SPECT imaging of lung ischemia-reperfusion injury using [^99mTc]cFLFLF for molecular targeting of formyl peptide receptor 1

Abstract

Primary graft dysfunction after lung transplantation, a consequence of ischemia-reperfusion injury (IRI), is a major cause of morbidity and mortality. IRI involves acute inflammation and innate immune cell activation, leading to rapid infiltration of neutrophils. Formyl peptide receptor 1 (FPR1) expressed by phagocytic leukocytes plays an important role in neutrophil function. The cell surface expression of FPR1 is rapidly and robustly upregulated on neutrophils in response to inflammatory stimuli. Thus, we hypothesized that use of [^99mTc]cFLFLF, a selective FPR1 peptide ligand, would permit in vivo neutrophil labeling and noninvasive imaging of IRI using single-photon emission computed tomography (SPECT). A murine model of left lung IRI was utilized. Lung function, neutrophil infiltration, and SPECT imaging were assessed after 1 h of ischemia and 2, 12, or 24 h of reperfusion. [^99mTc]cFLFLF was injected 2 h before SPECT. Signal intensity by SPECT and total probe uptake by gamma counts were 3.9- and 2.3-fold higher, respectively, in left lungs after ischemia and 2 h of reperfusion versus sham. These values significantly decreased with longer reperfusion times, correlating with resolution of IRI as shown by improved lung function and decreased neutrophil infiltration. SPECT results were confirmed using Cy7-cFLFLF-based fluorescence imaging of lungs. Immunofluorescence microscopy confirmed cFLFLF binding primarily to activated neutrophils. These results demonstrate that [^99mTc]cFLFLF SPECT enables noninvasive detection of lung IRI and permits monitoring of resolution of injury over time. Clinical application of [^99mTc]cFLFLF SPECT may permit diagnosis of lung IRI for timely intervention to improve outcomes after transplantation.

INTRODUCTION

Lung transplantation lags behind other solid organs in terms of donor organ utilization and transplant success rates (9, 19). As the supply of acceptable donor lungs remains low, wait list mortality rates continue to rise (44). One reason that most institutions have strict acceptance criteria for donor lungs is the high risk of primary graft dysfunction (PGD) (14). Clinically, PGD occurs within the early postoperative period after transplantation and is characterized by reduced oxygenation capacity and progressive lung failure (1). PGD occurs in up to 30% of patients and is the leading cause of early morbidity and mortality (1, 9, 10). Rates of chronic organ failure are also significantly higher in patients diagnosed with PGD (12).

The underlying pathophysiology responsible for PGD is ischemia-reperfusion injury (IRI) (5), which is also a common and severe postoperative complication after cardiopulmonary bypass, cardiopulmonary resuscitation, and pulmonary embolism (13). IRI occurs rapidly after transplant and involves disruption of endothelial and epithelial barriers as well as robust activation of innate immune cells including rapid infiltration of neutrophils (23, 34, 37, 55). Multiple inflammatory pathways are activated involving NADPH oxidase, oxidative stress, proinflammatory cytokines, receptor for advanced glycation end products (RAGE), and NF-κB activation (13, 16, 33, 35, 36, 54). These pathways, as well as upregulation of cell adhesion molecules, result in infiltration of neutrophils that further evoke tissue damage (28).

PGD is characterized by hypoxemia and alveolar infiltrates in the allograft within 72 h of lung transplantation (31). There are currently no clinically available methods for in vivo leukocyte labeling that can specifically identify inflammation in the transplanted lung (30). This inability to diagnose PGD early may contribute to high morbidity and mortality rates. Noninvasive imaging strategies are thus needed that allow for early and accurate detection of IRI, before onset of PGD, to facilitate early treatment as well as to monitor response to therapy.

Our laboratory has previously described the development of cinnamoyl-F-(D)L-F-(D)L-F-K (cFLFLF), a novel formyl peptide receptor 1 (FPR1) ligand, for in vivo neutrophil imaging (27). cFLFLF is a synthetic peptide with high-affinity binding of FPR1 that does not induce neutrophil activation and, when linked with polyethylene glycol (PEG), has a favorable pharmacokinetic profile (27, 41, 47, 48, 52). FPR1 is a G protein-coupled cell surface receptor that is rapidly upregulated on neutrophils in response to inflammatory stimuli and mediates efficient activation and recruitment of neutrophils to sites of infection/inflammation (11). When complexed with technetium-99m (^99mTc), [^99mTc]cFLFLF uptake can be visualized using single-photon emission computed tomography (SPECT), allowing for noninvasive localization of activated neutrophils at sites of inflammation (52). The objective of this study was to investigate the diagnostic utility of in vivo neutrophil labeling with [^99mTc]cFLFLF using a murine model of lung IRI. We hypothesized that [^99mTc]cFLFLF SPECT would permit imaging of lung IRI and also allow for monitoring of resolution of injury over time.

MATERIALS AND METHODS

cFLFLF peptide for imaging.

