculated by dividing each of these areas by the corresponding plaque area.

The vulnerability index of different segments of the aorta

The proportions of macrophages, smooth muscle cells, necrotic cores, and collagen fibers in the aorta plaque area of mice in different groups were counted, and then the vulnerability index of the aorta plaque of mice in each group was obtained according to the vulnerability formula³⁹:

$${{{\rm{Vulnerability}}}}\; {{{\rm{index}}}} \, ({VI})=\frac{{{{\rm{Macrophages}}}} \, \left(\%\right)+{{{\rm{Necrotic}}}}\; {{{\rm{cores}}}} \, (\%)}{{{{\rm{Smooth}}}}\; {{{\rm{muscle}}}}\; {{{\rm{cells}}}} \, \left(\%\right)+{{{\rm{Collagen}}}}\; {{{\rm{fibers}}}} \, (\%)}$$

(3)

The proportion of vulnerable components and the vulnerability index of each segment are shown in Table S2.

In vivo photoacoustic imaging of neutrophil elastase (NE) in carotid plaque

Establishment of a carotid atherosclerosis mouse model

For the establishment of a carotid atherosclerosis model, ApoE^-/- (Apolipoprotein E knockout) mice (male, 8 weeks old, 18-20 g) were procured from GemPharmatech Co., Ltd (Nanjing, China). After the mice were anesthetized, an incision, approximately 3-4 mm in length, was made in the center of the neck of the mice, the left carotid artery was separated with a blunt instrument, and the left artery was ligated with a 7-0-braided polyester fiber suture. The wound was subsequently closed with sutures. Subsequently, mice were fed with high-fat diet for 8 weeks to obtain carotid plaque mice.

In vivo photoacoustic imaging of the carotid artery

Carotid plaque mice were randomly divided into three groups (n = 5): (i) AS mice: PBS was administered daily by gavage from days 8 to 21. (ii) AS mice with inflammation: ApoE^-/- male mice were given LPS (2 mg/mL, 50 μL) by nasal feeding on the days 1, 4, and 7, and the mice were treated with PBS by gavage from day 8 to 21. (iii) AS mice with inflammation and drug: ApoE^-/- male mice were given LPS (2 mg/mL, 50 μL) by nasal feeding for day 1, 4, and 7, and then rosuvastatin (10 mg/kg/d) and dexamethasone (2 mg/kg/d) were administered by gavage from day 8 to 21. Then all of the mice were i.v. injected with PBS containing SPNE (200 μL, 200 μM), and PA images at 695 nm and 800 nm were collected before and post-injection (15, 30, and 60 min).

During the PA imaging process, the MSOT system will generate the corresponding organ distribution map for each photoacoustic tomography cross-section, so as to match the signal in the photoacoustic image with the organ. Therefore, we imaged the carotid artery area through photoacoustic tomography and located the carotid artery area by the organ anatomical distribution map provided by the MSOT system. Finally, the NE content in this area was precisely analyzed using the photoacoustic ratio signal.

Cholesterol content in the serum of mice

To test cholesterol levels in the serum of mice, the carotid artery model mice and the healthy normal mice were anesthetized. Subsequently, mouse blood was collected and left for 20 min at room temperature. Then, mouse serum was obtained by centrifugation (1200 × g) for 15 min. The serum of each group was detected by a total cholesterol assay kit (Nanjing Jiancheng Bioengineering Institute, A111-1-1). The specific methods are as follows:

2.5 μL of the test sample, deionized water, and standard were added to the 96-well plate together with 250 μL of the working solution. The Wells were then shaken and incubated at 37 °C for 10 min. Finally, the absorption of each well at 500 nm was tested in a microplate reader.

The total cholesterol content of serum samples was calculated using the following formula:

$${{{\rm{Cholesterol}}}}\; {{{\rm{content}}}}\left({{{\rm{mmol}}}}/{{{\rm{L}}}}\right)=\frac{{{{{\rm{A}}}}}_{{{{\rm{s}}}}{{{\rm{ample}}}}}-{{{{\rm{A}}}}}_{{{{\rm{b}}}}{{{\rm{lank}}}}}}{{{{{\rm{A}}}}}_{{{{\rm{standard}}}}}-{{{{\rm{A}}}}}_{{{{\rm{b}}}}{{{\rm{lank}}}}}}\times {{{{\rm{C}}}}}_{{{{\rm{standard}}}}}$$

(4)

Among them, A_sample represents the absorbance of the test sample at 500 nm; A_blank represents the absorbance of the sample of deionized water at 500 nm, and A_standard represents the absorbance of the standard sample at 500 nm. C_standard represents the concentration of the standard sample (mmol/L).

