|Home | About | Journals | Submit | Contact Us | Français|
The synthesis and validation of a new, highly potent 64Cu-labeled peptide, cFLFLFK-PEG-64Cu, that targets the formyl peptide receptor (FPR) on leukocytes is described. The peptide ligand is an antagonist of the FPR, designed not to elicit a chemotactic response resulting in neutropenia. Evidence for the selective binding of this synthesized ligand to neutrophils is provided. PET imaging properties of the compound was evaluated in a mouse model of lung inflammation.
The FPR-specific peptide, N-cinnamoyl-F-(D)L-F-(D)L-F (cFLFLF), was sequentially conjugated at ω-NH2 of the lysine (K) with a bifunctional polyethylene glycol moiety (PEG, 3.4 kD) and a 2,2',2”,2'”-(1,4,7,10-tetraazacyclododecane-1,4,7,10-tetrayl) tetraacetic acid (DOTA), and finally labeled with [64Cu]CuCl2 to form cFLFLFK-PEG-64Cu. The binding affinity and stimulation potency of the ligand towards human neutrophils were assessed in vitro. Blood kinetic and organ biodistribution properties of the peptide were studied in the mouse. Ten male C57BL/6 mice were used for imaging; four control mice and six administered Klebsiella pneumonia. PET/CT scans were performed to assess the localization properties of the labeled peptide in lungs 18 hr after tracer administration. Lung standardized uptake values (SUVs) were correlated with lung neutrophil activity as measured by myeloperoxidase (MPO) assays. Immunohistochemistry (IHC) was performed to confirm that neutrophils constitute the majority of infiltrating leukocytes in lung tissue 24 hr post Klebsiella exposure.
In vitro binding assays of the compound cFLFLFK-PEG-64Cu to the neutrophil FPR yielded a Kd of 17.7 nM. The functional superoxide stimulation assay exhibited negligible agonist activity of the ligand with respect to neutrophil superoxide production. The pegylated peptide ligand exhibited a blood clearance half-life of 55 ± 8 min. PET imaging 18 hr post tracer administration revealed mean lung SUVs and lung MPO activities for Klebsiella-infected mice that were 5- and 6-fold higher, respectively, compared with control mice. IHC staining confirmed that the cellular infiltrate in lungs of Klebsiella-infected mice was almost exclusively neutrophils at the time of imaging.
This new radiolabeled peptide targeting the FPR binds to neutrophils in vitro and accumulates at sites of inflammation in vivo. This modified peptide may prove to be a useful tool to probe inflammation or injury.
Tissue damage caused by trauma, infection, or other stimuli triggers a complex sequence of events collectively known as the inflammatory response. This response includes the directed migration of neutrophils from circulation to the site of injury with the ultimate goal of killing pathogens. Although inflammation is critical to survival, exaggerated or persistent inflammation causes collateral damage that plays a key role in the progression of many major diseases. Neutrophils are among the first leukocytes to reach a site of injury and they are abundant at focal sites of infection (1). The ability to detect and quantify neutrophilic accumulation could be important not only in locating and identifying inflammatory lesions, but also to facilitate the development and testing of anti-inflammatory agents.
