Imaging of MMP Activity in Postischemic Cardiac ... - ACS Publications

Mar 18, 2014 - Department of Biomedical Engineering, Eindhoven University of Technology, 5656 ... Center for Imaging Research and Education (CIRE), 56...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/molecularpharmaceutics

Imaging of MMP Activity in Postischemic Cardiac Remodeling Using Radiolabeled MMP-2/9 Activatable Peptide Probes Sander M.J. van Duijnhoven,†,‡ Marc S. Robillard,‡,§ Sven Hermann,∥ Michael T. Kuhlmann,∥ Michael Schaf̈ ers,∥ Klaas Nicolay,†,‡ and Holger Grüll*,†,‡,§ †

Department of Biomedical Engineering, Eindhoven University of Technology, 5656 Eindhoven, The Netherlands Center for Imaging Research and Education (CIRE), 5656 Eindhoven, The Netherlands § Department of Minimally Invasive Healthcare, Philips Research, 5656 Eindhoven, The Netherlands ∥ European Institute for Molecular Imaging (EIMI), D-48149 Münster, Germany ‡

S Supporting Information *

ABSTRACT: The noninvasive imaging of matrix metalloproteinases (MMPs) activity in postischemic myocardial tissue holds great promise to predict cardiac function postmyocardial infarction. Consequently, development of MMP specific molecular imaging probes for noninvasive visualization and quantification of MMP activity is of great interest. A novel MMP imaging strategy is based on activatable cell-penetrating peptide probes (ACPP) that are sensitive to the proteolytic activity of MMP-2 and -9. The MMP-mediated activation of these ACPPs drives probe accumulation at the target site. The aim of this study was the development and characterization of radiolabeled MMP-2/9 sensitive ACPPs to assess MMP activity in myocardial remodeling in vivo. Specifically, a short and long-circulating MMP activatable cell-penetrating imaging probe (ACPP and Alb-ACPP, respectively; the latter is an ACPP modified with an albumin binding ligand that prolongs blood clearance) were successfully synthesized and radiolabeled. Subsequently, their biodistributions were determined in vivo in a Swiss mouse model of myocardial infarction. Both peptide probes showed a significantly higher uptake in infarcted myocardium compared to remote myocardium. The biodistribution for dual-isotope radiolabeled probes, which allowed us to discriminate between uncleaved ACPP and activated ACPP, showed increased retention of activated ACPP and activated Alb-ACPP in infarcted myocardium compared to remote myocardium. The enhanced retention correlated to gelatinase levels determined by gelatin zymography, whereas no correlation was found for the negative control: an MMP-2/9 insensitive non-ACPP. In conclusion, radiolabeled MMP sensitive ACPP probes enable to assess MMP activity in the course of remodeling post-myocardial infarction in vivo. Future research should evaluate the feasibility and the predictive value of the ACPP strategy for assessing MMP activity as a main player in postinfarction myocardial remodeling in vivo. KEYWORDS: myocardial infarction, matrix metalloproteinases, molecular imaging, activatable imaging probes, radioisotopes, myocardial remodeling

■

INTRODUCTION Myocardial infarction (MI) commonly leads to maladaptive remodeling of the myocardial extracellular matrix (ECM), resulting in congestive heart failure.1 The matrix metalloproteinases (MMPs), a family of extracellular matrix degrading enzymes, are involved in this remodeling process.2,3 For example, the gelatinases MMP-2 and MMP-9 are increased in activity and abundance and act as crucial modulators in adverse cardiac remodeling.4 MMP-9 predominantly plays a role in the early wound healing and inflammation response shortly after MI, whereas MMP-2 is mainly involved in the cardiac ECM remodeling phases.2 In a recent study, it was illustrated that a noninvasive methodology to image the temporal and spatial levels of MMP-2/9 in myocardial remodeling holds great promise to predict cardiac function post-MI in a pig model.5 The MMP level was visualized by © 2014 American Chemical Society

single photon emission computed tomography (SPECT) using a radiolabeled MMP-inhibitor, which displayed a 1:1 probetarget binding, and this showed correlation to the cardiac function measured by magnetic resonance imaging (MRI). We hypothesized that imaging of the MMP catalytic activity using activatable probes, which enable signal amplification, will further improve the monitoring of the myocardial remodeling process. Chen et al. have demonstrated that a quenched nearinfrared fluorescent (NIRF) probe can successfully be used to monitor MMP-2/9 activity in MI-mice.6 However, in the human heart, NIRF imaging would only be feasible by Received: Revised: Accepted: Published: 1415

September 25, 2013 March 18, 2014 March 18, 2014 March 18, 2014 dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

Figure 1. Mechanism and structure of radiolabeled MMP-2/9 activatable ACPPs. (A) The cell-penetrating property of a polycationic peptide is masked by a polyanionic peptide. Cleavage of the linker by MMP-2/9 releases the polycationic cell-penetrating peptide, which will transfer its radionuclide cargo (in blue) across the cell membrane. The polyanionic peptide and its radionuclide cargo (in red) will be cleared from the target tissue. The ratio of the radionuclide cargoes (blue-over-red) is a measure for local ACPP activation. Peptide structure of (B) 177Lu/125I-ACPP and (C) 177Lu/125I-Alb-ACPP. The MMP-2/9 cleavage site is indicated by the arrow. D-amino acids are denoted in lower case. X, SHPP, e*, and DOCA represent 6-aminohexanoic acid, 3-(4-hydroxyphenyl)propionic acid, γ-glutamic acid, and deoxycholic acid, respectively. Molecular structure of AlbACPP is shown in Supporting Information Supplemental Figure 1.