H&E staining of the lungs in different groups

Mice in each group were sacrificed on the second day after imaging, those were sacrificed and the lungs of each group were collected and fixed with paraformaldehyde immediately. And then H&E staining of lungs was performed (Wuhan Service Bio Technology Co., Ltd.).

Analysis of the vulnerability index of carotid plaque

Histopathological analysis of carotid arteries in each group of atherosclerotic mice

On day 22 after the first LPS stimulation in AS mice with inflammation, the carotid artery of each group of mice was harvested post-euthanasia and immediately embedded in an optimal cutting temperature compound (OCT). Samples were frozen at −80 °C for 30 min and sectioned at 10 μm thickness using a Leica CM1950 freezing microtome.

For H&E staining and Sirius red staining, refer to ethe xperimental section in the previous text.

For Oil Red O staining, the frozen sections were placed in the paraformaldehyde fixative solution (G1101, Wuhan Service Bio Technology Co., Ltd.) for 30 min and washed with PBS. Then the tissue sections were stained with hematoxylin at room temperature for 5 min. Subsequently, the tissue sections were washed with deionized water and stained with Oil Red O staining solution at room temperature for 30 min. After washing with 60% isopropanol, the sections were mounted with a drop of neutral balsam and sealed with a coverslip.

Finally, ImageJ was used to calculate the extracellular vacuoles and lacunae area of the plaque stained by H&E staining sections, the red area of the plaque stained by Sirius red staining sections, and the red area of the plaque stained by Oil red O. The proportion of necrotic cells, collagen fibers, and lipid content in the plaque was calculated via dividing each of these areas by the corresponding plaque area.

Immunofluorescence staining of carotid arteries in each group of atherosclerotic mice

On day 22 after the first LPS stimulation in AS mice with inflammation, the carotid artery of each group of mice was harvested post-euthanasia and immediately embedded in an optimal cutting temperature compound (OCT). Samples were frozen at -80 °C for 30 min and sectioned at 10 μm thickness using a Leica CM1950 freezing microtome.

For Ly6G staining, the frozen sections were placed in the paraformaldehyde fixative solution (G1101, Wuhan Service Bio Technology Co., Ltd.) for 30 min and washed with PBS. Then the tissue sections were blocked with 5% BSA at room temperature for 30 min. Next, the tissue sections were incubated with antibody Alexa Fluor® 488 anti-mouse Ly-6G (Biolegend, 127625, 1:200) for 1 hour at 37 °C. Then, the tissue sections were washed three times with deionized water and dried at 37 °C. At last, sections were stained with DAPI for 15 min at 37 °C and sealed with a coverslip.

For macrophage staining and smooth muscle cell staining, refer to the experimental section in the previous text.

Finally, the immunofluorescence images were obtained using a Laser Confocal Microscope (Nikon, Japan). DAPI used an excitation wavelength of 405 nm and emission wavelength of 420–475 nm; F4/80 and Ly6G used an excitation wavelength of 488 nm and emission wavelength of 500–550 nm. α-SMA used an excitation wavelength of 560 nm and an emission wavelength of 570–620 nm.

ImageJ was used to calculate the area of green fluorescence in F4/80 immunofluorescence staining and the area of orange fluorescence in α-SMA immunofluorescence staining, respectively. The above area was divided by the corresponding plaque area to obtain the proportion of macrophages and smooth muscle cells in the plaque. The average intensity of red fluorescence in Ly6G immunofluorescence staining was calculated by ImageJ to obtain neutrophil infiltration.

The vulnerability index of different groups of carotid arteries

The statistical method of the proportion of plaque vulnerable components and the calculation method of plaque vulnerable index are referred to experiment section in the previous text.

The proportion of vulnerable components and the vulnerability index of each group are shown in Table S3.

H&E staining of major organs in different groups of mice

Mice in each group on the second day after imaging, the major organs (heart, liver, spleen, and kidney) of the mice were collected immediately and fixed in paraformaldehyde, and H&E staining was performed (Wuhan Service Bio Technology Co., Ltd.).