Currently available clinical nuclear imaging probes for targeting and diagnosing inflammatory lesions include 67Ga citrate and 111In or 99mTc leukocytes labeled ex vivo (2). Although each of these agents can yield useful results in specific situations, each possesses significant drawbacks. In general, techniques utilizing in vitro labeling of white blood cells suffer the disadvantage of lengthy, laborious, and potentially hazardous labeling procedures. In contrast, injection of peptides that have a high affinity for surface receptors on leukocytes, have emerged as an attractive option for the in vivo detection of inflammation. Formyl peptides, synthetic analogs of natural bacteria products, have been extensively studied as a possible replacement for current techniques for imaging inflammation. Since peptide probes specifically target leukocytes in vivo, the disadvantages associated with ex vivo laboratory labeling procedures are avoided. Although prior studies have shown promising results detecting leukocyte accumulation in response to inflammatory stimuli with peptide probes in vivo, several problems remain. For example, some of these peptides are potent receptor agonists, with the potential for causing neutrophil activation and neutropenia at high doses (3). Several 99mTc and 111In labeled chemotactic peptide ligands including agonist formyl-methionyl-leucyl-phenylalanine (fMLF) (4, 5) and the antagonist i-Boc-MLF (6) have been investigated for imaging inflammation in vivo. fMLF–based agonist ligands have high affinity for neutrophils; however, they were found to induce chemotaxis, cell adhesion, and degranulation of leukocytes; responses associated with infection and inflammation (7). On the other hand, i-Boc-MLF did not exhibit undesirable neutrophil-activating effects, but exhibited weak binding affinity. An ideal imaging peptide ligand to detect neutrophilic inflammation would exhibit high binding affinity for neutrophils and could be used at doses less then their binding Kd without significantly perturbing their function or influencing their distribution.
The peptide N-cinnamoyl-F-(D)L-F-(D)L-F (cFLFLF) was reported as an antagonist to the neutrophil FPR with a high binding affinity (Kd = 2 nM) (3). However, due to its high hydrophobicity it demonstrated relatively poor target-to-background ratios compared to peptide agonists in imaging focal sites of infection in rabbits. To address this, we modified the peptide by conjugating it with a polyethylene-glycol (PEG, molecular weight = 3.4 kD) to enhance its hydrophilicity (8). The PEG was terminated with DOTA to chelate to 64Cu. In this study, we characterize the binding affinity of this modified peptide to the FPR and determine its functional ability to detect neutrophils. Once we demonstrated the peptide had the desired in vitro properties, in vivo imaging was performed in a mouse model of pulmonary inflammation.
All chemicals obtained commercially were of analytical grade and used without further purification. [64Cu]CuCl2 was purchased from Isotrace, Inc (O'Fallon, MO). The peptide N-cinnamoyl-phe-(D)-Leu-phe-(D)-Leu-phe-lys-CONH2 (cFLFLFKCONH2) was synthesized via solid-phase Fmoc method by Biomolecular Research Facility at the University of Virginia, and the structure was confirmed by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectroscopy, as reported previously (9). Bifunctional t-butoxycarbonyl-protected PEG-succinimidyl ester (t-Boc-PEG-NHS; molecular weight, 3.4 kD) was obtained from Laysan Bio, Inc.. 2,2',2”,2'”-(1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetrayl)tetraacetic acid (DOTA) was obtained from Macrocyclics, Inc.. N-hydroxysulfosuccinimide and 1-ethyl-3-[3-(dimethylamino)-propyl]carbodiimide were purchased from Pierce. All other chemical reagents and solvents were obtained from Sigma-Aldrich. Semipreparative reversed-phase (RP) high-performance liquid chromatography (HPLC) was performed on a Varian system with an ABI Spetroflow 783 ultraviolet detector and a Bioscan Flow Count Radio-HPLC detector with an Apollo C18 RP column (5μ, 250 × 10 mm). The mobile phase changed from 40% Solvent A (0.1% trifluoroacetic acid in 80% water) and 60% Solvent B (0.1% trifluoroacetic acid in 80% aqueous acetonitrile) to 100% solvent B at 30 min at a flow rate 3mL/min. MALDI-TOF mass spectroscopy analysis was performed on samples of peptide products at W.M. Keck biomedical mass spectrometry laboratory at the University of Virginia, and the data were obtained on a Bruker Daltonics system.
