MMP-14-targeted Nanoprobe for in vivo Fluorescence Imaging of Rupture-prone Carotid Plaques


 Background: The identification of rupture-prone carotid plaques for preventing stroke remains a clinical challenge. Macrophage matrix metalloproteinase (MMP)-14, which contributes to plaque progression and destabilisation, could be a promising biomarker for plaque imaging. This study aimed to design and synthesise an MMP-14-targeted nanoprobe to noninvasively visualise the behaviour of M1 macrophages in atherosclerotic plaques.Methods: A fluorescence molecular imaging probe (AuNPs@PEG-Peptide-Cy5.5) was constructed by covalently attaching the fluorescent dye cyanine (Cy) 5.5, an MMP-14 substrate, and polyethylene glycol (PEG) 5000-wrapped gold nanoparticles (AuNPs), and then administered via tail vein injection to carotid atherosclerosis models for in vivo fluorescence imaging. Additionally, carotid tissues and cultured macrophages were analysed for nanoprobe binding, and MMP-14 and inflammation-related marker expression was evaluated by polymerase chain reaction, western blotting, and immunohistochemistry.Results: MMP-14 expression significantly increased with plaque progression, along with the upregulation of MMP-2 and inflammatory M1 markers, CD68 and F4/80, and significant downregulation of the M2 marker CD206. All of cell, tissue and in vivo fluorescence imaging exhibited a favourable targeting efficacy of AuNPs@PEG-Peptide-Cy5.5 for MMP-14.Conclusions: MMP-14, a cell membrane-anchoring enzyme, can serve as a biomarker of vulnerable plaques, and MMP-14 substrate-based AuNPs@PEG-Peptide-Cy5.5, with an intense fluorescence signal after activation and good biocompatibility, can be applied to screen for and monitor plaque progression in vivo.


Background
Carotid atherosclerosis is a strong risk factor for ischemic stroke [1]. As a multifactorial and complex systemic disease, carotid atherosclerosis is characterised by the formation of atherosclerotic plaques, which is a lengthy process that involves the deposition of cholesterol, in ammation, and cell in ltration into vessel walls [2,3]. After a plaque buildup, its enlargement may result in a severe occlusion of the lumen and cause distal hypoperfusion or rupture to embolise a distal branch, leading to morbidity and mortality [4]. Therefore, there is an urgent need to identify plaques that are prone to enlargement or rupture.
Matrix metalloproteinase (MMP)-14, a cell membrane-anchoring enzyme, has been reported to be overexpressed in vulnerable human atherosclerotic plaques [5][6][7][8]. MMP-14 can activate pro-MMP-2 [9] and pro-MMP-13 [10], as well as interact with the hyaluronate receptor CD44 to activate MMP-9 [11] to break down the extracellular matrix within the lesion, converting chronic stable plaques into acute unstable plaques, with the potential for thromboembolism [12]. Macrophages contribute to a broad range of chronic in ammatory processes and furnish most MMPs in atheromata [6,13], and therefore, many atherosclerosis-related studies have focused on macrophages. Immunohistochemical results have con rmed that MMP-14 is mainly codistributed with macrophages in the fragile region of rupture-prone plaques [7,14], although other types of cells, such as smooth muscle cells, also express small amounts of MMP-14. Furthermore, MMP-14-directed macrophage invasion increases the likelihood of the rupture at the shoulder of atherosclerotic plaques [5,15]. Thus, MMP-14 from macrophages could be a promising biomarker for plaque imaging.
Current imaging technologies, such as computer tomography, magnetic resonance imaging, and positron emission tomography (PET), enable the determination of the severity of luminal narrowing and the assessment of plaque vulnerability at the morphological level; however, these technologies have a poor accuracy and speci city and cannot provide information about molecular signatures of rupture-prone plaques at the cellular level [4,16]. Consequently, more advanced imaging modes are needed to demonstrate the extent of the disease spread. In vivo molecular imaging strategies provide a way beyond the anatomy to visualise atherosclerotic in ammation, angiogenesis, apoptosis, oxidative stress, and calci cation, thus overcoming the limitations of traditional imaging methods [17]. However, primary intravascular molecular imaging uses radioisotope-derivatised monoclonal antibodies, whose large sizes do not allow fast on/off rates and produce high target-to-background ratios in vivo [18,19]. Near-infrared (NIR) uorescence imaging with a targeted nanoprobe is a recent and promising imaging technology that allows noninvasive visualisation of the metabolic state of various types of cells, local in ammatory responses, and intraplaque angiogenesis [20][21][22]. Moreover, its real-time modality, high sensitivity, and high speci city, as well as the absence of ionising radiation, make NIR uorescence a preferred option for the diagnosis of rupture-prone plaques to assist in clinical decision making [20,23].
The main aim of this study was to investigate and validate the mechanism by which MMP-14, a promising biomarker of M1 macrophages, contributes to the progression and rupture of atherosclerotic plaques. For this purpose, an MMP-14-targeted nanoprobe, AuNPs@PEG-Peptide-Cy5.5, was designed and synthesised to noninvasively visualise the behaviour of M1 macrophages in atherosclerotic plaques.

