Characterizing Cathepsin Activity and Macrophage Subtypes in Excised Human Carotid Plaques
Background and Purpose—Atherosclerosis is a leading cause of mortality worldwide, contributing to both strokes and heart attacks. Macrophages are key players in atherogenesis, promoting vascular inflammation and arterial remodeling through cysteine cathepsin proteases. We used a cathepsin-targeted activity-based probe in human carotid plaque to assess its diagnostic potential and evaluate macrophage subtypes ex vivo.
Methods—Carotid plaque specimens surgically removed during endarterectomy from 62 patients (age range, 38% female, 28% symptomatic) were graded pathologically as either stable (Grade 1) or unstable (Grade 2 or 3). A cathepsin activity-based probe was used to quantify individual cathepsins in plaque tissue and macrophage subtypes.
Results—Cathepsin B and S activities were increased in unstable carotid plaques. They were quantified using the probe to biochemically investigate individual cathepsins (Cathepsin B and S: 0.97 and 0.90 for grade 3 versus 0.51 and 0.59 for grade 1; P=0.006 and P=0.03 arbitrary units (AU), respectively). Higher cathepsin activity was observed in carotid plaques from symptomatic patients (Cathepsin B and S: 0.65 and 0.77 for asymptomatic, 0.99 and 1.17 for symptomatic; P=0.008 and P=0.005 AU, respectively). Additionally, it was demonstrated that M2 macrophages from unstable plaques express cathepsin activity 5-fold higher than M2 macrophages from stable plaques (25.52 versus 5.22; P=0.008 AU).
Conclusions—Targeting cathepsin activity in human carotid plaques may present a novel diagnostic tool for characterizing high-risk plaques. Novel cathepsin activity patterns within plaques and macrophage subpopulations suggest their involvement in the transition to active disease.
The formation and progression of atherosclerotic plaques is characterized by accumulation of oxidized lipids and inflammatory cells covered by a fibrous cap of smooth muscle cells and collagen-rich matrix.1 Most severe clinical events result from plaque rupture, which exposes prothrombotic material in the plaque to the blood and causes intraluminal thrombus formation. Plaques vulnerable to rupture are characterized by a thin fibrous cap, a large lipid/necrotic core, positive vascular remodeling, and substantial infiltration of inflammatory cells.2 In particular, macrophages have emerged as a key cellular element of atherosclerotic plaque pathogenesis and pose a significant risk for causing plaque rupture.3 Recently, studies have observed heterogeneous subsets of macrophages, such as M1 and M2, in human atherosclerotic lesions.4
Progressive necrosis of foam cells causes secretion of several factors that sustain the inflammatory processes and produce tissue-degrading proteases, such as matrix metalloproteinases and elastases.5 Plaque destabilization is attributed mostly to extracellular matrix (ECM) degradation (collagenolysis elastinolysis) by multiple ECM degrading enzymes, such as matrix metalloproteinases, serine hydrolyses, and cysteine proteases. Cysteine proteases, including cathepsins B, L, and S, display profound ECM catabolic activity in vitro and were found to be overexpressed in activated macrophages and advanced atherosclerotic lesions.6,7
Cathepsin B, L, and S play key roles in macrophage function, with their mRNA expression and activity increasing significantly in the M2 phenotype.8 Additionally, M1 macrophages show elevated cathepsin S and L mRNA levels.9 Moreover, both mRNA and protein levels of cathepsin B, L, and S are overexpressed in murine atherosclerotic plaques. Similarly, in advanced human atherosclerotic plaques, cathepsin B and S were found overexpressed within macrophages, endothelial cells, and fibrous cap smooth muscle cells; they were shown to be sensitive to proinflammatory stimuli.9,10 In normal human arteries, however, cathepsin B and S protein are poorly expressed.11,12 Furthermore, the cathepsin activity was shown to be involved in inflammation in IL1β processing and to contribute to the degradation of the ECM in the fibrous cap and destabilization of the plaque.11,13–15 Therefore, these proteases serve as markers for plaque inflammation and vulnerability.