Human tumor necrosis factor-alpha (TNF-α) was procured from Peprotech and fMLF was purchased from Sigma. Aliquots of both samples were taken (TNF-α, 10 U/mL, and fMLP, 10 mM) and stored at -20°C. For every assay the solutions were thawed to ambient temperature and freshly diluted with hepatic arterial (HA) buffer before use. Multiscreen high throughput screening (HTS) with glass fiber filter (FC) 96-well plates, type C, with 1.2-μm glass filters were purchased from Millipore. Filtration from 96-well plates was performed under vacuum on Brandel filtration device. The membranes from each well were collected by punching with the Millipore Multiscreen punching instrument. The radioactivity from 64Cu-bound ligand was measured with either Minaxi (Packard), Autogamma 5000 series (Packard), or Wallac 1420 Wizard (Perkin-Elmer), γ-counters. Radioactivity was measured for 1 min per sample and was not corrected for decay.
cFLFLFK-PEG-t-Boc was prepared by incubating a mixture of cFLFLFK(NH2) (10 mg, 10.6 μmol) dissolved in 2 mL of acetonitrile and t-Boc-PEG-NHS (30 mg, 8.8 μmol) dissolved in 2 mL of sodium borate buffer (0.1N, pH 8.5) overnight at 4°C. Removal of volatiles under reduced pressure using rotary evaporator afforded a residue which was dissolved in 2 mL of TFA and left at room temperature for 2 hr to remove t-Boc protecting group. Concentration of the mixture under reduced pressure followed by reconstitution in 50% actontrile:water (2.0 ml) yielded stock solution. This solution was subjected to multiple injections (~5-6) on semipreparative RP-HPLC to collect fractions containing pure cFLFLFK-PEG-NH2 (retention time 18.4 min.). The fractions were concentrated under reduced pressure to yield pure sample, which was further characterized by MALDI-TOF mass spectroscopy. The average molecular weight distribution of cFLFLFK-PEG-NH2 was centered at 4.3 kD and major m/z peaks were observed at 4240, 4284, 4372, and 4416. The average calculated mass was 4331.
cFLFLFK-PEG-NH2 (16.5 mg, 3.8 μmol) was dissolved in 1 mL of H2O and the pH was adjusted to 8.5 with 0.1N NaOH. To this solution was added DOTA-NHS (19 μmol in 20 μL of water), prepared according to a previously reported method (9). The mixture was incubated overnight at 4°C. The solution was subjected to HPLC purification (retention time 16.8 min.) to yield pure cFLFLF-PEG-DOTA (7.6 mg, 43%). Characterization of the peptide by MALDI-TOF revealed an average molecular weight distribution centered at 4.8 kD and major m/z peaks were observed at 4644, 4688, 4776, and 4820. The average calculated mass was 4718.
The radiolabeling was accomplished by addition of 200-800 μCi (7.4 - 29.6 MBq) of 64CuCl2 to 5-20 μg of cFLFLF- PEG-DOTA in 0.1N ammonium acetate (pH 5.5) buffer and the mixture was incubated at 40°C for 30 minutes. The mixture was injected as is for RP-HPLC purification. The column eluate was monitored by UV absorbance at 215 nm and with a gamma detector. The collected product eluted at 17.2 min with a radiochemical yield higher than 95% and the specific activity of 1.1×106 MBq/mmol (yield > 90%). Pure fractions were concentrated under reduced pressure. The radiolabeled peptide was further characterized by comparing its chromatographic properties with non-radioactive copper labeled compound synthesized independently using copper chloride in the same process. Analysis by MALDI-TOF revealed an average MW distribution of about 4.8 kD and major m/z peaks were observed at 4692, 4736, 4778, and 4794. The average calculated mass was 4782, which is in strong agreement with experimental values.
To test for compound stability, we incubated the compound in serum at 37 °C for 1, 3, and 6 hrs. Following incubation, we monitored the compound with HPLC. To determine the partition coefficient of the pegylated and non-pegylated compound, we dissolved about 350 kBq of cFLFLF-PEG76-DOTA-64Cu (or cFLFLF-DOTA-64Cu) in 500μL of water and mixed the solution with 500μL of octanol in an Eppendorf microcentrifuge tube. The tube was sonicated for 10 min and then was centrifuged at 4,000 rpm for 5 min (model Fisher Scientific Marathon Micro-A). Radioactivity was measured in 100μL aliquots of both octanol and water layers in triplicate.