Synthesis of AuNPs@PEG-Peptide-Cy5.5
A cyanide dye (Cy) 5.5-labelled Gly-Arg-Ile-Gly-Phe-Leu-Arg-Thr-Ala-Lys-Gly-Gly peptide (5 mg, 0·0028 mmol) was synthesised as previously reported [24,25], and commercially available SHpolyethylene glycol (PEG)-COOH-modi ed Au nanoparticles (NPs) were purchased from XFNANO (Nanjing, Jiangsu, China). Next, 10 µL of 0·1% sodium dodecyl sulphate (SDS) and 0·5 mL of the Cy5.5labelled peptide (2·453 mg/mL) were added to 10 mL of the AuNPs [0·1 mg/mL, dispersed in 2-(Nmorpholino) ethanesulphonic acid], and stirred for 30 min. Thereafter, 0·5 mg of 1-ethyl-3-(3dimethylaminopropyl) carbodiimide (10 mg/mL) was added, and the mixture was stirred at room temperature overnight. At the end of the reaction, the sample was puri ed via ultra ltration. The AuNPs were dispersed in pure water to a nal concentration of 0·05 mg/mL in 20 mL, and a small amount of SDS was added to the suspension as a stabiliser. The coupling rate of the uorescent peptide to the AuNPs was calculated based on the volume of the suspension and the amount of the uorescent peptide.

Probe characterisation
The NP morphology and size were assessed using transmission electron microscopy (TEM) and dynamic light scattering (DLS). For sample preparation, AuNPs@PEG and AuNPs@PEG-Peptide-Cy5.5 in deionised water were placed on a 400-mesh copper grid and dried. TEM imaging was carried out using a FEI Tecnai F20 transmission electron microscope at an acceleration voltage of 200 kV. DLS measurements were performed using a Zetasizer Nano ZS90 (Malvern Instruments, Worcestershire, UK), and hydrodynamic sizes of samples were calculated. Fluorescence spectra were recorded using a uorescence spectrometer (FLS980; Edinburgh, UK). NIR uorescence imaging was performed using an IVIS Spectrum system (PerkinElmer).

Animals
Twenty male 6-week-old Apoe −/− C57BL/6J mice (Huafukang Biotechnology Co., Beijing, China) were fed a standard laboratory chow diet (Cat #1010009; Xietong Pharmaceuticals Co., Jiangsu, China) for 1 week. After acclimation, the mice were randomly allocated into a normal (control) group and a model group (n = 10 each). The control group was fed the standard laboratory chow diet for 12 weeks. The model mice received a rigid, silicone, perivascular constrictive cuff placed around the right common carotid artery, as previously reported [26], and were fed a high-fat diet (HFD; 0·25% cholesterol and 15% cocoa butter). The animal handling and care and all procedures used in this study were performed in accordance with the regulations of the Animal Ethics Committee of Shandong University (IACUC Issue NO. DWLL-2021-001).