invasively monitoring the activation of the fluorescent probe by real-time intravascular catheter detection.7 Here, we tested a dual-isotope radiolabeled activatable cell-penetrating molecular imaging probe (ACPP) that is sensitive to the proteolytic activity of MMP-2 and -9 (Figure 1).8 These ACPP probes consists of an MMP-2/9 substrate inserted in between a polyanionic inhibitory peptide and a polycationic cell penetrating peptide (CPP). Cleavage of the substrate, which demonstrated specific sensitivity toward MMP-2/9 but did not cross-react with other members of the MMP-family, releases the polycationic from the polyanionic domain, thereby triggering tissue adhesion and subsequent retention of the polycationic peptide domain at the target site.8,9 We further developed and assessed the in vivo behavior of a long circulating ACPP (Alb-ACPP) that consists of the ACPP moiety conjugated to the albumin ligand deoxycholic acid (Figure 1c). In these probes, the polycationic cell-penetrating peptide domain and the polyanionic peptide domain were labeled with the orthogonal radioisotopes 177Lu (energy of the γ-ray emitted by isotope is Eγ = 208 keV) and 125I (Eγ = 35 keV), respectively. A large ratio of 177 Lu over 125I would indicate cleavage of the probe and subsequent retention of the 177Lu-radiolabeled CPP, thereby facilitating a direct read-out of the relative level of MMP activity. We and others previously showed the usefulness of 177 Lu and 125I in dual-isotope preclinical biodistribution studies.8,10,11 The hybrid dual isotope labeled probes were administrated in all experiments described in this study, except for one imaging study using autoradiography where the single 177 Lu radiolabeled ACPP probe was employed.

ACPP is reported in the Supporting Information (Supplemental Figures 1−2). Animal Studies. All animal procedures were approved by the ethical review committee of the Maastricht University Hospital (The Netherlands) and were performed according to the principles of laboratory animal care (NIH publication 85− 23, revised 1985),12 and the Dutch national law “Wet op Dierproeven” (Stb 1985, 336). Male Swiss mice (body weight >25g, Charles River Laboratories) were housed in an enriched environment under standard conditions: 21−23 °C, 50−60% humidity, and 12 h light−dark cycles for >1 week. Food and water were freely available. Myocardial Infarction (MI) Model. MI was induced by permanent ligation of the left anterior descending coronary artery (LAD) using published procedures.13 In short, animals were subcutaneously injected with buprenorphine (0.1 mg/kg) and 30 min later anesthetized with isoflurane. Animals were intubated and ventilated with 100% oxygen with a rodent respirator. After left thoracotomy between ribs four and five, the LAD was ligated with a 6-0 prolene suture. The chest and skin were closed with 5-0 silk sutures. The animal’s temperature was continuously measured rectally and maintained at 36.5− 37.5 °C during surgery. Sham-operated animals underwent the same procedure, except that the 6-0 prolene suture was passed through the myocardium without ligating the LAD and served as controls. After surgery, animals were allowed to recover at 30 °C overnight. Ten days (for ACPP) or eleven days (for AlbACPP) post-MI, the animals were used for biodistribution studies. Radiolabeling for In Vivo Studies. 177Lu/125I-ACPP and 177 Lu/125I-non-ACPP. 177LuCl3 in 0.05 M HCl (4.0 μL, 20.0 MBq) was mixed with ACPP or non-ACPP in Milli-Q water (3.7 μL, 25 nmol) and metal-free 0.9% NaCl (242.3 μL) for 20 min, at 600 rpm and 90 °C. 125I (PerkinElmer) in phosphate buffered saline (PBS) pH 7.4 (7.0 μL, 12.5 MBq) was mixed with ACPP or non-ACPP in Milli-Q water (3.7 μL, 25 nmol) and PBS (239.3 μL) in a Pierce iodination tube (Thermo

■

METHODS Probe Synthesis. Peptides Ac-y-e9-x-PLGLAG-r9-x-k(DOTA)-NH2 (ACPP), Ac-y-e9-x-LALGPG-r9-x-k(DOTA)NH2 (non-ACPP), and Ac-LAG-r9-x-k(DOTA)-NH2 (CPP) were synthesized as previously reported.8 The synthetic procedure and analysis for peptides Alb-ACPP and non-Alb1416