The present study utilized PEGylated cFLFLF conjugated with hydrazinonicotinamide and labeled with ^99mTc for SPECT imaging to target activated neutrophils (27, 47, 48, 52). cFLFLF-PEG coupled with commercially available Cy5- or Cy7-NHS ester (Sigma-Aldrich, St. Louis, MO) was used for fluorescence imaging (48).

Animals and lung IRI model.

The current study conformed to the standards set by the National Institutes of Health in the 2011 Guide for the Care and Use of Laboratory Animals, 8th edition. The study protocol was approved by the Animal Care and Use Committee at the University of Virginia. C57BL/6 wild-type mice (9–12 wk old, The Jackson Laboratory, Bar Harbor, ME) were used. An established model of left lung IRI via temporary hilar occlusion was used as previously described (Fig. 1) (37, 50). Briefly, mice were anesthetized with 2% inhaled isoflurane, intubated, and ventilated with room air at 120 breaths/min. Heparin (20 U/kg; Hospira Inc., Lake Forest, IL) was administered via right external jugular vein. A left anterolateral thoracotomy was made through the fourth intercostal space to expose the left hilum. Using a small olive-tip J-hook cannula, a 6–0 prolene suture was passed around the hilum. A Rummel tourniquet was fashioned with PE60 polyethylene tubing, and a surgical clip was used to temporarily occlude the hilar structures. The incision was closed with surgical clips, analgesia was administered (intraperitoneal buprenorphine, 0.2 mg/kg), and the animal was weaned from anesthesia and extubated when breathing spontaneously (Fig. 1A). Animals were returned to their cage for the prescribed 1 h of left lung ischemia. Animals were then briefly reanesthetized, and the Rummel tourniquet was removed to initiate reperfusion (Fig. 1B). Animals were returned to their cages and underwent 2, 12, or 24 h of reperfusion (using separate groups of animals). Sham animals underwent all steps described except for application of the Rummel tourniquet. Two hours before completion of reperfusion, animals received a tail-vein injection of imaging probe ([^99mTc]cFLFLF, Cy7-cFLFLF, or Cy5-cFLFLF) (Fig. 1C). At the conclusion of the reperfusion period, animals underwent SPECT imaging (Fig. 1D), and separate groups were used for lung function evaluation, lung histology, or immunofluorescence microscopy.

Fig. 1.

Murine model of left lung ischemia-reperfusion injury. A: through a left thoracotomy, application of a Rummel tourniquet and clip occludes the hilum of left lung [pulmonary artery (PA), pulmonary vein (PV), and bronchus] to induce ischemia. B: after 1 h of ischemia, the tourniquet is removed to begin reperfusion. C: two hours before the end of the reperfusion period (2, 12, or 24 h), mice are injected with the [^99mTc]cFLFLF single-photon emission computed tomography (SPECT) imaging probe (or Cy7-cFLFLF or Cy5-cFLFLF fluorescent probes). D: mice then undergo SPECT (or fluorescence) imaging.

Pulmonary function.

After reperfusion, pulmonary function [compliance, airway resistance, and pulmonary artery (PA) pressure] was measured (n = 5–6/group) using an IPL-1 isolated, buffer-perfused lung system (Harvard Apparatus, Holliston, MA) as previously described, which assesses the function of combined left and right lungs (38). Briefly, mice were anesthetized with ketamine-xylazine, intubated via tracheotomy, and ventilated (tidal volume: 7 μL/g body wt, rate: 100 breaths/min, positive end-expiratory pressure: 2 cmH₂O), and then exsanguinated via transection of the abdominal aorta and inferior vena cava. The main PA via the right ventricle was cannulated to allow perfusion of the lungs with 37°C Krebs-Henseleit buffer (flow rate: 60 μL·g body wt⁻¹·min⁻¹), and a left ventriculotomy allowed perfusate drainage. After a 5-min equilibration period, lung function data were captured over an additional 5-min period using PULMODYN data acquisition software (Harvard Apparatus).

SPECT imaging.

Separate groups of animals (n = 5–7/group) were used for SPECT imaging. Two hours before the end of reperfusion, animals received a 200 μL tail-vein injection of freshly labeled [^99mTc]cFLFLF (40–60 MBq). Two hours later, animals were anesthetized with 1% isoflurane, ventilated, and SPECT/CT imaging was performed. A microSPECT/CT scanner designed and built at the University of Virginia was used as previously described (52, 53). Briefly, the mouse was positioned supine on a carbon fiber half-cylinder tube located at the axis of rotation of the scanner gantry. Anesthesia was maintained with continuous inflow of isoflurane. Sequential CT and SPECT scans were acquired by transitioning the animal from one subsystem to the other along the axis of rotation, allowing for consistent fusion of CT and SPECT images using stored offset parameters. An interactive data language (IDL) program (Harris Geospatial Solutions, Broomfield, CO) was used with the CT images for detector sensitivity uniformity correction and dark count subtraction. The images were reconstructed using a Feldkamp three-dimensional filtered back projection algorithm (COBRA, Exxim, Pleasanton, CA). A custom-written interface using Kmax software (Sparrow Corp., Port Orange, FL) was used to acquire SPECT images using a 10 cm × 10 cm Nal(TI) gamma camera equipped with a pinhole collimator (1.0 mm), and a maximum-likelihood expectation-maximization algorithm was used to reconstruct SPECT images as previously described (52, 56). Reconstructed CT and SPECT images were converted to DICOM format and processed using medical image viewing software Horos (https://www.horosproject.org). CT images were thresholded with a lung window and used to manually draw anatomical boundaries for body and lungs for each animal. The anatomical boundaries were then superimposed on the corresponding SPECT images.