Histopathology
Mice were sacri ced at 4, 8, and 12 weeks (n = 3 each) after cuff implantation to observe the formation of atherosclerotic plaques. Carotid artery samples were xed with 10% formalin and embedded in para n at room temperature. The para n blocks were cut into 4-µm-thick slices, which were stained with haematoxylin and eosin (H&E) to calculate the areas of the vessel lumen and the inner and outer tunica media for each animal. Slices adjacent to the H&E-stained slices were selected for immunohistochemical analysis for MMP-14, MMP-2, CD68, CD206, and F4/80. After depara nisation and rehydration, the slices were incubated with foetal bovine serum at 37°C for 30 min, followed by incubation with primary rabbit and mouse monoclonal antibodies overnight at 4°C, washing, and incubation with horseradish peroxidase-conjugated mouse anti-rabbit secondary antibodies. Finally, the slices were stained with 3, 3′diaminobenzidine (Solarbio, China), counterstained with haematoxylin, and scanned using a microscope (BX53; Olympus, Japan).

Cytotoxicity assessment
For the cell viability assay, cells were seeded at a density of 4 × 10 4 per well in 96-well plates and incubated at 37°C in a 5% CO 2 incubator for 24 h. Thereafter, the cells were treated with increasing concentrations (0 to 50 µg/mL) of the uorescent probe for 24 h, and cell viability was evaluated using Cell Counting Kit-8 (CCK-8; Sigma-Aldrich, St. Louis, MO, USA). For apoptosis detection, cells were seeded in 6-well plates and cultured for 24 h, followed by staining with annexin V-PI (BD Biosciences, Shanghai, China) and ow cytometry. To detect organ necrosis, model mice injected with saline and with probe were sacri ced 24 h post-injection and major organs were carefully harvested. All samples were stained with H&E, placed on black paper, and imaged under a microscope.

Immuno uorescence
Raw 264.7 cells were seeded into an 8-well chamber slide (Ibidi, Germany) and then treated with or without 75 µg/mL oxidized low-density lipoprotein (ox-LDL) or GM6001. After 24 h of treatment, the cells were xed with 4% paraformaldehyde and incubated with an MMP-14 antibody (1:500, Cat #ab51074; Abcam) at 4°C overnight. The samples were then washed with cold phosphate-buffered saline (PBS) three times and incubated with an Alexa Fluor-conjugated secondary antibody at room temperature for 2 h. Afterwards, the cells were incubated with AuNPs@PEG-Peptide-Cy5.5 for another 1 h, washed, and mounted with a glue solution containing 4′, 6-diamidino-2-phenylindole (DAPI). Images were captured using a uorescence microscope.

In vivo imaging and ex vivo biodistribution studies
In vivo images were acquired and analysed using the IVIS Spectrum system (PerkinElmer). All mice were shaved at the neck area, and those from the model group (n = 3) were administered AuNPs@PEG-Peptide-Cy5.5 at a dose of 5 mg/kg via tail vein injection, while mice from the control group (n = 3) were injected saline. Imaging was performed at 10 and 30 min and 1, 2, 4, 8, 12, and 24 h after injection. During the injection and image acquisition process, the mice were anesthetised with 2·5% iso urane in oxygen at a ow rate of 1·5 L/min. All images were normalised and analysed using the Maestro software. For quantitative comparison, regions of interest were drawn over the necks, and the average signal for each area was measured. For an ex vivo biodistribution study, mice were sacri ced at 24 and 48 h after injection of the probe. The major organs were carefully harvested and rinsed with PBS (pH 7·4), then placed on black paper, and immediately imaged using the IVIS Spectrum.