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

Fischer Scientific) for 20 min, at 600 rpm and 23 °C, and transferred to a siliconized 1.5 mL tube. 177 Lu-CPP. 177LuCl3 in 0.05 M HCl (2.0 μL, 10.0 MBq) was mixed with CPP in Milli-Q water (3.0 uL, 25 nmol) and metalfree 0.9% NaCl (245 μL) for 20 min at 600 rpm and 90 °C. 177 Lu/125I-Alb-ACPP and 177Lu/125I-non-Alb-ACPP. 177LuCl3 in 0.05 M HCl (3.0 μL, 15.0 MBq) was mixed with Alb-ACPP and non-Alb-ACPP in Milli-Q water (2.4 μL, 15 nmol) and metal-free 0.9% NaCl (144.6 μL) for 20 min at 600 rpm and 90 °C. 125I (PerkinElmer) in 1 mM NaOH (3.0 μL, 6.0 MBq) was mixed with Alb-ACPP and non-Alb-ACPP in Milli-Q water (2.4 μL, 15 nmol) and PBS (144.6 μL) in an iodination tube for 20 min at 600 rpm and 23 °C and transferred to a siliconized 1.5mL tube. The 177Lu and 125I labeling yields were determined by radioTLC using iTLC-SG strips (Pall) eluted with 200 mM EDTA in 0.9% NaCl and 20 mM citric acid at pH 5.2, respectively, imaged on a phosphor imager (FLA-7000, Fujifilm), and quantified with AIDA Image Analyzer software. Analytical radio-HPLC was carried out on an Agilent 1100 system equipped with a C18 Eclipse XBD-column (length = 150 mm, diameter = 4.6 mm, particle size = 5.0 μm) and a Gabi radioactive detector (Raytest). The radiochemical purities were 95% or higher, and typically at least 96% (Supporting Information Supplemental Figures 3−5). 177 Lu-labeled ACPP/non-ACPP/Alb-ACPP/non-Alb-ACPP was mixed with 125 I-labeled ACPP/non-ACPP/Alb-ACPP/non-Alb-ACPP in a 1:1 molar ratio. Blood Kinetic Measurements and Biodistribution Experiments. Blood kinetic and biodistribution experiments were performed on MI-mice (n = 3−5) or sham-mice (n = 4) by i.v. injection of 177Lu/125I-ACPP (10 nmol/100 μL, ca. 4.0 MBq 177Lu, ca. 2.5 MBq 125I), 177Lu/125I-non-ACPP (10 nmol/ 100 μL, ca. 4.0 MBq 177Lu, ca. 2.5 MBq 125I), 177Lu-CPP (10 nmol/100 μL, ca. 4.0 MBq), 177Lu/125I-Alb-ACPP (10 nmol/ 100 μL, ca. 5.0 MBq 177Lu, ca. 2.0 MBq 125I), or 177Lu/125I-nonAlb-ACPP (10 nmol/100 μL, ca. 5.0 MBq 177Lu, ca. 2.0 MBq 125 I). At selected time points (2, 5, 10, 30, 60, and 150 min for ACPP (n = 3) and 4, 10, 30, 60, 180, and 360 min for the AlbACPP (n = 3)), blood samples were withdrawn from the vena saphena, weighed, and diluted to 1 mL with Milli-Q water. The mice were anesthetized with isoflurane 5, 20, or 50 h after i.v. injection, subjected to 2% (w/v) Evans Blue i.v. injection (50 mg/kg), and sacrificed 2 min later by cervical dislocation. Organs and tissues of interest were harvested, weighed, and measured along with standards in a γ-counter (Wizard 1480; PerkinElmer) using a dual-isotope protocol (10−80 keV and 155−380 keV energy windows for 125I and 177Lu, respectively) with a cross-contamination correction to determine the injected dose per gram (%ID/g). The blood clearance data was fitted to a two-phase exponential decay, y = A·exp(−k1t) + B·exp(−k2t) + y∞, using GraphPad Prism. The area under the curve (AUC) was determined for the fitted 177Lu blood kinetic profiles for all probes. The AUC half-lives were subsequently derived using MatLab. The free volume of distribution per probe was calculated using the formula VD = dose/(mean body weight·C0) [L/kg]. Because of the ∼100-fold lower concentration of AlbACPP (∼3−4 μM) relative to mouse serum albumin (0.4 mM)14 and the presence of multiple binding sites per protein molecule,15 it was assumed that the binding capacity of albumin was far from saturated. Therefore, the unbound probe fraction, α, was calculated from α = 1/(1 + CaKa), where Ca is the total concentration of albumin and Ka is the equilibrium binding

constant.16 The hearts were cooled to 4 °C, cut in 1 mm slices from apex to base. Randomly heart slices were incubated in 1.0% (w/v) 2,3,5-triphenyltetrazolium chloride (TTC) in PBS at 37 °C, 300 rpm for 20 min to prove consistency between in vivo Evans Blue and ex vivo TTC staining (Supporting Information Supplemental Figure 6). Infarct, border, and remote areas were isolated based on Evans Blue staining, and sample radioactivity was counted in a γ-counter (Wizard 1480, PerkinElmer) along with standards to determine the % injected dose per gram (%ID/g). Specifically, a dual-isotope protocol (10−80 keV and 155−380 keV energy windows for 125I and 177 Lu, respectively) with cross-contamination correction was used. TTC-stained slices were excluded for γ-counting due to wash-out of the radiolabeled probes during the incubation. The 125 I-activity in stomachs and thyroids was measured to exclude 125 I-ACPP in vivo dehalogenation (Supporting Information Supplemental Tables 1−3). Sample Preparation and Tissue Homogenates. After γcounting, tissue samples were homogenized in 20 volumes of 50 mM Tris, 200 mM NaCl, 10 mM CaCl2, and 10 μM ZnCl2 at pH 7.5 at 4 °C at 25 Hz for 2 × 5 min, subsequently mixed at 5 Hz for 2 × 15 min using a tissue lyser (Qiagen), and then centrifuged at 5000 rpm for 5 min at 4 °C. The supernatants were aliquoted and stored at −80 °C until zymography analysis. Typically, 14 μL of tissue supernatant (from 0.7 mg of tissue homogenate) was used per analysis. Gelatin Zymography. Samples were analyzed on 10% SDS-PAGE gel containing 0.1% (w/v) gelatin (Biorad). MMP2 (Calbiochem) (0.21 ng) was loaded as an internal standard used to normalize activities between gels. Electrophoresis was performed at 150 V for 1.5 h, after which the gels were washed with Milli-Q water, incubated for 3 × 20 min in 2.5% Triton-X (50 rpm) to remove SDS, washed with Milli-Q water, and incubated in 50 mM Tris, 200 mM NaCl, 5 mM CaCl2, 0.1% (w/v) NaN3, and 0.02% (w/v) Brij-35 at pH 7.6 and 37 °C for 2 days. Gels were stained for 2 h with 0.25% Coomassie Blue in 60% (v/v) Milli-Q water, 30% (v/v) methanol, 10% (v/v) acetic acid, and destained for >24h with 67.5% (v/v) Milli-Q water, 25% (v/v) methanol, 7.5% (v/v) acetic acid. Gelatinatic activity showed up as clear bands against a dark background. Zymograms were imaged (Epson Perfection V700 Photo scanner) and band intensities were quantified using ImageJ. Autoradiography and Histology. MI-mice (n = 4) were injected i.v. with 177Lu-ACPP (10 nmol/100 μL, ca. 20.0 MBq 177 Lu). The mice were anesthetized with isoflurane 20 h after i.v. injection, and tissue perfusion fixation was performed with 4% PFA in sterile PBS. The hearts were excised and incubated in 4% PFA at 4 °C for 2 h, washed with ice-cold PBS, and incubated in 15% sucrose in PBS for 16−48 h. The hearts were cut in half (either in coronal or transversal direction) and sliced into sections with a cryomicrotome (repetitions of 1 × 40 μm and 10 × 10 μm sections for autoradiography and histology, respectively). Autoradiography was performed for a maximum of 80 h using a MicroImager (BioSpace Lab). The radioactivity within each region of interest (ROI) was quantified by ImageJ. Azan staining was performed using standard procedures. Antibodies against MMP-2 and Mac3 (for macrophages) were used for immunostaining. Statistical Methods. Quantitative data were expressed as mean ± SD. Standard one-way or repeated measures ANOVA, with Bonferroni’s post hoc testing, was used for multiple group comparisons. Single groups were compared with the one-tailed 1417