For SPECT analysis of probe uptake, 10 consecutive transaxial SPECT slices through the region of enhanced probe uptake in the left lung were extracted using the heart as a reference and CT as guidance. Within each slice, a region of interest (ROI) centered on the pixel with maximum signal intensity was drawn by hand using Asipro VM software (Siemens Preclinical Solutions, Knoxville, TN), with the ROI boundary outlining the region of enhanced probe uptake. For quantification of probe uptake in lungs, total signal intensity values (counts/pixel normalized by ROI area) were obtained on all 10 slices in lungs of animals after sham or IR surgery. An identical set of ROI boundary locations were then drawn within the corresponding region of 10 slices in the right lung, following left-right geometric inversion of the ROIs. The sum of the total signal intensity values within the ROIs of 10 slices from each lung were then used to calculate the ratio of signal intensity between the left and right lungs.

At the completion of SPECT/CT, animals were euthanized, and major organs were collected, weighed, and placed in a gamma-well counter (PerkinElmer, Waltham, MA) to assess probe biodistribution (n = 6/group). The radioactivity of each organ was normalized to injected dose, tissue weight, and body weight. Organ distribution of [^99mTc]cFLFLF is expressed as a percentage of total radioactivity (the sum radioactivity in all organs assessed). Lung ^99mTc uptake is reported as the ratio of left-to-right lung relative percent radioactivity.

Fluorescence imaging with Cy7-cFLFLF.

Separate groups of animals (n = 5/group) received a 100 μL tail-vein injection of 2 nM Cy7-cFLFLF. Two hours later, animals were euthanized and both lungs collected. Because of the limited penetration and scattering of fluorescence signal in whole mice, lungs were imaged ex vivo using an IVIS Spectrum Imaging System (excitation 745 nm, emission 775 nm; PerkinElmer). Total fluorescence radiance was obtained for each lung by designating a region of interest that incorporated the entire lung field.

In vitro neutrophil binding of cFLFLF.

Spleens from wild-type mice were dissociated into a single-cell suspension using a gentleMACS Dissociator (Miltenyi Biotec, Auburn, CA). Neutrophils (>95% pure) were isolated via positive selection using a murine neutrophil isolation kit (Miltenyi Biotec). Neutrophils were then plated (1 × 10⁵ cells/chamber) onto chamber slides (Thermo Fisher Scientific, Waltham, MA). Phorbol myristate acetate (PMA, 32 nM, Sigma-Aldrich) was added to select wells to stimulate neutrophil activation, and slides were incubated for 2 h at 37°C. All cells were then fixed using a 50:50 methanol:acetone mixture and washed. Cy5-cFLFLF (1 nM), anti-Ly6G-FITC (5 μg/mL; Biolegend, San Diego, CA), and DAPI (1 μg/mL; Invitrogen, Carlsbad, CA) were added to all wells and incubated overnight at 4°C along with blocking buffer (5% Blotto in PBS, 0.5 mL/well, Santa Cruz Biotechnology, Dallas, TX) to reduce nonspecific binding. Images were obtained with an Olympus IX81 inverted confocal microscope (Olympus, Tokyo, Japan) with a CCD camera at ×200 magnification.

Neutrophil immunohistochemistry.

Left lungs from separate groups of animals (n = 5/group) were fixed in 10% buffered formalin, paraffin-embedded, and sectioned. A Vectastain ABC kit (Vector Laboratories, Burlingame, CA) was used to immunostain neutrophils as previously described (43). The primary antibody used was rat anti-mouse neutrophil antibody (AbD Serotec, Raleigh, NC) and secondary antibody used was alkaline phosphatase-conjugated anti-rat IgG (Sigma-Aldrich). Fast-Red (Sigma-Aldrich) was used for neutrophil staining. Purified normal rat IgG (eBioscience, San Diego, CA) was used as a negative control. Neutrophils in five semistandardized fields per lung section were counted manually in a blinded fashion at ×20 magnification and averaged to obtain a final count per high-powered field (HPF) per lung.

Immunofluorescence analysis of Cy5-cFLFLF binding in lungs.