Statistical analysis
Continuous data were compared between the two groups using a t-test. A chi-squared or Fisher's exact test was used for categorical data. The nonparametric Mann-Whitney test was used for comparison of unpaired data. ANOVA test was used to compare means of three or more samples. Statistical signi cance was set at p < 0·05. All statistical analyses were performed using GraphPad Prism version 8.0 (GraphPad Software, La Jolla, CA, USA) or the SPSS software version 21.0 (SPSS, Chicago, IL, USA) and presented as the mean ± standard deviation. 21520011691806). The funders had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication. The corresponding authors have full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Results
3.1. Design and characterisation of AuNPs@PEG-Peptide- Cy5.5 In this study, we rst synthesised an MMP-14-speci c and sensitive substrate peptide with a sequence of Gly-Arg-Ile-Gly-Phe-Leu-Arg-Thr-Ala-Lys-Gly-Gly, which was then labelled at the N-terminus with a NIR uorescent dye, Cy5.5. Subsequently, the Cy5.5-conjugated peptide was coupled to the surface of PEG 5000-wrapped AuNPs to construct a uorescence resonance energy transfer system. Owing to the surface energy transfer properties [27], the uorescence of Cy5.5 was quenched until the substrate peptide was degraded by the MMP-14 enzyme, which was highly expressed by M1 macrophages, as shown in A TEM image of AuNPs@PEG-Peptide-Cy5.5 is shown in Fig. 2a. The probe exhibited good dispersion in water, with an average particle diameter of 10 nm. Compared with AuNPs@PEG, which had an average hydrodynamic diameter of 28·76 ± 7·37 nm (Fig. 2b), the probe displayed a similar or even higher monodispersity, with an average hydrodynamic diameter of 32·07 ± 6·71 nm (Fig. 2c). The optical properties of the nanoprobe were determined by ultraviolet-visible (UV-Vis) spectroscopy and are displayed in Fig. 2d, which shows a peak at 520 nm, which is the typical absorbance of AuNPs@PEG, and peaks at 642 and 698 nm, which are minimally different from the Cy5.5 absorbance at 634 and 677 nm, indicating the successful conjugation of AuNPs and the Cy5.5-labelled peptide. The slight shifts in the absorption peaks of Cy5.5 were caused by its strong interaction with AuNPs. The peptide concentration in the puri ed supernatant of AuNPs@PEG-Peptide-Cy5.5 was 31 mg/mL when the AuNPs@PEG concentration was 0·05 mg/mL; therefore, we estimated that there were 1·77 × 10 2 molecules of the Cy5.5-labelled peptide per AuNP. The intensity of the AuNPs@PEG-Peptide-Cy5.5 peak, which appeared in an extremely narrow range, from 673 to 676 nm (Fig. 2e) was 56% lower than that of the Cy5.5-labelled peptide, indicating an obvious quenching e ciency. To investigate the responsiveness of the probe to MMP-14, in vitro enzymatic assays were performed by incubating recombinant MMP-14 (320 ng/mL) and its inhibitor GM6001 with AuNPs@PEG-Peptide-Cy5.5 in PBS buffer (pH 7·4) at 37°C for 1 h. As shown in Fig. 2f, a rapid uorescence recovery of the nanoprobe was triggered by the addition of MMP-14, while the e cient uorescence recovery of the nanoprobe was signi cantly prevented when the enzyme activity was inhibited by GM6001. To con rm that the NIR uorescent signal of the probe was ampli ed as the MMP-14 concentration increased, an incubation was carried out with different concentrations (0, 10, 20, 40, 80, 160, and 320 ng/mL) of MMP-14. The NIR uorescence emission signals of Cy5.5 were measured using a spectro uorometer for 60 min. As shown in Fig. 2g, the uorescence was progressively intensi ed and positively linearly correlated with the concentrations of MMP-14 (R 2 = 0·94). To con rm the biological safety of the nanoprobe, in vitro cytotoxicity of AuNPs@PEG-Peptide-Cy5.5 was investigated using the CCK-8 assay (Fig. 2h) and ow cytometry analysis (Supplementary Fig. S1). Both assays demonstrated no signi cant cytotoxicity against RAW 264.7 cells after incubation with AuNPs@PEG-Peptide-Cy5.5 at different concentrations (0 to 50 µg/mL) for 24 h, and there were no signi cant differences in the populations of RAW 264.7 cells that underwent apoptosis and necrosis compared with those in the control without AuNPs@PEG-Peptide-Cy5.5 treatment.