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

Figure 2. Biodistribution of 10 nmol 177Lu/125I-ACPP and 10 nmol 177Lu/125I-Alb-ACPP. (A) 177Lu-biodistribution of ACPP in infarct, border, and remote myocardium of MI-mice (n = 3) at 5 h postinjection. (B) 177Lu-biodistribution of ACPP in infarct, border, and remote myocardium of MImice (n = 5) and in hearts of sham-mice (n = 4) at 20 h postinjection. (C) 177Lu-biodistribution of Alb-ACPP in infarct, border, and remote myocardium of MI-mice (n = 5) at 50 h postinjection. Data are mean ± SD (D, E, F) Correlation between 177Lu-biodistribution of ACPP at 5 h (n = 3) and at 20 h (n = 5), of Alb-ACPP at 50 h (n = 5) postinjection, and of gelatinase levels in heart tissue of mice 10 days post-MI. Pearson correlation coefficients are r = 0.92, r = 0.78, and r = 0.82, respectively.

unpaired Welch’s t test.17 Groups with p < 0.05 were considered significantly different. Linear regression was performed to assess linear relationship between gelatinase expression levels and 177Lu-uptake for ACPP and Alb-ACPP and between gelatinase expression and 177Lu-to-125I ratios for ACPP and Alb-ACPP. GraphPad Prism was used for all statistical calculations.

■

RESULTS In Vivo Distribution of ACPP and Non-ACPP in MIand Sham-Mice. Following successful synthesis8 and radiolabeling (Supporting Information Supplemental Figures 3−5), 177 Lu/125I-ACPP and its negative control 177Lu/125I-non-ACPP, containing a scrambled linker (LALGPG), were studied in mice 10 days after induction of a myocardial infarction (MI) or 10 days after sham-surgery. Biodistribution was determined at both 5 h and 20 h postinjection. In this paragraph, we will focus on the 177Lu biodistribution of the hybrid dual isotope ACPP probe. In the next section, the 177Lu-to-125I ratios for the ACPP probes will be discussed. The absolute 177Lu-uptake of 177 Lu/125I-ACPP probe was significantly higher in infarct than in remote myocardium (Figure 2A and B, p < 0.001), uptake in the latter was similar to that in hearts of sham-mice. In addition, imaging of the biodistribution of single 177Lu radiolabeled ACPP in MI-hearts using autoradiography revealed a 6-fold higher probe retention in infarcted myocardium compared to remote myocardium 20 h postinjection (Figure 3). The whole body biodistribution data (5 h and 20 h) for 177Lu/125I-ACPP in MI- and sham-mice and 177Lu/125I-non-ACPP in MI-mice can be found in Supporting Information Supplemental Tables 1−3. Importantly, 177Lu showed a significantly higher uptake in

Figure 3. Histology and autoradiography of a coronal section of the mouse heart 20 h postinjection of 10 nmol 177Lu-ACPP 10 days postMI. (A) Representative azan staining showing the infarct scar in blue and remote myocardium in red. (B) Autoradiography showing enhanced uptake of 177Lu-ACPP in the infarct zone in an adjacent section.

infarcts for ACPP (1.43 ± 0.21%ID/g) compared to the negative control non-ACPP (0.27 ± 0.13% ID/g, p < 0.001). Though for sham-mice, gelatinatic activity in the heart was nearly undetectable by gelatin zymography analysis, gelatinase levels were highly increased in infarcted myocardium and slightly elevated in remote myocardium in MI-mice (Figure 4A). Quantification of gelatin zymograms revealed a significantly higher gelatinase level in infarcted heart tissue compared to remote heart tissue and sham-hearts (Figure 4B, p < 0.001). A strong correlation was found between the gelatinase level and 177 Lu-ACPP uptake in infarcted, border zone, and remote myocardium (Figure 2D and E, p < 0.001). 1418

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

uptake of the entire probe and specific activation. In MI-mice, which received 177Lu/125I-ACPP, infarcted heart tissue showed a significantly higher 177Lu-to-125I ratio compared to remote myocardium at both 5 h and 20 h postinjection (Figure 5A and B, Supporting Information Supplemental Tables 1 and 3, p < 0.05). This points toward a higher degree of probe activation in the infarct zone. Furthermore, the 177Lu-to-125I ratio was significantly elevated in infarcted myocardium compared to skeletal muscle of MI-mice and heart tissue of sham-mice (p < 0.05). In animals injected with the negative control 177Lu/125Inon-ACPP, the 177Lu-to-125I ratio was near unity and similar in infarcted, remote myocardium, and skeletal muscle (Supporting Information Supplemental Table 2, p > 0.05). A significant linear dependency was found between the 177Lu-to-125I ratios for ACPP and gelatinase levels in MI-hearts (Figure 5D and E, p < 0.001), whereas no correlation was found for non-ACPP (Figure 6).

Figure 4. (A) Representative gelatin zymogram of heart homogenates. Lanes 1 and 4 = infarcted myocardium, 2 and 5 = border myocardium, 3 and 6 = remote myocardium, and 7 = active MMP-2 (0.21 ng). (B) MMP-2 expression in MI-hearts (n = 8) and sham-hearts (n = 4). Data are mean ± SD. 177 Lu-to-125I Ratios for ACPP and Non-ACPP. The administration of dual-isotope labeled 177Lu/125I-ACPP and 177 Lu/125I-non-ACPP probes enabled us to study the in vivo probe activation in more detail and to discriminate between

Figure 6. Correlation between gelatinase level and 177Lu-to-125I ratios for 177Lu/125I-non-ACPP (n = 4) in MI-mice 20 h postinjection. Pearson correlation coefficient r = 0.03.