Given the previously identified role of cathepsins in plaque progression, compounds that can report on cathepsin activity in atherosclerotic plaques should allow studies of their involvement in plaque progression and distinguish between vulnerable and stable plaques.
We hypothesize that a fluorescent cathepsin activity-based probe (ABP), GB123, which we have previously developed,16,17 may enable identification of individual cathepsin activity in unstable plaques and the contribution of different macrophage subpopulations to this pathological state. ABPs are small molecules that covalently bind to a target enzyme in an activity-dependent manner. Because ABPs covalently bind to their enzyme targets, they remain at the active site, allowing biochemical analysis of the target enzymes.
Detailed methods are described in methods sectionin the online-only Data Supplement.
Carotid plaque specimens were collected from patients who underwent carotid endarterectomy after providing informed consent. The carotid endarterectomy specimens were collected from a total of 62 patients classified as with or without a history of cerebrovascular symptoms; details in SI in the online-only Data Supplement.
Pathological Plaque Classification
Human carotid plaque samples were collected immediately after endarterectomy. Samples were divided into 3 parts that were flash frozen in liquid nitrogen, cryosectioned, and histopathologically analyzed. All plaques were classified into 3 groups on the basis of their morphology as described in Li et al.18 Grade 1 to 3 plaques were defined by fibrous cap thickness and integrity, lipid pool, leukocyte infiltration, internal elastic lamina, necrotic core, cholesterol crystal clefts, internal plaque hemorrhage, thrombosis, and neovascularization. Plaques were referred to as stable if graded 1, vulnerable if graded 2, and ruptured if graded 3. Unstable plaques were defined as Grades 2 and 3; details in SII in the online-only Data Supplement.
Biochemical Analysis of Cathepsin Activity in Human Carotid Lysates
Frozen tissue sample lysates (n=25) were treated with cathepsin B, L, and S ABP, GB12316 (probe structure in Figure I in the online-only Data Supplement), and inhibitor controls were run. SDS-PAGE was fluorescently scanned; details in SIII in the online-only Data Supplement.
Measuring Cathepsin Activity in Carotid Plaques
Fresh carotid specimens were washed, frozen in optimal cutting temperature compound (OCT), and cryosectioned. Serial sections of tissue samples were washed and treated with GB123 for 1 hour; inhibitor controls were applied. Samples were fixed, washed, and analyzed by florescence confocal microscopy; details in SIV in the online-only Data Supplement.
Ex Vivo Labeling of Plaque Samples With Fluorescent ABPs
Freshly resected carotid samples (n=19) were incubated with GB123 and then washed for 4 or 24 hours. Cathepsin inhibitor controls were applied. Samples were imaged using an IVIS Kinetic. Plaque serial cryosections stained with cathepsin B, cathepsin S, and CD68 antibodies were analyzed by fluorescent microscopy; details in SV and SVI in the online-only Data Supplement.
Isolation of Cells From Carotid Plaque for Fluorescence-Activated Cell Sorter Analysis
Freshly resected human carotid samples (n=12) were incubated with GB123 for 1 hour and then embedded in OCT. After collecting stained plaques, the OCT-embedded specimens were thawed and washed. Single cell suspension was analyzed by fluorescence-activated cell sorter using CD45, CD86, or CD206 (mannose receptor) antibodies and GB123 labeling; details in SVII in the online-only Data Supplement.
Data were expressed as mean±SD values. Statistical analyses were done in Graphpad prism 6. Kolmogorov–Smirnov and Shapiro–Wilk tests were used to determine whether the data follows the normal distribution, and comparisons were done using 2-tailed Student’s t test and 1-way analysis of variance with false discovery rate correction set to 0.05 and Dunnett’s test, respectively, to account for multiple comparisons. Receiver–operator characteristic curve computed nonparametrically with 95% confidence intervals using the null hypothesis that the area under the curve equals 0.50. All experiments were performed in a double-blind fashion.