After sham or IR surgery, mice were injected with Cy5-cFLFLF (i.v., 5 nmoles in 100 μL saline). At the end of 2 h of reperfusion, mice were anesthetized and lungs were slowly flushed with 5 mL PBS with calcium and magnesium via the right ventricle followed by flushing with 3 mL of fixing solution (4% sucrose, 3% paraformaldehyde in PBS with calcium and magnesium). Fixing solution (1 mL) was also gently instilled through the trachea until the lungs were inflated. The heart-lung blocks were then immersed in fixing solution for 15 min. The fixing solution instilled into lungs was then withdrawn and replaced with ~1 mL optimum cutting temperature (OCT) compound (Thermo Fisher Scientific, Middletown, VA) through the trachea. The heart-lung blocks were then immersed in OCT compound and snap-frozen by dipping into isopentane (submerged in a liquid nitrogen bath) for 10 s. The frozen blocks (wrapped with aluminum foil and stored at −80°C) were then cut into 5 μm sections using a Leica Cryostat. For immunofluorescence staining, sections were washed with PBS without calcium and magnesium and blocked with 1% BSA in PBS at room temperature for 1 h. For neutrophil staining, sections were incubated overnight at 4°C with FITC-conjugated rat anti-mouse neutrophil (Ly-6B.2) antibody (1 μg/mL, Abcam, Cambridge, MA) and Cy5-cFLFLF (5 μg/mL). For macrophage staining, sections were incubated overnight at 4°C with rat anti-mouse macrophage (Mac-2) antibody (0.5 μg/mL, Cedarlane, Burlington, NC) and Cy5-cFLFLF (5 μg/mL). Sections were then washed with PBS and incubated with Alexa Fluor 488-conjugated goat anti-rat IgG (1:1,000 dilution, Cell Signaling, Danvers, MA) secondary antibody for 1 h at room temperature. For nuclear staining, all sections were washed three times with PBS and incubated with DAPI (1 μg/mL, Sigma-Aldrich, St. Louis, MO) for 15 min at room temperature in the dark. Slow Fade mounting medium was applied to the slides, and images were obtained using an Olympus BX51 fluorescence microscope equipped with an Olympus DP70 digital camera (Minneapolis, MN).

Statistical analysis.

Groups were compared using two-tailed Student’s t test or one-way analysis of variance (ANOVA) with a post-hoc Bonferroni correction for multiple comparisons. A P value of <0.05 was used to determine statistical significance. All values are reported as means ± SD. Prism 7 software (GraphPad Software, La Jolla, CA) was used for analyses.

RESULTS

Lung function after IR.

Pulmonary function measurements after IR revealed peak injury (dysfunction) after 2 h of reperfusion with significant resolution of injury over time. Pulmonary compliance was significantly reduced after IR versus sham (IR 2-h: 3.0 ± 0.4 vs. Sham: 6.5 ± 0.7 μL/cmH₂O, P < 0.0001) and improved with increasing reperfusion time (IR 12-h: 3.9 ± 0.5 μL/cmH₂O, P > 0.05 vs. IR 2-h; IR 24-h: 5.0 ± 0.5 μL/cmH₂O, P = 0.02 vs. IR 12-h) (Fig. 2A). Airway resistance was significantly higher after IR versus sham (IR 2-h: 1.91 ± 0.13 vs. Sham: 1.07 ± 0.09 cmH₂O·μL⁻¹·s⁻¹, P < 0.0001) and improved with increasing reperfusion time (IR 12-h: 1.45 ± 0.17 cmH₂O·μL⁻¹·s⁻¹, P = 0.0005 vs. IR 2-h; IR 24-h: 1.28 ± 0.14 cmH₂O·μL⁻¹·s⁻¹, P > 0.05 vs. Sham) (Fig. 2B). PA pressure was significantly higher after IR versus sham (IR 2-h: 12.3 ± 1.0 vs. Sham: 5.7 ± 0.4 cmH₂O, P < 0.0001) and improved with increasing reperfusion time (IR 12-h: 8.9 ± 1.6 cmH₂O, P = 0.0004 vs. IR 2-h; IR 24-h: 6.1 ± 0.9 cmH₂O, P > 0.05 vs. Sham) (Fig. 2C).

Fig. 2.

Lung function after ischemia-reperfusion (IR). Measurement of lung function demonstrated peak injury (dysfunction) after ischemia and 2 h of reperfusion, with resolution of injury over time. A: Pulmonary compliance, B: airway resistance, and C: pulmonary artery pressure were measured after IR compared with sham animals. Mice after 1 h of left lung ischemia and 2, 12, or 24 h of reperfusion are represented as IR 2-h, IR 12-h, and IR 24-h, respectively. Sham animals underwent 2 h of perfusion following sham surgery. *P < 0.05 vs. Sham, ^P < 0.05 vs. IR 2-h, ^#P < 0.05 vs. IR 12-h. n = 5–6/group.

Neutrophil infiltration after IR.

Immunostaining of left lungs demonstrated significant neutrophil infiltration after IR at all three time points (IR 2-h: 222 ± 30, P < 0.0001; IR 12-h: 175 ± 45, P < 0.0001; IR 24-h: 145 ± 21 neutrophils/HPF, P < 0.0001) compared with sham (52 ± 23 neutrophils/HPF) (Fig. 3). Neutrophil infiltration was highest after 2 h of reperfusion and decreased over time. There was a significant difference between IR 2-h and IR 24-h neutrophil counts (P = 0.01).

Fig. 3.