Figure 5. Biodistribution of 10 nmol 177Lu/125I-ACPP and 10 nmol 177Lu/125I-Alb-ACPP. (A) 177Lu-to-125I- ratios of ACPP in infarct, border, and remote myocardium of MI-mice (n = 3) at 5 h postinjection. (B) 177Lu-to-125I- ratios of ACPP in infarct, border, and remote myocardium of MImice (n = 5) and in hearts of sham-mice (n = 4) at 20 h postinjection. (C) 177Lu-to-125I- ratios of Alb-ACPP in infarct, border, and remote myocardium of MI-mice (n = 5) at 50 h postinjection. Data are mean ± SD (D, E, F) Correlation between 177Lu-to-125I- ratios of ACPP at 5 h (n = 3) and at 20 h (n = 5), of Alb-ACPP at 50 h (n = 5) postinjection, and of gelatinase levels in heart tissue of mice 10 days post-MI. Pearson correlation coefficients are r = 0.80, r = 0.60, and r = 0.77, respectively. 1419

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

Figure 7. Blood kinetic profiles of 10 nmol (A) 177Lu-ACPP (n = 3) and 177Lu-Alb-ACPP (n = 3), (B) 177Lu/125I-ACPP (n = 3), and (C) 177Lu/125IAlb-ACPP (n = 3) in MI-mice. The data are presented as mean %ID/g ± SD. Blood kinetics were fitted with a two-phase exponential decay function (see Supporting Information Supplemental Table 4).

to-125I ratios for 177Lu/125I-Alb-ACPP revealed a significantly higher ratio in the infarcted area compared to remote areas of the heart (Figure 5C, p < 0.05). Furthermore, 177Lu uptake and 177 Lu/125I ratios for 177Lu/125I-Alb-ACPP correlated significantly with gelatinase levels (Figures 2F and 5F), suggesting MMP mediated local activation of 177Lu/125I-Alb-ACPP in infarcted heart tissue. The 177Lu-to-125I ratios for 177Lu/125IAlb-ACPP are several folds lower than observed for 177Lu/125IACPP in both infarct and remote. The latter observation is extensively discussed in the Supporting Information.

Long Circulating ACPP. In a previous study, we showed that the ACPP probe has a relatively short circulation time in nude mice.8 To increase the exposure time of ACPP to the target proteases, we synthesized a long circulating MMP-2/9 sensitive ACPP (Supporting Information Supplemental Figures 1 and 2). This low molecular weight albumin binding ACPP probe (Alb-ACPP) contains an deoxycholic acid (DOCA) moiety that can reversibly bind to albumin with a Ka of 7.4 × 10 4 M −1 , allowing the extravasation of the unbound compound.14 Probe design considerations, the molecular structure, the synthesis, and the radiolabeling of Alb-ACPP and the negative control non-Alb-ACPP are discussed in the Supporting Information (Supplemental Figures 1−5). In vitro experiments demonstrated a slightly lower rate of MMP-2mediated activation of radiolabeled Alb-ACPP compared to the parent ACPP probe in the presence of albumin (Supporting Information Supplemental Figure 7). Blood Kinetics of ACPP and Alb-ACPP. The blood kinetics of 177Lu/125I-ACPP and 177Lu/125I-Alb-ACPP showed a significant increase in circulation time for Alb-ACPP compared to ACPP in MI-mice (p < 0.01, Figure 7A). Specifically, the probe half-lives (based on 177Lu counting) were 22 min for 177 Lu/125 I-ACPP and 67 min for 177 Lu/ 125 I-Alb-ACPP (Supporting Information Supplemental Table 4). Both probes displayed a biphasic elimination from the circulation. The calculated volume of distribution (VD) was 0.16 L/kg for ACPP and 0.12 L/kg for Alb-ACPP. This suggests an extracellular extravascular biodistribution for ACPP, as was earlier observed in tumor-bearing mice, 9 and a restricted extracellular extravascular biodistribution for Alb-ACPP.18,19 The blood kinetics of 177Lu and 125I for both ACPP and Alb-ACPP follow different trends, with a ∼2.5-fold faster clearance of 177Lu compared to 125I, which would suggest probe activation and tissue entrapment of the activated 177Lu-CPP domain (Figure 7). Biodistribution of 177Lu/125I-Alb-ACPP. Biodistribution was determined at 50 h postinjection, at which time the blood presence of Alb-ACPP was negligible. A significantly higher uptake of 177Lu-Alb-ACPP was observed in infarcted areas of the heart compared to remote regions (p < 0.01, Figure 2C, Supporting Information Supplemental Table 5), which was comparable to ACPP at 20 h postinjection. Correspondingly, the infarct-to-remote ratios for both probes were in the same range (6.5 ± 1.1 vs 6.6 ± 1.3 for 177Lu-ACPP and 177Lu-AlbACPP, respectively). Furthermore, 177Lu-Alb-ACPP uptake in infarcted myocardium was significantly higher compared to the negative control non-Alb-ACPP (p < 0.01, Supporting Information Supplemental Table 5). Analysis of the 177Lu-

■

DISCUSSION Noninvasive methodologies to image the temporal and spatial levels of MMP-2/9 in myocardial remodeling hold great promise to predict cardiac function post-MI as has been shown by Sinusas and co-workers.6 In that study, a radiolabeled MMPtargeted tracer was studied in pigs with a myocardial infarction and a 3−4 fold higher tracer uptake was observed in infarcted myocardium compared to remote myocardium. To extend the current toolbox for in vivo monitoring of MMP levels, which is mostly based on probes showing a 1:1 probe-target binding fashion, we focused on the detection of MMP catalytic activity using activatable probes that facilitate an amplification of the imaging signal. Here, we employed a dual-isotope radiolabeled peptide probe that is cleavable by MMP-2/9 and subsequently is trapped in tissue.8 The probe consists of an MMP-2/9 substrate inserted in between a polycationic cell-penetrating peptide (CPP) and a polyanionic peptide that were labeled with the orthogonal radioisotopes 177Lu and 125I, respectively, which have earlier been successfully employed in dual-isotope approaches for dissection-based in vivo biodistribution studies.8,10,11 The cell-penetrating property of the polycationic peptide was masked by electrostatic interactions with the polyanionic domain, preventing tissue association of the probe. Activation of the probe should result in efficient tissue retention of the 177Lu-labeled polycationic peptide, whereas the 125I labeled counterpart is cleared by the kidneys. Therefore, a large ratio of 177Lu over 125I would indicate cleavage of the probe and subsequent retention of the 177Lu-radiolabeled CPP, thereby facilitating a direct read-out of the relative level of MMP activity. We choose to conjugate 125I to the polyanionic noncell penetrating domain to circumvent potential intracellularassociated dehalogenation and subsequent cellular release of 125 I. Previously, we indeed showed that such a radioiodinated ACPP probe was stable and not subject to significant dehalogenation, as evidenced by the low amount of 125I measured in thyroids and the stomach.8 In this study, we also 1420