Human neutrophils were prepared from normal heparinized (10 U/mL) venous blood by a one-step Ficoll-Hypaque separation procedure (10, 11), yielding approximately 98% neutrophils; greater than 95% viable as determined with trypan blue containing less than 50 pg·ml-1 of endotoxin. Following separation, neutrophils were washed with Hank's balanced salt solution (HBSS) with heparin (10 U/mL) three times. After the third wash, neutrophils were resuspended in HA Buffer, which was HBSS supplemented with 0.1% human albumin (Bayer Healthcare). Neutrophil experiments were completed in HA Buffer.
Freshly isolated human neutrophils (4×106 cells/mL) were treated with 10 U/mL of TNF-alpha (Peprotech) twenty minutes prior to binding studies and transferred to a 96 well plate (Multiscreen® HTS FC by Millipore, Billerica, MA. 1.2 μ glass filter type C, 50.0 μL,~ 2.0 × 105 cells/well). Saturation assays were carried out using eight different concentrations of cFLFLF-PEG-64Cu (specific activity = 139 mCi/μg or 0.89 μCi/mmol) ranging from 0.001 to 100 nM. Neutrophils were incubated with the radioligand at 25 °C for 90 minutes to obtain total binding. Following incubation, 96 well plates were filtered rapidly under vacuum using Brandel filtration device (Brandel Inc. Gaithersburg, MD), washed three times with cold Tris-Mg buffer (-5 °C, 10mM, 150 μL each time/well) to remove the unbound radioligand, and dried under vacuum. The membranes from each well were collected by Millipore multiscreen punching instrument (Billerica, MA). The bound radioactivity remaining on the membranes was measured in a gamma counter. Specific binding was calculated as the difference between total binding and nonspecific binding. Non-specific binding was assessed using the highest concentration of radiolabeled ligand used in the binding experiment following pre-incubation with cold compound (100 μM of cFLFLF-PEG-Cu). Binding parameters (Kd and Bmax values) were calculated using PRISM 4.0 (GraphPad).
The biological activity of cFLFLF-PEG-Cu or fMLP was assessed by measuring the stimulated release of superoxide by neutrophils after exposure to a range of concentrations. The neutrophil oxidative activity (luminol-enhanced chemiluminescence) was measured using a microtitre polymorphonuclear (PMN) chemiluminescence assay (11). Activated neutrophils emit light from unstable high-energy oxygen species produced by the plasma membrane associated NADPH oxidase, and release myeloperoxidase from primary granules. The light signal from activated neutrophils can be enhanced by the addition of luminol to the samples. Luminol-enhanced emission of light is stimulated by singlet oxygen, a reactive oxygen species, dependent on both the production of superoxide and mobilization of myeloperoxidase (12).
To prime the PMNs, purified cells (2×106/ml) were incubated in a water bath (37°C) for 15 min with rhTNF (10 U/ml). Following priming, aliquots of the PMNs were transferred to a microtitre plate (white walled clear bottom 96 well tissue culture plates) containing luminol (100 μM) and (0.0001-10 μM) of cFLFLF-PEG-Cu or fMLF. Peak stimulated chemiluminescence was determined with a Victor 1420 Multilabel Counter set for chemiluminescence mode using Wallac Workstation software. Sigmoidal dose-response curves for fMLP and cFLFLF-PEG-Cu stimulation of PMN oxidative activity were fit using PRISM 4.0 (GraphPad). EC50 values were derived from concentration-response curves using PRISM software. We compared relative agonist potency (EC50) of cFLFLF-PEG-Cu to fMLF on PMN oxidative activity.