Cathepsin Activity in Human Atherosclerotic Plaques
Plaques were collected from carotid endarterectomy patients between the ages of 57 and 87 years, 38% female, and 28% symptomatic. A representative computed tomographic angiogram before surgery is shown in Figure IIA in the online-only Data Supplement. Carotid plaques were divided into 3 stages according to their morphological alterations18 as described in the Methods section, with representative images shown in Figure IIB in the online-only Data Supplement. ABP GB12316, which labels active cathepsin B, L, and S enzymes, was applied to measure cathepsin activity in the different types of carotid plaques (n=25; for probe structures, see Figure I in the online-only Data Supplement). Importantly, using the novel fluorescently labeled cathepsin ABP conjugates allowed us to perform a specific biochemical analysis of the activity of individual cathepsins by running SDS-PAGE of tissue lysates. A representative fluorescent scan of a gel containing lysates of various lesion grades is shown in Figure 1A. The intensity of the bands at 31 and 26 kDa corresponds to the molecular weights of cathepsins B and S, respectively, as confirmed by immunoprecipitation (Figure III in the online-only Data Supplement). Importantly, it was found that the surrounding tissue near the plaque had lower cathepsin activity than the plaques themselves in almost all cases of Grade 2 and 3 (Figure 1A). Cathepsin L activity was found in low levels in lysates of excised plaques, probably because of the labeling of lysates in which cathepsin L activity is often damaged by the lysing process itself. It was, therefore, decided to pursue with cathepsin B and S only.
Quantification of the cathepsin B and S activities, as measured by the gel analysis, showed good correlation of cathepsin activity with plaque severity (Figure 1B). For cathepsin B, activity increased by ≈2-fold in unstable Grade 2 and 3 plaques in comparison to stable Grade 1 plaques: 0.90 and 0.97 versus 0.51, respectively, AU (P<0.02). For cathepsin S, activity increased close to 1.5-fold, but only in Grade 3 plaques: 0.59 versus 0.90 AU (P=0.03; Figure 1B). These findings were further corroborated by Western blot analysis of cathepsin protein expression in the human carotid samples (Figure IV in the online-only Data Supplement). Using the GB123 probe, cathepsin B and S activity was compared in carotid plaques from symptomatic (n=7) and asymptomatic patients (n=18). When cathepsin B and S activities were quantified, it was found that carotid plaques from symptomatic patients displayed significantly higher cathepsin activity (≈1.5 fold) compared with plaques from asymptomatic patients (cathepsin B: 0.65 versus 0.99 AU, P=0.008, and cathepsin S: 0.77 versus 1.17 AU, P=0.005; Figure 1C). Using receiver–operator characteristic curves, further testing was performed to assay whether cathepsin B and S activity could differentiate between symptomatic and asymptomatic patients. The area under the curve was found to be 0.84 and 0.88 for cathepsin B and cathepsin S, respectively, thus indicating the ability of GB123 to discriminate symptomatic patients (Figure 1D). In addition, the quantified percentage of symptomatic and asymptomatic patients in each Grade showed an increase in symptoms as the grade progressed (Figure V in the online-only Data Supplement).
Cathepsin Protein Expression Correlates With Human Carotid Plaque Grade
The magnitude of cathepsins B and S protein expression was measured in serial sections (n=21) of human carotid plaque specimens. Similar to the activity, differential expression patterns between cathepsin B and cathepsin S were found. Although cathepsin B increased in both Grade 2 and Grade 3 plaques, cathepsin S only increased in Grade 3 plaques. As expected, expression of both cathepsins B and S was found to colocalize with CD68-positive macrophages, in both Grade 2 and 3 plaques (Figure IIC in the online-only Data Supplement). To further validate the findings, gene expression profiling, GEO (GSE41571), of macrophages obtained from stable or ruptured human carotid plaques was performed. Significant enrichment was found in genes involved in lysosomal activity, including cathepsins B and S in ruptured versus stable plaques (Figure VI in the online-only Data Supplement). Taken together, these data demonstrate that expression and activity of cathepsins B and S are progressive with atherosclerotic plaque severity.