MMP-14 production in carotid samples and plaque formation
To detect MMP-14 expression and visualise plaque formation, right carotid arteries were harvested every 4 weeks and analysed by qPCR and western blotting for in ammatory markers; by H&E staining for morphological evidence; and by immunohistochemical staining for molecular signatures. As shown in Fig. 4a and b, the results of qPCR and western blotting demonstrated gradually increased expression of MMP-14, MMP-2, CD-68, and F4/80 and reduced levels of CD206 in carotid samples collected from atherosclerotic model mice at 0, 4, 8, and 12 weeks. Quantitative analysis (Fig. 4c) showed that the protein levels of MMP-14, MMP-2, CD68, and F4/80 were signi cantly upregulated (almost vefold, p < 0.0001) in carotid samples from the model mice at 12 weeks compared with those at 0 weeks, while that of CD206 was largely reduced, indicating that the expression of MMP-14 increased with the enhancement of in ammatory activities of M1 macrophages. In addition, H&E staining revealed that the enlargement of atherosclerotic plaques was accompanied by a thinning of brous caps and an increase of necrotic cores, which increased the likelihood of shoulder rupture, and immunohistochemical staining further con rmed increased levels of MMP-14, MMP-2, CD-68, and F4/80 and reduced deposition of CD206 (Fig. 4d). Thus, MMP-14 is a good marker that re ects the in ammatory responses of macrophages and the formation of rupture-prone plaques. We then simultaneously performed immuno uorescence staining using an MMP-14 antibody and nanoprobe staining of carotid samples and found that the stained areas overlapped. Furthermore, after preincubation of tissue sections with the MMP-14 inhibitor GM6001 for 30 min, the NIR uorescent signal was signi cantly reduced, suggesting that AuNPs@PEG-Peptide-Cy5.5 had a good binding ability that could be used for MMP-14 detection (Fig. 4e).

In vivo and ex vivo molecular imaging of carotid plaques
To evaluate the feasibility of using AuNPs@PEG-Peptide-Cy5.5 for in vivo imaging of MMP-14 activity, the probe was intravenously administered to carotid atherosclerosis model mice and to control mice of the same age, sex, and background (Fig. 5a). Plaque formation was rst assessed by microultrasound imaging to ensure that the carotid atherosclerosis model was successfully constructed ( Supplementary  Fig. S2). Next, the AuNPs@PEG-Peptide-Cy5.5 nanoprobe was injected into the tail vein of mice at a dose of 2·5 mg/kg. In vivo uorescence images were recorded over a period of 24 h at an excitation of 675 nm using an IVIS system. Representative images are shown in Fig. 5a at selected time points (10 and 30 min and 1, 2, 4, 8, 12, and 24 h) after injection, with strong activation of the NIR uorescence of AuNPs@PEG-Peptide-Cy5.5 observed in the neck of the carotid atherosclerosis model mice. It should be noted that the probe showed a rapid, early onset of the activation (in less than 10 min), and the uorescence remained relatively steady for 2 h, with a high average radiant e ciency compared with that in the control mice (Fig. 5b). After 2 h, the uorescent signal was quickly reduced and almost completely cleared after 12 h. Additionally, ex vivo imaging showed stronger uorescence signals in the right carotid arteries of the model mice than in those of the control mice treated with AuNPs@PEG-Peptide-Cy5.5 (Fig. 5c), which was consistent with the in vivo results shown in Fig. 5a. Collectively, these results strongly demonstrate that AuNPs@PEG-Peptide-Cy5.5 have a great potential as a NIR probe for sensitive detection of MMP-14 and for the diagnosis of the progression of atherosclerotic plaques.

Ex vivo histological and biodistribution studies
Ex vivo molecular imaging of various organs was performed at 24 and 48 h after in vivo imaging to detect biodistribution of the probe. As shown in Fig. 6a and b, the biodistribution study indicated that AuNPs@PEG-Peptide-Cy5.5 was predominantly metabolised in the liver, kidney, and intestine. After 48 h, the average radiant e ciency declined by more than half, but that of the kidney was still higher than that of the other organs, indicating that the kidney was the main organ for the metabolism of the nanoprobe. H&E staining images of various organs obtained from the control mice after intravenous injection of a PBS solution and from the model mice after AuNPs@PEG-Peptide-Cy5.5 injection at a dose of 2·5 mg/kg showed no apparent abnormalities in the heart, pancreas, liver, kidney, muscle, lung, spleen, intestine, and brain (Fig. 6a). Therefore, our data con rmed that AuNPs@PEG-Peptide-Cy5.5 were notably effective in detecting MMP-14 activity and were not toxic to normal organs, suggesting that this probe can serve as an agent for rupture-prone plaque detection, with great clinical translational potential.