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

found low 125I-activity in stomachs and thyroids (Supporting Information Supplemental Tables 1−3). In this study, we determined the biodistribution of 177Lu/125I radiolabeled MMP-2/9 sensitive activatable CPPs (ACPPs) in a mouse model of myocardial infarction to detect elevated tissue levels of MMP activity. Future development of single labeled ACCP probes with 111In, 99mTc, or 68Ga radiolabels may allow potential clinical translation toward single photon emission computed tomography (SPECT) or positron emission tomography (PET), respectively. Here, both whole-body biodistribution analyses using γ-counting (Figure 2) and imaging of 177Lu-ACPP biodistribution in MI-hearts using autoradiography (Figure 3) showed a 6-fold increased uptake of the peptide-based probe in infarcted regions of the heart compared to remote areas 20 h postinjection. At 5 h postinjection an impressive 10-fold higher uptake in infarcted myocardium was observed compared to remote myocardium, which is more than a 2-fold larger difference than reported for (nonactivatable) radiolabeled MMP targeted tracers in pig and mice 1 week post-MI6,20 and for a fluorescent activatable probe in mice 1 week post-MI.7 We also found increased gelatinase levels in infarcted myocardium compared to remote myocardium and the gelatinase levels showed a linear dependency with both the 177Lu uptake as well as 177Lu-to-125I ratios for 177 Lu/125I-ACPP at 5 h and 20 h postinjection (Figure 5). For the MMP-2/9 insensitive negative control 177Lu/125I-nonACPP, the 177Lu-to-125I ratios were close to unity and similar in both infarct and remote. These data indicate MMP-specific activation of the cleavable probe in the MI-hearts. Next to gelatin zymography, histology was performed on MIhearts to assess levels of MMP-2 activity and macrophages. Qualitative histological analysis of MMP-2 and macrophage levels was, however, not performed due to reduced quality of structure preservation. Nevertheless, the histological data seem to indicate elevated levels of active MMP-2 and macrophages in areas of increased 177Lu-ACPP uptake (Supporting Information Supplemental Figure 8). The biodistribution data for 5 h and 20 h postinjection suggest a probe washout from MI-hearts overtime. As mentioned earlier, we found linear dependencies between the MMP-2 level on one hand and probe uptake (177Lu uptake) and probe activation (177Lu-to-125I ratio) on the other hand at both time points. Stronger Pearson correlation coefficients (r) were observed at 5 h postinjection due to higher variation for the 20 h postinjection biodistribution data. The observed probe clearance from the target tissue overtime most likely has contributed to this larger variation. The circulation time of ACPP peptide probe was efficiently enhanced via in vivo protein binding using the albumin ligand deoxycholic acid. The elongated blood circulation time of this new probe, Alb-ACPP, is most likely caused by a reduction in renal clearance due to albumin binding. It was calculated that 96.7% of Alb-ACPP was noncovalently bound to albumin directly after its intravenous administration. Despite the increased circulation time, biodistribution in MI-hearts did not show improvements for Alb-ACPP compared to ACPP. In this respect, the restricted extravasation from the circulation of Alb-ACPP, as suggested by its volume of distribution, may have reduced the effective concentration of this probe in the remodeling myocardium, resulting in an overall lower amount of activated probes. Nevertheless, the 177Lu-to-125I ratio for AlbACPP in infarcted myocardium was higher compared to remote myocardium and correlated with gelatinase levels.

We recently showed that ACPP and the positive control peptide CPP are characterized by a similar biodistribution pattern in mice bearing subcutaneous MMP-positive HT-1080 tumors, suggesting that ACPP activation in the vascular compartment was the main mechanism for the observed elevated probe uptake in this tumor model 8. Here, we found a significant higher 177Lu infarct-to-remote (p < 0.01) and 177Lu infarct-to-muscle (p < 0.05) ratio for ACPP compared to the positive control CPP (Supporting Information Supplemental Figure 9, Supplemental Tables 1 and 2), indicating a significant degree of infarct-specific activation of ACPP. Interestingly, no significant difference in gelatinase levels was observed for HT1080 tumors and myocardial infarcted tissue by gelatin zymography.8 However, it was recently demonstrated that a significant part of the MMPs in the tumor microenvironment of subcutaneous HT-1080 tumors is inhibited, most likely by the family of tissue inhibitors of metalloproteinases (TIMPs), thereby significantly slowing down the cleavage rate of MMP probes,21 which may contribute to the differences observed in probe activation in tumor tissue and postischemic remodeling myocardium. Despite infarct-specific activation, the biodistribution data of 177 Lu/125I-ACPP in MI-mice also suggest, in agreement with our earlier tumor study,8 a degree of background ACPP activation in the vasculature indicated by 177Lu-to-125I ratios >1 in various tissues that show low levels of MMP expression, for example, sham-hearts (177Lu/125I ratio = 5.0 ± 1.5) and remote myocardium (177Lu/125I ratio = 6.2 ± 0.8). The activation of the probe in the vasculature also resulted in a strong accumulation of the activated domain, 177Lu-CPP, in the liver (20.4 ± 4.2%ID/g), an organ adjacent to the heart, which may complicate imaging applications of ACPP probes in the heart. An improvement of the target-to-non target ratios may be achieved with an ACPP-based cleavable imaging probe that is activated in the infarct without activation in the vasculature. Such a probe may display reduced background liver uptake as observed for non-ACPP, an ACPP analog that showed no detectable cleavage in the vasculature, whereas the local infarctspecific activation may still result in relatively high target uptake. Next to the gelatinases MMP-2 and -9, other MMPs such as neutrophil collagenase MMP-8 and membrane-type MMP-14 play also pivotal roles in adverse cardiac remodeling after MI.22−24 As MMP-8 is present in the circulation, it is not suitable for the ACPP imaging concept, which requires local activation of a probe with subsequent accumulation of the probe in the lesion. MMP-14 is a cardiac-restricted cell membrane-associated biomarker and is for that reason a particular attractive target for future ACPP imaging probes. Other research groups already have undertaken attempts to develop an MMP-14 sensitive ACPP (ACPP-14) by incorporating the peptide SGRIGFLRTA as MMP-14 substrate.25 Although in vitro ACPP-14 activation was observed in MMP-14 positive cell cultures, an MMP-14 negative control cell line also activated the probe, and therefore, MMP-14 selectivity could not be demonstrated. Therefore, we are currently designing new MMP-14 sensitive ACPPs following a similar concept as shown here.