Cathepsin ABP for Ex Vivo Plaque Imaging
To establish our ABP as an accurate diagnostic imaging tool for atherosclerosis, fresh dissected human carotid samples (n=19) were cut into pieces (Figure 2A, left and right), then incubated with GB123 applying inhibitor controls (Figure 2B). As the GB123 probe is constitutively fluorescent, even when unbound, no significant difference in fluorescent signals was detected between plaques pretreated with or without an inhibitor in the 4-hour washout time. However, after 24 hours of washout, a significant difference in probe retention was observed, suggesting that extended washing is required to remove the unbound nonquenched probe. Specifically, there was an increase in the Cy5 signal relative to inhibited samples in unstable plaques (Grade 2 and 3 combined; n=12) compared with stable plaques (Grade 1; n=7; 1.95 versus 1.19 AU; P=0.008; Figure 2B and 2C).
Cathepsin Activity Localizes to Macrophages in Human Carotid Plaques
Although our quantitative analyses established a strong correlation between cathepsins B and S activities and vascular pathology, their potential cell source remains to be identified. Using confocal microscopy, it was shown that macrophages were responsible for increased cathepsin activity. First, it was determined that GB123 renders a reliable estimate of cathepsin activity by treating tissue sections (n=12) with the probe in the presence or absence of a selective inhibitor (Figure 3A). A clear and specific signal was detected in frozen sections, and inhibitor pretreatment almost completely abrogated the GB123 fluorescent signal. Then, to investigate probe penetration, whole tissue samples were treated with GB123, washed for 24 hours, after which fresh frozen OCT samples were prepared and sectioned. Here, the fluorescent images similarly demonstrated a strong Cy5 signal within the plaque (Figure 3B). In addition, the immunoreactive CD68, cathepsins B, and cathepsin S colocalized with the majority of the Cy5 signal, showing that the GB123 signal reflects macrophage-derived cathepsin activity and also that local macrophages exhibit strong cathepsin activity in human carotid plaques (Figure 3C).
Plaque macrophages exhibited 2 distinct phenotypes commonly referred to as M1 (proinflammatory) or M2 (anti-inflammatory). Although both cell types contribute to ECM remodeling, their relative contribution to atherogenesis in human subjects is incompletely understood. For this reason, it was decided to explore the phenotype of macrophages with high cathepsin activity using our novel GB123 ABP. To understand whether the macrophages within the plaques were principally M1- or M2-polarized macrophages, a phenotypic characterization of the relative abundance of macrophage subtypes in atherosclerotic plaques was performed. Tissue samples treated with GB123 were analyzed for the costimulatory molecule CD86 as a marker for M1 macrophages and the mannose receptor CD206 (MR) as a marker for M2 macrophages. Interestingly, from the CD45+ gated monocyte population, the CD86+ cells were found in similar percentages in both stable and unstable plaques, 86±4% and 88±2% respectively. In contrast, an increase in MR+ macrophages (so-called M2) was found in unstable plaques (46±1.5% versus 59±3%; P=0.017; Figure 4A). Most dramatically, when cathepsin activity in the CD86+ and MR+ cells isolated from the human carotid plaques were measured, there was a significant increase in MR+ cells labeled with GB123 observed in unstable compared with stable plaques (25±4% versus 5.0±0.8%; P=0.008). However, there was almost no difference in cathepsin activity of CD86+ cells in stable and unstable plaques (25±6% versus 28±6%, P>0.05; Figure 4B), indicating that a unique population of M2 macrophages may be responsible for the upregulation in cathepsin activity within the unstable plaque. Finally, flow cytometric analysis was used to compare the mean fluorescent intensity of Cy5 signal in CD86+ versus MR+ cell populations. Confirming our previous results, there was an increase observed in the ratio of Cy5 mean fluorescent intensity for MR+/CD86+ in unstable compared with stable plaques (1.6±0.1 versus 0.8±0.2 AU; P=0.015; Figure 4C). Representative dot plots are showed in Figure VII in the online-only Data Supplement.