Klebsiella pneumoniae strain 43816, serotype 2 (American Type Culture Collection) was grown in trypticase soy broth (TSB) overnight, then subcultured for 2 hr to log phase growth. Following extensive rinsing, bacteria were diluted in sterile normal saline for inoculation. C57BL/6 mice (male, 8-10 weeks old, Charles River) were inoculated by oropharyngeal aspiration of 50 μl of bacterial suspension (approximately 3×105 CFU) under light inhalational anesthesia with methoxyflurane. The size of the inoculum was quantitated by plating serial dilutions on MacConkey agar plates and counting colony-forming units (CFU) after overnight incubation. Mice showed signs of moderate illness 18-36 hr after inoculation, when imaging was performed.
Distribution of radioactivity in the body was determined in both control (n=4) and Klebsiella-infected (n=6) mice 18 hr post injection of cFLFLF- PEG-64Cu. After taking a single blood sample from the tail vein, mice were euthanized by deep halothane anesthesia. The pulmonary circulation was flushed with 3 mL of sterile normal saline via the right ventricle and the following organs and tissues were removed and washed: heart, lungs, muscle, bone, liver, kidney, spleen, small intestine, and stomach. The dissected tissues were placed in a pre-weighed vial and later assayed in a gamma well counter. The measured radioactivity for each sample was decay corrected back to the time of tracer injection. Biodistribution values are expressed as a percentage of the injected activity (%ID) and normalized for body and organ/tissue mass (13).
Blood kinetics of cFLFLF-PEG-64Cu was studied in three control mice. Approximately 50 μL of blood from the contralateral tail vein was collected in capillary tubes at 5, 15, 30, 60, 120, and 180 minutes after tracer injection (0.37 – 0.74 MBq). The capillary tubes were placed in a vial which was pre- and post-weighed. Activity in each blood sample was measured in a gamma counter, standardized for injected dose and animal body weight, and expressed as percent injected dose per gram of blood (%ID/g).
To estimate the number of intrapulmonary neutrophils, myeloperoxidase assays were performed. Immediately after being imaged, mice were euthanized by deep halothane anesthesia and their pulmonary circulation flushed with 3 mL of sterile normal saline via the right ventricle. The lungs were removed and snap frozen at -80°C until later assayed. Lungs were weighed and placed in homogenization buffer [hexadecyltrimethylammonium bromide], and homogenized followed by sonication and centrifugation. Five microliters of supernatant was added to assay buffer (o-dianisidine hydrochloride in potassium phosphate) in a 96 well plate and optical density kinetic measurements at wavelength 490 were made using a μQuant spectrophotometer. MPO activity is reported as change in OD/min/mg lung tissue.
Immunohistochemical (IHC) analysis was performed on harvested lung tissue 42 hrs post Klebsiella administration, which matches the time point of imaging post infection (24 hrs plus an additional 18 hrs of tracer clearance time). IHC was used to assess the relative amount and distribution of neutrophils compared to macrophages in the lungs of control versus infected mice. Prior to removal, the pulmonary circulation was flushed with saline via the right ventricle to eliminate non-adherent white blood cells. Lungs were then inflated with formalin to distend the alveolar spaces uniformly. The trachea was cannulated and 10% phosphate-buffered formalin infused at a pressure of 25 cm H20. After fixation, the lung was dissected coronally in the plane of the mainstem bronchus. Adjacent histological sections (3 μm thick) were specifically stained for either neutrophils with a monoclonal rat anti-mouse neutrophil IgG (MCA771G; Serotec) or for macrophages with anti MAC-2 IgG (ACL8942P; Accurate). Stained cells were observed under a light microscope (Microphot, Nikon, LRI Instruments AB, Tokyo, Japan).