■

CONCLUSION MMP activity in cardiac remodeling of postischemic myocardium was successfully detected in vivo using radiolabeled MMP activatable cell-penetrating peptide probes (ACPPs). We 1421

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

metalloproteinase-9 transcription after myocardial infarction. Am. J. Physiol. Heart Circ. Physiol. 2006, 291, H2216−H2228. (5) Sahul, Z. H.; Mukherjee, R.; Song, J.; McAteer, J.; Stroud, R. E.; Dione, D. P.; Staib, L.; Papademetris, X.; Dobrucki, L. W.; Duncan, J. S.; Spinale, F. G.; Sinusas, A. J. Targeted imaging of spatial and temporal variation of matrix metalloproteinase activity in porcine model of post-infarct remodeling: relationship to myocardial dysfunction. Circ. Cardiovasc. Imaging 2011, 4, 381−391. (6) Chen, J.; Tung, C. H.; Allport, J. R.; Chen, S.; Weissleder, R.; Huang, P. L. Near-infrared fluorescent imaging of matrix metalloproteinase activity after myocardial infarction. Circulation 2005, 111, 1800−1805. (7) Jaffer, F. A.; Vinegoni, C.; John, M. C.; Aikawa, E.; Gold, H. K.; Finn, A. V.; Ntziachristos, V.; Libby, P.; Weissleder, R. Real-time catheter molecular sensing of inflammation in proteolytically active atherosclerosis. Circulation 2008, 118, 1802−1809. (8) van Duijnhoven, S. M. J.; Robillard, M. S.; Nicolay, K.; Grüll, H. Tumor targeting of MMP-2/9 activatable cell-penetrating imaging probes is caused by tumor-independent activation. J. Nucl. Med. 2011, 52, 279−286. (9) Jiang, T.; Olson, E. S.; Nguyen, Q. T.; Roy, M.; Jennings, P. A.; Tsien, R. Y. Tumor imaging by means of proteolytic activation of cellpenetrating peptides. Proc. Nat. Acad. Sci. U.S.A. 2004, 101, 17867− 17872. (10) Tijink, B. M.; Laeremans, T.; Budde, M.; Stigter-van Walsum, M.; Dreier, T.; de Haard, H. J.; Leemans, C. R.; van Dongen, G. A. Improved tumor targeting of anti-epidermal growth factor receptor nanobodies through albumin binding: taking advantage of modular nanobody technology. Mol. Cancer. Ther. 2008, 7, 2288−2297. (11) van den Bosch, S. M.; Rossin, R.; Renart Verkerk, P.; ten Hoeve, W.; Janssen, H. M.; Lub, J.; Robillard, M. S. Evaluation of strained alkynes for Cu-free click reaction in live mice. Nucl. Med. Biol. 2013, 40, 415−423. (12) Guide for the Care and Use of Laboratory Animals. NIH, Government Printing Office: Washington, DC, 1985; pp 86−23. (13) Lutgens, E.; Daemen, M.; de Muinck, E. D.; Debets, J.; Leenders, P.; Smits, J. F. Chronic myocardial infarction in the mouse: cardiac structural and functional changes. Cardiovasc. Res. 1999, 41, 586−593. (14) Peeples, E. S.; Schopfer, L. M.; Duysen, E. G.; Spaulding, R.; Voelker, T.; Thompson, C. M.; Lockridge, O. Albumin, a new biomarker of organophosphorus toxicant exposure, identified by mass spectrometry. Toxicol. Sci. 2005, 83, 303−312. (15) Rudman, D.; Kendall, F. E. Bile acid content of human serum. II. The binding of cholanic acids by human plasma proteins. J. Clin. Invest. 1957, 36, 538−542. (16) Elmadhoun, B. M.; Wang, G. Q.; Templeton, J. F.; Burczynski, F. J. Binding of [3H]palmitate to BSA. Am. J. Physiol. Gastrointest. Liver Physiol. 1998, 275, G638−G644. (17) Ruxton, G. D. The unequal variance t-test is an underused alternative to Student’s t-test and the Mann-Whitney U test. Behav. Ecol. 2006, 17, 688−690. (18) Riches, A. C.; Sharp, J. G.; Brynmor Thomas, D.; Vaughan Smith, S. Blood volume determination in the mouse. J. Physiol. 1973, 228, 279−284. (19) Durbin, P. W.; Jeung, N.; Kullgren, B.; Clemons, G. K. Gross composition and plasma and extracellular water volumes of tissues of a reference mouse. Health Phys. 1992, 63, 427−442. (20) Su, H.; Spinale, F. G.; Dobrucki, L. W.; Song, J.; Hua, J.; Sweterlitsch, S.; Dione, N. P.; Cavaliere, P.; Chow, C.; Boruke, B. N.; Hu, X. Y.; Azure, M.; Yalamanchili, P.; Liu, R.; Cheesman, E. H.; Robinson, S.; Edwards, D. S.; Sinusas, A. J. Noninvasive targeted imaging of matrix metalloproteinase activation in a murine model of postinfarction remodeling. Circulation 2005, 112, 3157−3167. (21) Xia, Z.; Xing, Y.; Jeon, J.; Kim, Y. P.; Gall, J.; DragulescuAndrasi, A.; Gambhir, S. S.; Rao, J. Immobilizing reporters for molecular imaging of the extracellular microenvironment in living animals. ACS Chem. Biol. 2011, 6, 1117−1126.