The application of the small molecular cathepsin ABP GB123 for characterizing plaque stability has been demonstrated. A significant correlation between cathepsin activity and plaque grade was found, with the highest cathepsin activity observed in advanced plaques. Although cathepsin S expression was highly elevated only in ruptured Grade 3 plaques, cathepsin B was found to be highly elevated in unstable Grade 2 and 3 plaques relative to Grade 1, thus suggesting potentially different roles for these 2 cathepsins during plaque progression. Furthermore, there was a significant correlation between patients with symptoms and high cathepsin activity. The above results indicate that cathepsins B and S are involved in, and can be detected in, plaque progression and rupture and correlate with symptomatic events.
Cathepsin L was reported to be highly expressed in rupture-prone regions of atheromatous in 3 major cell types encompassing vascular lesions, such as endothelial cells, smooth muscle cells, and macrophages.19,20 In this study, however, cathepsin L activity was not reported because, initially, plaques were lysed, thus leading to a significant loss in cathepsin L activity (G. Blum and M. Bogyo, unpublished data, 2005). Thus, the significant contribution of cathepsin L in vascular injury should not be excluded.
Because macrophages play a key role in inflammatory cardiovascular diseases, there have been several classes of probes that have been described for imaging macrophages as a way to detect vulnerable plaques. These include iodinated, fluorescent, fluorinated, and magnetic nanoparticles (based on an iron-oxide core) that have been shown to accumulate in macrophages in plaques and in injured myocardium. These probes were shown to be useful as contrast reagents for optical, magnetic resonance imaging, positron emission tomography, or computed tomography modalities.2,5,21–29 Furthermore, macrophage plasticity is the focus of several recent studies.30 Different subsets of macrophages, such as M1 and M2, have been observed in human atherosclerotic lesions.4 Fluorescence-activated cell sorter analysis showed upregulation in levels of MR+ (M2) cells in unstable compared with stable plaques, whereas CD86+ (M1) macrophages were present in higher but similar quantities in stable and unstable plaques. By using the ABPs, it was shown that a specific phenotype of MR+ macrophages of unstable plaques have a 5-fold increase in cathepsin activity compared with stable plaque MR+ cells, whereas the cathepsin activity in CD86+ cells remained unchanged. This increase in MR+ cathepsin activity of unstable plaques suggests that the major source of elevated cathepsin activities in these plaques originates from MR+ cells, implying that they are involved in plaque rupture. In stable plaques, however, the cathepsin activity is most likely generated by CD86+ cells because they were found in substantially higher levels than MR+ cells. Taken together, these data highlight the contribution of cathepsins B and S to the clinical sequelae of atherosclerosis and suggests that M2 macrophages play a key role in advanced vascular lesions.
The fluorescently labeled probe described here is unique because it enables ex vivo imaging of the vulnerable plaque with multiple utilities: The probe forms a covalent bond with its targets, therefore, allowing direct analysis of cathepsin activity in tissues using biochemical methods, fluorescent microscopy, and fluorescence-activated cell sorter. Thus, this probe can be used to analyze activity of multiple cathepsins simultaneously, as well as determine the location of these activities within the plaque tissue and the cell type expressing the cathepsin activity.
As the degree of stenosis provides an imperfect estimate to the risk of clinical events (eg, myocardial infarction, stroke), there is a need for methods that can distinguish between vulnerable and stable atherosclerotic plaques. GB123 has been used previously in vivo for noninvasive molecular imaging of cancer and has shown high potential to be applied for preclinical imaging of cathepsin activity in atherosclerotic mice models.16 Similarly, high molecular weight polymeric fluorescently quenched cathepsin substrate probes have been previously described for cancer and atherosclerosis applications and found to be extremely useful.