Lung neutrophil infiltration after ischemia-reperfusion (IR). A: representative lung sections immunostained for neutrophils (red color, ×40 magnification). B: quantification of neutrophils per high-powered field (HPF). Mice after 1 h of left lung ischemia and 2, 12, or 24 h of reperfusion are represented as IR 2-h, IR 12-h, and IR 24-h, respectively. Sham animals underwent 2 h of perfusion following sham surgery. *P < 0.0001 vs. Sham, ^P = 0.01 vs. IR 2-h. n = 5/group.

SPECT imaging of IRI.

SPECT with [^99mTc]cFLFLF permitted semiquantitative imaging of left lung IRI after 2, 12, and 24 h of reperfusion. Signal intensity after IR in the left lung peaked after 2 h of reperfusion with reduction in signal intensity over time while minimal signal occurred in the right lung. Representative coronal SPECT images are shown in Fig. 4A. The left-to-right lung SPECT signal intensity ratios after IR and 2 and 12 h of reperfusion were significantly higher than sham mice (2-h: 4.24 ± 0.81 vs. 1.08 ± 0.06, P < 0.0001; 12-h: 2.14 ± 0.55 vs. 1.0 ± 0.10, P = 0.0002) (Fig. 4B). By 24 h of reperfusion there was no significant difference in left-to-right lung signal intensity ratios between IR and sham animals (1.42 ± 0.38 vs. 1.18 ± 0.08, P = 0.24) (Fig. 4B). Total uptake of [^99mTc]cFLFLF in left lungs by gamma counts was significantly higher after IR at all three time points compared with sham (2-h: 2.21 ± 0.53 vs. 0.95 ± 0.10, P = 0.02; 12-h: 1.80 ± 0.12 vs. 1.03 ± 0.11, P < 0.0001; 24-h: 1.38 ± 0.10 vs. 1.11 ± 0.12, P = 0.003) (Fig. 4C). The left-to-right lung signal intensity ratios by SPECT were 3.9-, 2.1-, and 1.2-fold higher following 2, 12, and 24 h reperfusion, respectively, versus sham. The left-to-right lung probe uptake by gamma counts was 2.3-, 1.7-, and 1.2-fold higher following 2, 12, and 24 h reperfusion, respectively, versus sham. Importantly, SPECT signal intensity ratio and [^99mTc]cFLFLF uptake diminished over reperfusion time, correlating with changes in pulmonary function and neutrophil counts.

Fig. 4.

Single-photon emission computed tomography (SPECT) imaging of lung ischemia-reperfusion (IR) injury. SPECT demonstrated peak signal after ischemia and 2 h of reperfusion, which diminished over time. A: representative SPECT images (coronal views shown) after sham surgery or 1 h of left lung ischemia followed by 2, 12, or 24 h of reperfusion. Mice were injected with [^99mTc]cFLFLF 2 h before the end of reperfusion. Outlines of the body (solid green line) and lungs (dashed green lines) are shown. B: semiquantitative measurements of SPECT signal intensity, reported as the ratio of left-to-right lung signal intensity as described in materials and methods (n = 5–7/group). C: quantification of [^99mTc]cFLFLF uptake in lungs measured ex vivo using a gamma-well counter and reported as left-to-right lung ratio of radioactivity ([^99mTc]cFLFLF uptake) as described in materials and methods (n = 5–6/group). Mice after 1 h of left lung ischemia and 2, 12, or 24 h of reperfusion are represented as IR 2-h, IR 12-h, and IR 24-h, respectively. Sham animals underwent appropriate 2, 12, or 24 h of perfusion following sham surgery as indicated.

Organ biodistribution of [^99mTc]cFLFLF.

Organ biodistribution of [^99mTc]cFLFLF was measured in mice undergoing 2 h of reperfusion following sham surgery or ischemia (Fig. 5). Uptake was highest overall in kidneys representing excretion of excess probe. Left lung uptake after IR was significantly greater compared with sham left lung (12.4 ± 1.3 vs. 5.6 ± 2.1 relative % radioactivity, P = 0.02) or right lung after IR (12.4 ± 1.3 vs. 3.9 ± 0.4 relative % radioactivity, P < 0.0001). Uptake was not significantly different in right lungs or in other organs between sham and IR animals.

Fig. 5.

Organ biodistribution of [^99mTc]cFLFLF. Organ uptake of [^99mTc]cFLFLF after left lung ischemia and 2 h of reperfusion was measured and is reported as a percentage of total radioactivity of all organs harvested and normalized to injected dose, tissue weight, and animal body weight. Results demonstrate highest uptake in the kidneys as well as significant increase in left lung uptake in the ischemia-reperfusion (IR) animals. IR mice underwent 1 h of left lung ischemia and 2 h of reperfusion while the Sham mice underwent 2 h of perfusion following sham surgery. *P = 0.02 vs. Sham left lung, ^P < 0.0001 vs. IR right lung. n = 6/group.

Ex vivo fluorescence imaging of IRI with Cy7-cFLFLF.