demonstrated a significant contribution of infarct-associated ACPP activation to infarct targeting of ACPP, most likely due to MMPs. An impressive 10-fold higher probe uptake was observed in infarcted tissue compared to remote myocardium. Future research should address whether ACPP can be used for in vivo nuclear imaging of MMPs in postischemic myocardium.

■

ASSOCIATED CONTENT

S Supporting Information *

Alb-ACPP probe design considerations, synthesis, and in vitro results. Radiolabeling results, in vivo biodistribution data, and blood kinetic profiles of all probes and discussion on 177Luto-125I ratios for ACPP versus Alb-ACPP. This information is available free of charge via the Internet at http://pubs.acs.org/.

■

AUTHOR INFORMATION

Corresponding Author

*H. Grüll. E-mail: [email protected]. Telephone: +31 (0) 61 293 2309. Fax: +31 (0) 40 243 2598. Address: Eindhoven University of Technology, Department of Biomedical Engineering, Biomedical NMR High Tech Campus 11, Room WBC2p265, 5656 AE Eindhoven, The Netherlands. Notes

The authors declare the following competing financial interest(s): Marc Robillard and Holger Gruell are employed at Philips Research.

■

ACKNOWLEDGMENTS We thank Iris Verel (Philips Research) for valuable discussions, Caren van Kammen, Leonie Niesen, Peter Leenders, Carlijn van Helvert, Melanie Blonk, Marleen Hendriks (University Hospital Maastricht), and Monique Berben (Philips Research) for their support with the in vivo experiments, and Irmgard Hoppe (European Institute for Molecular Imaging, Münster) for autoradiography and histology experiments. This research was performed within the framework of the Center for Translational Molecular Medicine (CTMM, [www.ctmm.nl]), project “Translational Initiative on Unique and Novel Strategies for Management of Patients with Heart Failure” (TRIUMPH, grant 01C-103) and supported by the Dutch Heart Foundation. The study was partially supported by the core unit “Preclinical Imaging eXperts” (PIX) of the Interdisciplinary Center of Clinical Research, Münster, Germany, and the Collaborative Research Center 656 (SFB, projects C06 and Z02), Münster, Germany.

■

REFERENCES

(1) Deschamps, A. M.; Spinale, F. G. Pathways of matrix metalloproteinase induction in heart failure: Bioactive molecules and transcriptional regulation. Cardiovasc. Res. 2006, 69, 666−676. (2) Vanhoutte, D.; Schellings, M.; Pinto, Y.; Heymans, S. Relevance of matrix metalloproteinases and their inhibitors after myocardial infarction: a temporal and spatial window. Cardiovasc. Res. 2006, 69, 604−613. (3) Zavadzkas, J. A.; Plyler, R. A.; Bouges, S.; Koval, C. N.; Rivers, W. T.; Beck, C. U.; Chang, E. I.; Stroud, R. E.; Mukherjee, R.; Spinale, F. G. Cardiac-restricted overexpression of extracellular matrix metalloproteinase inducer causes myocardial remodeling and dysfunction in aging mice. Am. J. Physiol. Heart Circ. Physiol. 2008, 295, H1394− H1402. (4) Mukherjee, R.; Mingoia, J. T.; Bruce, J. A.; Austin, J. S.; Stroud, R. E.; Escobar, G. P.; McClister, D. M., Jr.; Allen, C. M.; Alfonso-Jaume, M. A.; Fini, M. E.; Lovett, D. H.; Spinale, F. G. Selective spatiotemporal induction of matrix metalloproteinase-2 and matrix 1422

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423

Molecular Pharmaceutics

Article

(22) Fertin, M.; Lemesle, G.; Turkieh, A.; Beseme, O.; Chwastyniak, M.; Amouyel, P.; Bauters, C.; Pinet, F. Serum MMP-9: A novel indicator of left ventricular remodeling and cardiac outcome in patients after acute myocardial infarction. PLoS ONE 2013, 8, e71280. (23) Spinale, F. G.; Mukherjee, R.; Zavadzkas, J. A.; Koval, C. N.; Bouges, S.; Stroud, R. E.; Dobrucki, L. W.; Sinusas, A. J. Cardiac restricted overexpression of membrane type-1 matrix metalloproteinase causes adverse myocardial remodeling following myocardial infarction. J. Biol. Chem. 2010, 285, 30316−30327. (24) Koenig, G. C.; Rowe, R. G.; Day, S. M.; Sabeh, F.; Atkinson, J. J.; Cooke, K. R.; Weiss, S. J. MT1-MMP-dependant remodeling of cardiac extracellular matrix structure and function following myocardial infarction. Am. J. Pathol. 2012, 180, 1863−1878. (25) Watkins, G. A.; Jones, E. F.; Scott Shell, M.; VanBrocklin, H. F.; Pan, M. H.; Hanrahan, S. M.; Feng, J. J.; He, J.; Sounni, N. E.; Dill, K. A.; Contag, C. H.; Coussens, L. M.; Franc, B. L. Development of an optimized activatable MMP-14 targeted SPECT imaging probe. Bioorg. Med. Chem. 2009, 17, 653−659.

1423

dx.doi.org/10.1021/mp400569k | Mol. Pharmaceutics 2014, 11, 1415−1423