24 hr after administration of Klebsiella pneumoniae, cFLFLF-PEG-64Cu (100 ~ 150 μCi) (3.7-5.5 MBq) in 200 μl of saline was injected via the tail vein. Lung SUVs were measured at several time points post injection and fit to a mono-exponential curve, allowing for the calculation of ligand clearance in the control and infected lung. This analysis provides us with an estimate of the time window post injection for which the signal difference between control and infected lungs is maximized.
For accurate image co-registration, mice were placed in the prone position in a custom designed portable imaging tray, facilitating precise positioning between scanners. Anesthesia (1-2% isoflurane in oxygen) was delivered throughout the duration of imaging. Micro-X-ray computed tomography (14) images were acquired using a scanner developed in-house. Following CT acquisition, mice were transported to the PET scanner (Focus F-120, Siemens) and scanned for ~25 min. CT images were reconstructed with a 3D filtered back-projection algorithm using the COBRA software (Exxim, Inc.). The reconstructed pixel size was 0.15 × 0.15 × 0.15 mm on a 320 × 320 × 384 image matrix. Using MicroPET Manager (Siemens, version 22.214.171.124), PET data was reconstructed using the OSEM3D/MAP algorithm (zoom factor = 2.164). The reconstructed pixel size was 0.28 × 0.28 × 0.79 mm on a 128 × 128 × 95 image matrix. All microPET images were corrected for decay, but not attenuation.
CT-PET image co-registration was performed using ASIPRO (Siemens) and a transformation matrix previously obtained with an imaging phantom. To characterize the accumulation of the tracer in lungs, region-of-interest (ROI) analysis was performed. CT images were used to visualize lung boundaries and guide the placement of lung ROIs, which were drawn manually. 10 ± 2 contiguous transaxial lung ROIs were drawn to cover the entire lung volume. Lung ROIs were transferred to the PET images and the mean activity per milliliter of lung tissue was determined. Standardized uptake values (SUVs), defined as product of the mean lung ROI activity and the animal body weight divided by the injected dose were computed.
As a global index of lung inflammation, lung to muscle ratios were computed based on ROI analysis with the microCT image data sets. ROIs were drawn on transaxial slices, covering the entire lung volume. For every slice, the average pixel value was computed by dividing the total pixel values by the total number of pixels in that 2D ROI. The average pixel value in the whole lung is normalized to that of muscle to offset any potential fluctuations due to variability in CT acquisition parameters.
Group data are expressed as the mean ± SD. Student t-test was used to determine differences in SUV, %ID, microCT lung to muscle ratios, and MPO assay results among mice administered Klebsiella pneumonia and normal control mice. P value less than 0.05 was indicative of statistical significance. Sigma-Stat v3.0 (SPSS, Inc, Chicago, IL) was used for statistical calculations.
The desired cFLFLF-PEG-DOTA peptide containing a polyethyleneglycol linker was synthesized in a straightforward manner using a standard synthesis protocol of activating the carboxylic acid group on t-Boc protected-PEG with N-hydroxysuccinamide derivatization. Deprotection of t-Boc on PEG linker with TFA afforded a pegylated peptide with free NH2 which was further acylated with an N-hydroxylated-carboxy derivative of DOTA. At each step the intermediates were purified by RP-HPLC and the compounds were characterized by MALDI-TOF mass analysis. The final radiolabeled cFLFLF-PEG-64Cu ligand exhibited a pure single peak by RP-HPLC chromatogram and its radio purity was assigned to be more than 95%. The radiolabeled compound was compared with its non-radioactive counterpart for additional confirmation of the structure. HPLC analysis yielded no free 64Cu ions or radioligand fragments when testing the stability of the tracer in serum. The partition coefficients of cFLFLF-PEG-64Cu and cFLFLF-64Cu were measured to be -1.21 and 1.25, respectively, indicating the effectiveness of the PEG in enhancing the compound's water solubility.