SPECT imaging results were confirmed using Cy7-cFLFLF and near-infrared fluorescence imaging of lungs. Fluorescence in the left and right lungs of sham and IR mice after 2 h of reperfusion was measured ex vivo (Fig. 6A). Signal quantification demonstrated total radiance efficiency that was significantly higher in the IR 2-h mice (1.9-fold) compared with sham animals (1.66 ± 0.16 vs. 0.89 ± 0.10 left/right lung total radiance efficiency, P < 0.0001) (Fig. 6B).

Fig. 6.

Cy7-cFLFLF near-infrared fluorescence imaging of lung ischemia-reperfusion (IR) injury. Mice were injected with Cy7-cFLFLF 2 h before imaging. IR animals underwent 1 h of left lung ischemia and 2 h of reperfusion. Sham animals underwent 2 h of perfusion following sham surgery. A: representative fluorescence images of ex vivo lungs from Sham and IR mice (representative lungs from two mice of each group are shown). B: quantification of fluorescence intensities reported as ratio of left-to-right lung total radiance efficiency. A significant increase in fluorescence signal occurred in the left lungs of IR animals vs. Sham. n = 5/group.

Activated neutrophils bind cFLFLF.

Primary murine neutrophils were assessed in vitro to determine if cFLFLF (labeled with Cy5) specifically binds to activated neutrophils. Neutrophils activated by PMA demonstrated specific and substantial Cy5-cFLFLF binding that colocalized with DAPI and anti-Ly6G-FITC, while nonactivated neutrophils displayed minimal binding (Fig. 7).

Fig. 7.

Activated neutrophils selectively bind cFLFLF. Primary murine neutrophils were cultured with or without phorbol myristate acetate (PMA) for 2 h and fixed. Slides were then incubated with DAPI (blue, nuclei), anti-Ly6G-FITC (green, neutrophils), and Cy5-cFLFLF (red), and imaged by immunofluorescence microscopy. The far-right column shows an enlarged view of the insets shown in the “Merged” column.

cFLFLF primarily labels neutrophils in vivo after IR.

Immunofluorescence staining of left lung sections demonstrated that Cy5-cFLFLF binding was largely limited to neutrophils after ischemia and 2 h of reperfusion (Fig. 8). Nearly all neutrophils in IR lungs demonstrated strong cFLFLF binding as demonstrated by colocalization of Cy5-cFLFLF and anti-neutrophil-FITC fluorescence (arrows in Fig. 8A). A small fraction of macrophages in IR lungs was observed to bind cFLFLF (arrows in Fig. 8B). Lungs from sham mice demonstrated minimal Cy5-cFLFLF binding.

Fig. 8.

cFLFLF predominantly binds neutrophils in vivo after lung ischemia-reperfusion (IR). IR animals underwent 1 h of left lung ischemia and 2 h of reperfusion. Sham animals underwent 2 h of perfusion following sham surgery. Lung sections were immunostained with DAPI (blue, nuclei), Cy5-cFLFLF (red), and either anti-Ly-6B.2-FITC (green, neutrophils) (A) or anti-Mac-2-FITC (green, macrophages) (B) and imaged by fluorescence microscopy as described in materials and methods. A: nearly all neutrophils in lungs of IR animals corresponded with overlapping Cy5 and FITC signal (arrows). B: a small fraction of macrophages in lungs of IR animals corresponded with overlapping Cy5 and FITC signal (arrows). Very little Cy5-cFLFLF signal was visible in lungs of Sham animals. Representative images (×20) are shown. The far-right images show enlarged views of the insets depicted in the “Merged DAPI/FITC/Cy5” images.

DISCUSSION

The present study demonstrated the feasibility of using an FPR1-specific ligand ([^99mTc]cFLFLF) to label infiltrating neutrophils (and some macrophages) in vivo and thus image sites of acute pulmonary inflammation via SPECT. Using a mouse model of lung IRI that results in peak injury at 2 h with resolution over 24 h, SPECT imaging with [^99mTc]cFLFLF demonstrated a 3.9-fold increase in left lung signal intensity at 2 h of reperfusion, a 2.1-fold difference at 12-h, and no significant difference at 24 h. Importantly, SPECT results correlated with improvement in lung function and reduced neutrophil counts over time, thus allowing noninvasive monitoring of IRI and resolution of injury. Organ biodistribution revealed 1) increased uptake of [^99mTc]cFLFLF in left lungs after IR, 2) no significant differences in uptake among other organs between IR and sham, and 3) highest uptake in the kidney due to probe excretion. SPECT results were confirmed by Cy7-cFLFLF fluorescence imaging of lungs ex vivo, and immunofluorescence microscopy documented that cFLFLF binding is largely limited to infiltrating neutrophils in the lung after IR.

The risk of IRI leading to PGD in the early postoperative period drives transplant surgeons to be conservative in their acceptance of donor organs (5). When PGD occurs, early morbidity and mortality rises, as does the rate of long-term complications such as bronchiolitis obliterans (1, 9, 10, 12). Transplant clinicians are limited in their ability to diagnose and monitor PGD, having to rely mainly on nonspecific chest radiographs and arterial oxygenation values. Providing clinicians with the ability to noninvasively image active inflammation with cell-specific molecular probes may aid in earlier diagnosis, earlier treatment, and a means to monitor response to treatment.