Discussion
In the present study, we con rmed that MMP-14, a cell membrane-anchoring enzyme, could serve as a biomarker of vulnerable plaques. The data also demonstrated that the MMP-14 substrate-based uorescent nanoprobe AuNPs@PEG-Peptide-Cy5.5, which showed an intense uorescence signal after activation and good biocompatibility, could be applied to screen for and monitor plaque progression in vivo.
Satisfactory imaging techniques are urgently needed to detect vulnerable plaques, which are mediated by in ammatory activities and cause lethal cerebrovascular events [30,31]. In recent years, molecular targetbased imaging of the degree of plaque in ammation has been considered a promising approach for identifying rupture-prone plaques [32,33]. Many molecular biomarkers, such as MMPs [34], hypoxiarelated factors, and vascular cell adhesion molecules [35], have been proposed. Of these, MMPs are particularly attractive, as they are proteinases that directly degrade the matrix and erode the brous cap and are involved in a series of other regulatory mechanisms that promote plaque instability. Seifert et al. [34] implanted a tapered cuff around the right common carotid artery in Apoe −/− mice to induce a model of atherosclerosis and found that the MMP-2 and MMP-9 activities were signi cantly higher in upstream, low shear stress-induced unstable atherosclerotic plaques than in downstream, more stable plaque phenotypes. However, many uorogenic probes targeting MMP-2 and MMP-9 have been demonstrated to have limitations in in vivo applications. One important reason is that both MMP-2 and MMP-9 are extracellularly secreted, soluble types of MMPs. They are abundantly secreted into the bloodstream, providing high background signals, and probes are easily washed away, which thus lowers the imaging quality [36,37]. To compensate for these drawbacks, many reported MMP-2 and MMP-9 probes were designed to carry another recognition sequence to anchor these probes to the extracellular membrane of speci c cells, which increased the technical di culty of probe synthesis, while the effects might not have been as expected [38]. Unlike extracellular MMPs, membrane-type (MT)-MMPs are tethered to the plasma membrane via either a glycosylphosphatidylinositol linkage or a transmembrane domain [39]. Moreover, MMP-14, also known as MT1-MMP, is directly linked to the pathogenesis of plaque vulnerability, as MMP-14 has been shown to convert MMP-2 and MMP-9 proenzymes into their active forms and to act as an attractant to facilitate monocyte/macrophage in ltration into sites of experimentally induced in ammation and established atherosclerotic lesions [5,40]. To detect cardiac ischemia/reperfusion injury, van Duijnhoven et al. [41] developed a series of activatable cell-penetrating peptides (ACPPs) that were sensitive to MMP-2 and MMP-9 (ACPP-2/9) or to MMP-14 (ACPP-B). Although both ACPPs successfully detected regions of an infarcted myocardium in mouse models, the ACPP-2/9 probe showed a considerable degree of activation in all tissues, while ACPP-B was found to be tissue speci c. Therefore, MMP-14 is more suitable than MMP-2 and MMP-9 as a target for plaque in ammation imaging.
In this study, the core sequence, Arg-Ile-Gly-Phe-Leu-Arg, of the peptide was adopted from previous reports. This sequence has been veri ed to have a high speci city and selectivity for MT-MMPs, including MT1-MMP (MMP-14), MT2-MMP (MMP-15), and MT3-MMP (MMP-16), as well as for MMP-2 and MMP-9 [25,42]. Next, a NIR uorescent dye, Cy5.5, and PEG 5000-wrapped AuNPs were conjugated to the N-and C-termini of the peptide, respectively. Characterisation tests demonstrated that AuNPs@PEG-Peptide-Cy5.5 had a small hydrated size and homogeneous size distribution, which would allow the NPs to easily penetrate into a plaque through a damaged endothelium or the vasa vasorum. The uorescence intensity of AuNPs@PEG-Peptide-Cy5.5 was negligible in a static state because the signal could only be detected in a narrow range; however, after the activation, the signal could be ampli ed up to several folds, depending on the MMP-14 concentration. The biosafety assessment of AuNPs@PEG-Peptide-Cy5.5 showed its low cytotoxicity and no signi cant injuries to organs such as the lung, heart, liver, and kidney.
MMPs not only participate in the onset and progression of atherosclerosis but also re ect plaque vulnerability [43], and macrophages are the most important source of MMPs in atherosclerosis. Hence, in vitro biological evaluation of the MMP-14-targeting uorescent nanoprobe was performed using the macrophage cell line RAW 264.7. The results showed that MMP-14 was highly expressed in ox-LDLinduced foamy macrophages, at a level of 5·2-fold higher than that in normal macrophages. To verify the association of MMP-14 expression with in ammatory macrophages, we determined changes in the M1 macrophage markers CD68 and F4/80 [44] and the classical M2 macrophage marker CD206 [45] and con rmed that CD68 and F4/80 expression was signi cantly upregulated and that of CD206 was largely suppressed in ox-LDL-induced macrophages. These trends were reversed by the treatment with the nonselective MMP inhibitor GM6001 [46], which con rmed that MMP-14 could potentially serve as a marker for macrophage in ammatory activities. Consistent with the in vitro data, the expression of MMP-14 increased and the hyperactivity of macrophages was aggravated with plaque progression in the murine model of atherosclerosis. Cell and tissue uorescence imaging suggested a targeting ability of AuNPs@PEG-Peptide-Cy5.5 towards MMP-14, as the area of nanoprobe binding overlapped with that of MMP-14 expression. In vivo, a distinctive uorescence signal reached a peak value within 10 min after probe injection and disappeared at 12 h post-injection. To eliminate interference from other tissues and con rm that the uorescent signal was indeed associated with the right carotid artery, wherein plaques were induced, the right carotid artery was immediately isolated after AuNPs@PEG-Peptide-Cy5.5 injection, and ex vivo studies showed that the probe was preferentially accumulated in the plaque on the right artery wall. Biodistribution studies indicated that this nanoprobe had a short circulation time and could be rapidly cleared through the liver, kidney, and intestine, which lowered the risk of possible injury to organs and tissues.
There are some limitations to this study. First, the AuNPs@PEG-Peptide-Cy5.5 nanoprobe was used in a single imaging mode and only for the diagnosis of vulnerable plaques. Meanwhile, recently developed probes tend to have dual imaging modalities or be functional in both diagnosis and treatment, such as dual-functional, uorescent-radiolabelled, composite imaging agents [47,48], which can simultaneously detect PET and NIR uorescence signals, or photodynamic/photothermal therapeutic drugs, which are not only used for imaging but also provide thermotherapy to kill cancer cells [49,50]. Second, the implantation of the silicone cuff might have induced perivascular in ammation, thus causing falsepositive results, such as non-atherosclerosis-related background signals for imaging. More importantly, human vessels cannot be fully mimicked by mouse vessels, in which cardio-cerebrovascular events are rare, for unknown reasons. Third, in vitro uorescent imaging of plaque samples showed that the internal elastic membrane could also be stained by the nanoprobe, and further studies are needed to elucidate the mechanism of this off-target effect and make improvements.