In vivo imaging in mice with intimal hyperplasia

Establishment of an intimal hyperplasia mouse model

For the establishment of an intimal hyperplasia model, C57BL/6 mice (male, 8 weeks old, 18-20 g) were procured from GemPharmatech Co., Ltd (Nanjing, China). After the mice were anesthetized, an incision, approximately 3–4 mm in length, was made in the center of the neck of the mice, the left carotid artery was separated with a blunt instrument, and the left artery was ligated with a 7-0-braided polyester fiber suture. The wound was subsequently closed with sutures. Subsequently, intimal hyperplasia mice were obtained after 8 weeks of rearing.

In vivo photoacoustic imaging of intimal hyperplasia

C57BL/6 mice with carotid artery ligation were randomly divided into two groups after carotid artery ligation for 12 weeks (n = 4): (i) Intimal hyperplasia mice: No treatment was done during the first 7 days. (ii) Intimal hyperplasia in mice with inflammation: mice were given LPS by nasal feeding on days 1, 4, 7. Then all mice were i.v. injected with PBS containing SPNE (200 μL, 200 μM), and PA images were collected before and post-injection (5, 15, 30, 45, and 60 min).

During the PA imaging process, the MSOT system will generate the corresponding organ distribution map for each photoacoustic tomography cross-section, so as to match the signal in the photoacoustic image with the organ.

Histopathological analysis of carotid arteries in each group of Intimal hyperplasia mice

On day 8 after the first LPS stimulation in mice with intimal hyperplasia, the carotid artery of each group of mice was harvested post-euthanasia and immediately embedded in an optimal cutting temperature compound (OCT). Samples were frozen at −80 °C for 30 min and sectioned at 10 μm thickness using a Leica CM1950 freezing microtome.

For H&E staining, refer to the experiment section in the previous text.

Finally, all the stained sections were examined using a digital slice scanning system.

Immunofluorescence staining of carotid arteries in each group of Intimal hyperplasia mice

On day 8 after the first LPS stimulation in mice with intimal hyperplasia, the carotid artery of each group of mice was harvested post-euthanasia and immediately embedded in an optimal cutting temperature compound (OCT). Samples were frozen at −80 °C for 30 min and sectioned at 10 μm thickness using a Leica CM1950 freezing microtome.

The frozen sections were placed in the paraformaldehyde fixative solution (G1101, Wuhan Service Bio Technology Co., Ltd.) for 30 min and washed with PBS. Then the tissue sections were blocked with 5% BSA at room temperature for 30 min. Next, the tissue sections were incubated with rat anti-mouse F4/80 antibody (AbD Serotec, MCA497, 1:200) for 2 h at 37 °C. After washing with PBS, the tissue sections were incubated with an Alexa Fluor 488-labeled goat anti-rat IgG (CHAMOT, CM008-0.1A2F, 1:200) for 1 hour at 37 °C. Then, the sections were washed three times with deionized water. Next, the tissue sections were incubated with antibody Alexa Fluor® 647 anti-mouse Ly-6G (Biolegend, 127609, 1:500) for 1 hour at 37 °C. Subsequently, the tissue sections were washed three times with deionized water and dried at 37 °C. At last, the tissue sections were stained with DAPI for 15 min at 37 °C and sealed with a coverslip.

Finally, the immunofluorescence images were obtained using a Laser Confocal Microscope (Nikon, Japan). DAPI used an excitation wavelength of 405 nm and emission wavelength of 425–475 nm; F4/80 used an excitation wavelength of 488 nm and emission wavelength of 500–550 nm. Ly6G used an excitation wavelength of 639 nm and an emission wavelength of 663–738 nm.

Statistics analysis

All sample sizes are reflected in the figure legends. The results were expressed as mean ± SD unless otherwise stated. Statistical significance was determined using a two-tailed Student’s t-test for pairwise comparisons, and one-way ANOVA analysis of variance for multiple groups via Excel 2019. *P < 0.05, **P < 0.01, and ***P < 0.001, ‘ns’ indicating no significant difference (P > 0.05). ViewMSOT was used to analyze the photoacoustic images. Flow cytometry results were analyzed by FlowJo-V10. Living Mage 4.3 software PerkinElmer, was used to analyze the fluorescence images. ImageJ was used to analyze immunofluorescence images. Origin 2018 was used to represent the analyzed data in graphical formats.