Our laboratory and others have documented that lung IRI is a complex inflammatory process that entails innate immune cell activation (22, 37, 55). The release of HMGB1, TNF-α, IL-1β, and IL-17 leads to high levels of chemokine production and rapid, robust infiltration of activated neutrophils (29, 34). Because neutrophils are rapidly recruited and evoke tissue injury during lung IRI, we aimed to noninvasively image neutrophil infiltration as a means to provide early detection of lung IRI. Activated neutrophils highly express FPR1 that binds mitochondrial N-formyl peptides, which are known damage-associated molecular pattern molecules (DAMPs) (24, 45). We have developed cFLFLF as a ligand with high binding affinity for FPR1 that enables noninvasive in vivo imaging of activated neutrophils during inflammation (27). As shown in Fig. 8A, increased cFLFLF binding after IR was primarily limited to the large number of infiltrating neutrophils. However, macrophages have also been described as being capable of expressing FPR1 during inflammation (4), and cFLFLF has been used to image macrophages that infiltrate islet cells of diabetic mice (51) or herniated disks (46). In the current study, we observed a small fraction of pulmonary macrophages that did bind cFLFLF (Fig. 8B), which was a weaker signal compared with neutrophils. This could be explained if FPR1 expression is limited to infiltrating macrophages (versus resident alveolar macrophages). In support of this, we have previously reported that total macrophage numbers are not significantly elevated in murine lungs after 1 h of ischemia and 2 h of reperfusion (49). It is also possible that macrophages express relatively less FPR1 on a per cell basis versus neutrophils after lung IRI, which is suggested in Fig. 4 where the binding of cFLFLF to macrophages after IRI is substantially less intense versus neutrophils.

Conjugation of cFLFLF with various radioisotopes and fluorescent dyes has permitted visualization of probe uptake with different imaging modalities including PET, SPECT, optical imaging, and MRI (27, 42, 48, 52). [^99mTc]cFLFLF is a useful molecular probe that can be applied broadly to different disease processes involving acute inflammation and neutrophil activation and appears to be more specific than other nuclear medicine imaging techniques. For example, in a rat model of acute osteomyelitis, [^99mTc]cFLFLF SPECT was found to be superior to [^99mTc]methylene diphosphonate bone scanning and [¹⁸F]fluorodeoxyglucose ([¹⁸F]FDG) PET for diagnosis based on specificity and image quality, and [^99mTc]cFLFLF was effective at evaluating the therapeutic response to treatment (8). Although [¹⁸F]FDG PET is used extensively to evaluate primary and recurrent forms of cancer, it is not cell specific (15, 17, 18). Metabolically active cells, such as tumor cells, take up [¹⁸F]FDG, and thus [¹⁸F]FDG does not delineate specific cell types, which limits its utility for accurate cell type-specific diagnosis of inflammatory processes such as IRI. Nevertheless, PET as an imaging modality provides superior image quality compared with SPECT. The combination of a cell type-specific imaging probe with an imaging modality that produces high resolution and image quality will yield the most diagnostically accurate combination.

A consistent method of labeling activated leukocytes in vivo would be a significant breakthrough for the field of nuclear medicine as most successful methods of leukocyte labeling thus far have required ex vivo radiolabeling (30). Although currently the gold standard for radionuclide imaging of inflammation and infection, ex vivo methods are labor-intensive and have a high risk of contamination, requiring blood or bone marrow extraction, ex vivo leukocyte labeling, and injection of cells back into the patient. Given these limitations, attempts have been made using antigranulocyte antibodies for in vivo leukocyte labeling with limited success. Agents that have been investigated include: 1) Granuloscint, a monoclonal IgG1 antibody to the NCA-95 antigen on leukocytes, 2) Sulesomab, a monoclonal antibody fragment that binds to NCA-90 antigen, and 3) ^99mTc-Fanolesomab, an IgM antibody that binds to CD15 antigen (30). Unfavorable pharmacokinetics and varying sensitivities and specificities have limited the adoption of these agents into clinical practice.

Experimental means to noninvasively image lung rejection or lung inflammation using PET imaging have been recently described in several studies. Chen et al. (7) employed [¹⁸F]FDG PET in a murine lung transplant model to demonstrate that recipients with acute lung allograft rejection have increased [¹⁸F]FDG uptake driven primarily by accumulation of T cells in the graft. Using a canine model of oleic acid-induced acute lung injury, Chen et al. (6) also showed that “[¹⁸F]FDG uptake in these lungs reflects the state of neutrophil activation.” Rodrigues et al. (32) demonstrated uptake of [¹⁸F]FDG that paralleled neutrophil infiltration in a rat model of lipopolysaccharide-induced acute lung injury. These studies suggest that elevated [¹⁸F]FDG uptake by PET imaging, in the appropriate clinical context, could be used to guide patient management decisions. However, the limiting factor with the use of [¹⁸F]FDG is its nonspecificity; permitting only discrimination of cell populations or tissues with high glucose metabolic rate, usually active proliferation, from a quiescent cell population. In addition, higher uptake of FDG in the heart occurs due to the high metabolic rate of cardiomyocytes, which can affect accurate lung imaging. Thus, although [¹⁸F]FDG may be useful as a general indication of inflammation in the lung, knowing the activation status of specific innate immune cell populations at various stages would not only help to understand the physiopathology, but would also help to counteract any harmful effects caused by their activation or infiltration at the site of inflammation. Immune cell-specific molecular imaging probes such as cFLFLF would thus be more informative.