The binding assay of the cFLFLF-PEG-64Cu to freshly purified human neutrophils yielded a mean Kd value of 17.7 nM. A representative saturation curve of specifically bound cFLFLF-PEG-64Cu is shown in Fig. 1. The binding data is additionally shown as a Scatchard plot in Fig. 1. cFLFLF-PEG-64Cu showed minimal agonist activity as assessed by neutrophil superoxide production at all concentrations studied while fMLF displayed agonist activity with an EC50 of 5.1×10-7 M (Fig. 2).
The clearance of cFLFLF-PEG-64Cu in blood followed a mono-exponential elimination pattern. The mean biological half-life of the peptide was calculated to be 55 ± 8 minutes.
Excised tissue concentrations of radio tracer (%ID) at 18 hr post injection in control and mice administered Klebsiella pneumoniae are shown in Fig. 3. For Klebsiella infected mice, the highest mean concentrations were found in the liver, kidney, and small intestine. The following organs (or tissue) demonstrated statistically significant differences in mean %ID values between control and infected mice as determined by Student's T-test: heart, lungs, liver, kidney, small intestine, stomach, and blood. Muscle, bone, and spleen did not exhibit statistically significant differences at the time point observed. The mean ratio of radioactivity in the infected to control lungs was 3.8.
The imaging properties of cFLFLF-PEG-64Cu were studied in control (n=4) and Klebsiella infected (n=6) mice. To establish the optimal time for imaging post ligand injection, the rate of washout in the control and infected lung was computed. The ligand half-life was determined to be 4.8 ± 0.7 hours in the control lung and 10.3 ± 2.9 hours in the infected lung. From these clearance half lives, the optimal imaging time window post injection was determined to be between 14 and 20 hrs. The 18 hr time point was chosen for convenience. Fig. 4 shows representative CT and PET images of cFLFLF-PEG-64Cu 18 hr post injection. These images show significant tracer accumulation in the lungs of the Klebsiella infected mouse compared to the control. Average lung SUVs for Klebsiella infected and control mice were 0.142 ± 0.054 and 0.028 ± 0.003, respectively (*P<0.003), as shown in Fig. 5. High liver uptake was observed in the PET images regardless of whether lung infection was present or not. The mean lung SUV ratio in infected vs. control mice was approximately 5.
The average lung to muscle ratio for infected and control mice were 0.58 ± 0.06 and 0.68 ± 0.04, respectively (P<0.001). CT images of the infected mouse show dilated airways and extensive patchy regions of increased lung density compared to the control lung. In general, Klebsiella-infected mice exhibited lungs with greater radiopacity.
MPO values measured in excised lung tissue (post imaging) revealed significantly elevated enzyme activity in mice administered Klebsiella pneumoniae compared to controls (Fig. 6). Mean MPO assay values (measured in change in optical density per min per milligram lung tissue) for Klebsiella-infected and control mice were 0.78 ± 0.17 and 0.12 ± 0.05, respectively (*P<0.005). The ratio of mean MPO activity in the infected to the control lungs was 6.3, slightly higher than the mean lung SUV ratio measured in the same mice.
Immunohistochemical staining of lung tissue from a control mouse revealed very few neutrophils (Fig. 7A) or macrophages (7B), and normal alveolar wall structure. In contrast, the Klebsiella infected mouse (euthanized 42 hr post administration) had a significant lung neutrophil burden (7C), with very low numbers of macrophages (7D), as indicated by arrowheads.