With high molecular weight polymeric fluorescently quenched cathepsin substrate probes being translated to clinical use with the aid of endoscopic instruments,11,24,31,32 we consider fluorescent ABP methodology to be attractive for several reasons. First, our probe is a small molecule that can freely penetrate cells and, therefore, target a larger pool of active cellular enzymes, resulting in higher signals than high molecular weight substrate probes, as shown in cancer models.33 Second, the ABP allows for separation of individual cathepsin activities using biochemical methods. In addition, the ABP GB123 methodology can be translated to clinical use by changing the fluorescent tag with a 64Copper label that can be detected in greater depth with positron emission tomography imaging, as shown previously.34 Combining the potential clinical use of ABPs with present results correlating cathepsin activity with plaque grade and patient symptoms may allow accurate identification of individuals at risk for atherosclerosis-associated complications. The fluorescent cathepsin ABP described here enables assessment of vulnerable plaques and serves as an attractive tool for atherosclerosis research, with potential for both preclinical and clinical plaque imaging.
The molecular tools presented in this article provide significant advancements in atherosclerosis research, providing a novel diagnostic method and a novel activity pattern of cathepsin in the M2 macrophages of unstable plaques.
We thank Dr Israel Gotsman, Dr Seth Salpeter, and Prof Jonathan Leor for article reviewing.
Sources of Funding
We thank the United States–Israel Binational Science Foundation (BSF) (2009010 and 2011480 to G.B. and M.V.M.) and the Israel Ministry of Health Chief Scientist (3-00000-7035 to G.B.) for funding.
Dr McConnell has the following competing interest: cardiovascular magnetic resonance imaging research grant from GE Healthcare and preclinical research grant from Tiara Pharmaceuticals. Dr McConnell is currently an employee of Verily Life Sciences (Mountain View, CA) on partial leave from Stanford University. The other authors report no conflicts.
Current address for H.K.: Department of Cardiovascular Medicine, Tokyo Medical and Dental University, Bunkyo-ku, Tokyo, Japan.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.115.011573/-/DC1.
- Received September 21, 2015.
- Revision received January 20, 2016.
- Accepted January 21, 2016.
- © 2016 American Heart Association, Inc.
- Reddy VY,
- Zhang QY,
- Weiss SJ.
- Gocheva V,
- Wang HW,
- Gadea BB,
- Shree T,
- Hunter KE,
- Garfall AL,
- et al
- Liu J,
- Sukhova GK,
- Sun JS,
- Xu WH,
- Libby P,
- Shi GP.
- Chen J,
- Tung CH,
- Mahmood U,
- Ntziachristos V,
- Gyurko R,
- Fishman MC,
- et al
- Kitamoto S,
- Sukhova GK,
- Sun J,
- Yang M,
- Libby P,
- Love V,
- et al
- Amirbekian V,
- Lipinski MJ,
- Briley-Saebo KC,
- Amirbekian S,
- Aguinaldo JG,
- Weinreb DB,
- et al
- Trivedi RA,
- Mallawarachi C,
- U-King-Im JM,
- Graves MJ,
- Horsley J,
- Goddard MJ,
- et al
- Nahrendorf M,
- Jaffer FA,
- Kelly KA,
- Sosnovik DE,
- Aikawa E,
- Libby P,
- et al
- Chen W,
- Jarzyna PA,
- van Tilborg GA,
- Nguyen VA,
- Cormode DP,
- Klink A,
- et al
- Laitinen I,
- Saraste A,
- Weidl E,
- Poethko T,
- Weber AW,
- Nekolla SG,
- et al
- Michalska M,
- Machtoub L,
- Manthey HD,
- Bauer E,
- Herold V,
- Krohne G,
- et al
- Calfon MA,
- Rosenthal A,
- Mallas G,
- Mauskapf A,
- Nudelman RN,
- Ntziachristos V,
- et al