Using a murine lung transplant model, Liu et al. (26) reported the use of a CCR2-binding peptide ([⁶⁴Cu]DOTA-ECL1i) for PET imaging of lung IRI. This study, which targeted CCR2⁺ cells (monocytes), introduced the powerful potential of cell-specific molecular imaging techniques for noninvasive detection of lung IRI. Our current study demonstrates the effectiveness of using [^99mTc]cFLFLF to image activated neutrophils in vivo via FPR1 binding for noninvasive detection and monitoring of lung IRI by SPECT. Neutrophils typically dominate the initial leukocyte influx to sites of active inflammation, including lung IRI, and this first wave of neutrophil extravasation precedes a second wave of inflammatory monocyte extravasation (39). Thus, the use of cFLFLF SPECT to image neutrophils may allow earlier detection of IRI compared with imaging of monocytes with DOTA-ECL1i.

The current study has limitations. First, the lung IRI model does not entail transplantation, and our results will need to be confirmed in a transplant model. However, the clinical relevance of the murine IRI model is supported by the fact that there is remarkable synergy between this model and the inflammatory markers (e.g., neutrophil infiltration) observed in both the murine lung transplant model and in lung-transplant patients (3, 20, 21, 37, 40, 49, 55). Second, although the present study utilized SPECT, use of PET imaging could offer higher resolution and improved quantification of lung IRI, if needed. For example, cFLFLF has been labeled with PET radioisotopes such as ⁶⁴Cu (27). However, several advantages of SPECT imaging are that it is less expensive and that several different cell-specific probes (linked to different radioisotopes) could be simultaneously used to monitor the inflammatory state of multiple cell populations in the lung after transplant. Third, imaging shortly after transplantation (e.g., within an hour) would not be clinically feasible. Although the use of a planar SPECT imaging approach was not investigated in this study, it is possible that a planar SPECT imaging approach with portable technology that can be transported to the patient’s bedside may be more clinically applicable and efficient in some cases such as for those patients who cannot be moved in the first 24 h after transplant. Fourth, [^99mTc]cFLFLF SPECT, by itself, would not differentiate between other neutrophil-mediated inflammatory conditions in the lung such as infection or antibody-mediated rejection (2, 25). As an initial, proof-of-concept study, the current study aimed to interrogate the ability of [^99mTc]cFLFLF SPECT to semiquantitatively assess lung IRI by labeling neutrophils. Future studies, as stated above, will involve simultaneous SPECT imaging of activated neutrophils with imaging of other relevant cell types (e.g., T cells, M1 versus M2 macrophages, endothelium, etc.) to enable a more discriminatory diagnosis and to monitor temporal changes in lung injury and resolution after IR.

In conclusion, using a murine model of left lung IRI, the present study demonstrates that [^99mTc]cFLFLF SPECT permits semiquantitative, noninvasive detection of IRI. Importantly, changes in [^99mTc]cFLFLF uptake and SPECT signal intensity correlated with changes in lung function and neutrophil infiltration over time, suggesting that resolution of IRI can be monitored by [^99mTc]cFLFLF SPECT. Clinically, in vivo leukocyte labeling with cell-specific molecular probes such as cFLFLF may allow for early detection of IRI, which will permit rapid therapeutic interventions, monitoring of response to therapy, and improved outcomes after lung transplantation.

GRANTS

This work was supported by NIH National Heart, Lung, and Blood Institute Grants R01 HL130053 (V. E. Laubach, I. L. Kron), T32 HL007849 (V. E. Laubach), and UM1 HL088925 (I. L. Kron), and Pennsylvania Cystic Fibrosis Inc. (V. E. Laubach).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

I.L.K. and V.E.L. conceived and designed research; E.J.C., M.D.C., Y. Zhao, Y. Zhang, J.H.M., D.K.G., W.Z.C., A.K.S., and D.P. performed experiments; E.J.C., M.D.C., Y. Zhao, Y. Zhang, J.H.M., D.K.G., J.D., W.Z.C., A.K.S., D.P., and V.E.L. analyzed data; E.J.C., M.D.C., J.H.M., D.K.G., J.D., W.Z.C., A.K.S., I.L.K., D.P., and V.E.L. interpreted results of experiments; E.J.C., M.D.C., Y. Zhao, D.K.G., J.D., and V.E.L. prepared figures; E.J.C. and V.E.L. drafted manuscript; E.J.C., M.D.C., D.K.G., J.D., A.K.S., D.P., and V.E.L. edited and revised manuscript; E.J.C., M.D.C., Y. Zhao, Y. Zhang, J.H.M., D.K.G., J.D., W.Z.C., A.K.S., I.L.K., D.P., and V.E.L. approved final version of manuscript.