Although chemotactic peptide receptor agonists have been used successfully in numerous animal studies (5), they are not suitable inflammation imaging reagents due to their potential influence on leukocyte biological function. It has been previously demonstrated that the peptide FLFLF has a high binding affinity towards the neutrophil FPR and possesses antagonistic properties, thus does not induce neutropenia like other high affinity chemotactic peptide analogs (3). Another advantage is that radiolabeling procedures for antagonist peptides may be simpler and require less controlled generator elution and HPLC purification due to their minimal induced biological activity towards neutrophils. In this study it was shown that the peptide FLFLF can be conjugated with a PEG for the purpose of enhancing its hydrophilicity without significantly altering its binding affinity towards the neutrophil FPR. Binding studies to human neutrophils revealed a Kd of 17.7 nM, suggesting that the pegylation of peptide does not significantly alter the binding affinity towards the FPR when compared with unaltered parent peptide (Kd = 2 nM) as reported by Babich (3). In addition, the modification of peptide by pegylation may offer the prospect of fine-tuning pharmacokinetic parameters of the ligand to improve bioavailability and clearance. Additionally, the compound cFLFLF-PEG-64Cu showed no biological activity towards neutrophils as demonstrated by superoxide stimulation assays. In vivo biodistribution studies revealed high nontarget uptake in the liver, kidneys, and small intestine. Infected mice had significantly more tracer accumulation in liver and blood as compared to controls. This may be explained by elevated metabolism and enhanced recruitments of neutrophils in response to bacteria in the lungs. Blood clearance was determined to follow a mono-exponential pattern.
SUV measurements confirmed that the localization of the peptide was significantly higher in the lungs of Klebsiella infected mice as compared to controls 18 hrs post injection. Even though the blood half-life may be approximately 1 hour, the ligand does not clear from the lungs at this rate, as indicated in the Results.
To verify that the increase in measured lung SUVs is primarily due to infiltrating neutrophils responding to the bacteria, MPO assays were performed on post-imaged lung tissue. MPO analysis confirmed an increased population of leukocytes in the infected lungs, the magnitude of which correlated well with our average SUV results. Because MPO is an enzyme that is not exclusively found in neutrophils but also in macrophages, we sought additional evidence to show that neutrophils constitute the majority of infiltrating leukocytes in the lungs of this model. To determine this, we assessed the relative amounts of neutrophils and macrophages in both control and infected lungs by immunohistochemical (IHC) analysis. IHC analysis revealed that the primary cells infiltrating the infected lungs at the 42 hr time point post Klebsiella administration were neutrophils, with significantly fewer amounts of macrophages. We can therefore attribute our elevated lung SUV measurements in Klebsiella infected mice to infiltrating neutrophils. This is consistent with results reported by other groups (15).
Using microCT and computing lung to muscle ratios, we were able to detect significant changes in lung tissue characteristics as a result of administration of Klebsiella pneumoniae. Figure 4 exemplifies the marked increase in lung density that takes place 42 hr after infection. Although CT is sensitive to changes in lung tissue density, it cannot distinguish inflammation from fibrosis or edema, nor can it be used to identify which type of inflammatory cell is predominantly infiltrating the lungs.
We have demonstrated that the bioavailability and blood clearance properties of the peptide ligand cFLFLF can be improved by conjugation with a PEG modified linker without causing adverse affects to its binding affinity or antagonistic properties towards the formyl peptide receptor on neutrophils. It was also demonstrated that acute imaging of inflammation can be successfully achieved with this newly designed peptide with improved pharmacodynamic parameters and that the mechanism by which the compound accumulates appears to be by binding to receptors on neutrophils. Further biological evaluation of this novel imaging agent is ongoing, with the goal of refining the biological properties of the agent to best facilitate studies aimed to assess the efficacy of novel anti-inflammatory therapeutic drug candidates. Based on in vivo imaging results and in vitro cell function assays, this peptide appears to be a promising new radiopharmaceutical for the in vivo imaging of neutrophils.
The research was supported by Commonwealth Foundation for Cancer Research (DP) and NIH grants HL-073361 (JL) and HD051609 (KDF). This project was supported in part by a gift provided to the University of Virginia by Philip Morris USA. The review and approval process was overseen by an External Advisory Committee without any affiliation with the University, PM USA, or any other tobacco company. Funding for this project was based upon independent intramural and extramural reviews. The following people made contributions to this work: Joe Pole, Mark Williams, PhD, Li Xiao, and Ge Gao, PhD.
Financial Support: NIH P01 HL073361