Conclusions
In summary, in this study, we investigated and validated the mechanism by which MMP-14, a promising biomarker of M1 macrophages, contributes to the progression and rupture of atherosclerotic plaques. Then, we successfully developed an MMP-14-speci c, enzyme-activated NIR nanoprobe, AuNPs@PEG-Peptide-Cy5.5, for rapid and non-destructive in vivo visualisation of vulnerable atherosclerotic plaques, which would be a great help to make clinical decision.

Consent for publication
Not applicable.

Availability of data and materials
The data that support the ndings of this study are available in the manuscript and from the corresponding author upon request (drwangdonghai@sdu.edu.cn).

Competing interests
The authors have no con icts of interest to declare.

Funding
Financial support was provided by the Transverse Research Project Foundation of China (No. 21520011691806). The funders had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication. The corresponding authors have full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.  Schematic illustration of AuNPs@PEG-Peptide-Cy5.5 as a NIR uorescence imaging probe. The probe includes three parts: uorescence dye Cy5.5, MMP-14 substrate (peptide as a linker), and PEG5000 wrapped nanogold particles (AuNPs). Cy5.5 is covalently linked to the MMP-14 substrate and is incorporated into the surface of AuNPs, for which the uorescence will be quenched owing to surface energy transfer properties. However, upon proteolytic cleavage with MMP-14, Cy5.5 could emit strong near-infrared uorescence because the quenching effect was reduced as the distance increased.