Macrophage PET imaging in mouse models of cardiovascular disease and cancer with an apolipoprotein-inspired radiotracer

Macrophages are key inflammatory mediators in many pathological conditions, including cardiovascular disease (CVD) and cancer, the leading causes of morbidity and mortality worldwide. This makes macrophage burden a valuable diagnostic marker and several strategies to monitor these cells have been reported. However, such strategies are often high-priced, non-specific, invasive, and/or not quantitative. Here, we developed a positron emission tomography (PET) radiotracer based on apolipoprotein A1 (ApoA1), the main protein component of high-density lipoprotein (HDL), which has an inherent affinity for macrophages. We radiolabeled an ApoA1-mimetic peptide (mA1) with zirconium-89 (89Zr) to generate a lipoprotein-avid PET probe (89Zr-mA1). We first characterized 89Zr-mA1’s affinity for lipoproteins in vitro by size exclusion chromatography. To study 89Zr-mA1’s in vivo behavior and interaction with endogenous lipoproteins, we performed extensive studies in wildtype C57BL/6 and Apoe-/- hypercholesterolemic mice. Subsequently, we used in vivo PET imaging to study macrophages in melanoma and myocardial infarction using mouse models. The tracer’s cell specificity was assessed by histology and mass cytometry (CyTOF). Our data show that 89Zr-mA1 associates with lipoproteins in vitro. This is in line with our in vivo experiments, in which we observed longer 89Zr-mA1 circulation times in hypercholesterolemic mice compared to C57BL/6 controls. 89Zr-mA1 displayed a tissue distribution profile similar to ApoA1 and HDL, with high kidney and liver uptake as well as substantial signal in the bone marrow and spleen. The tracer also accumulated in tumors of melanoma-bearing mice and in the ischemic myocardium of infarcted animals. In these sites, CyTOF analyses revealed that natZr-mA1 was predominantly taken up by macrophages. Our results demonstrate that 89Zr-mA1 associates with lipoproteins and hence accumulates in macrophages in vivo. 89Zr-mA1’s high uptake in these cells makes it a promising radiotracer for non-invasively and quantitatively studying conditions characterized by marked changes in macrophage burden.


Materials and Methods
Radio-TLC assessment of tracer dissociation in fetal bovine serum (FBS). 89Zr-mA1 or bare 89 Zr were incubated in FBS at 37 ºC.The retention of 89 Zr by mA1 was assessed by radio-TLC using a Lablogic Scan-RAM Radio-TLC/HPLC detector at 1, 15, 60, 120, 240, 360 and 1440 minutes of incubation and using bare 89 Zr as control.
Mouse model of hypercholesterolemia. Female Apoe -/-mice (B6.129P2-Apoe tm1Unc /J, 8 weeks old, n = 8) were purchased from the Jackson Laboratory and fed a Western Diet (42% Kcal from fat TD88137, Envigo, Huntingdon, UK) for 12 weeks.These animals lack apolipoprotein E and develop severe hypercholesterolemia when fed a high-fat and highcholesterol diet 1,2 .
Mouse model of macrophage depletion.Female C57BL/6 mice (n=10) were purchased from Jackson Laboratory and randomly distributed over two groups.Twenty-four hours prior to the 89 Zr-mA1 injections, the experimental group received intravenous injections of Clodronate Liposomes (LIPOSOMA) at a dose of 100 µL of suspension / 10 grams of animal weight to deplete the macrophages.The control group received intravenous injections of 200 µL PBS.

Mouse model of myocardial infarction. Myocardial infarction in mice was induced by
permanent ligation of the left anterior descending (LAD) coronary artery of female C57BL/6 mice (n = 17) 3 .Animals were anesthetized with xylazine (10 mg/kg) and ketamine (100 mg/kg) and intubated using an endotracheal intubation kit from Braintree Scientific (Braintree, MA).
Left-sided thoracotomy and pericardial incision were performed.A 7-0 Silk suture was used to occlude the LAD.Incisions were closed with a 5-0 Silk suture.Infarcted animals were treated with 0.1 mg/kg of buprenorphine after surgery.
Histology.Following euthanasia, heart samples were embedded in OCT containing medium and were sectioned into 30 and 7 µm subsequent sections.30 µm slides were used for autoradiography and 10 µm slides were used for histology.Histology samples were stained with hematoxylin and eosin (H&E), and Mac3 staining (macrophages) using LAMP2 Antibody and 60 minutes of incubation (Invitrogen, Waltham, MA).TTC staining.Hearts were incubated with 2,3,5-Triphenyltetrazolium chloride (TTC) for 10 minutes at 37 °C, washed with PBS and sectioned into approximately 1 mm thick slices.
Autoradiography.Following euthanasia, tissue samples were excised and embedded in OCT mounting medium.OCT embedded samples were cut into 30 µm and 7 µm subsequent sections for autoradiography and histology, respectively.To determine radiotracer distribution, digital autoradiography was performed by placing tissue slices and sections in a film cassette against a phosphor imaging plate (BASMS-2325, Fujifilm, Valhalla, NY) for 10 minutes (TTC stained 1 mm slices) or 90 minutes (OCT embedded 30 µm sections) at -20 °C.
Phosphorimaging plates were read at a pixel resolution of 25 μm with a Typhoon 7000IP plate reader (GE Healthcare, Pittsburgh, PA).

1 .Supplemental Fig. 2 . 3 .
A) HPLC (C18 column) chromatograms of mA1 (UV at 278 nm) and mA1-DFO (UV at 278 nm).* Denotes the injection peak.B) LC-MS spectrum of mA1-DFO.C) Radio-TLC analysis of 89 Zr-mA1 using EDTA (50 mM) as the eluent.*Denotes the baseline.Bare 89 Zr would appear at the solvent front, at around 120 mm.D) Radio-TLC analysis of 89 Zr-mA1 or bare 89 Zr incubated in FBS at different timepoints.A) Representative 3D-rendered PET/CT images of infarcted animals scanned at 1 (left) and 24 hours (right) after i.v.tracer injection.Arrows indicate the heart.B) Representative PET (left), CT (middle) and fused PET/CT (right) images of infarcted animals scanned 24 hours after tracer injection showing different axis: axial (upper panel), coronal (middle panel) and sagittal (lower panel).Arrows indicate the infarct area.C) B) Biodistribution of 89 Zr-mA1 in the mouse model of myocardial infarction, 24 hours after injection, determined by ex vivo gamma counting (n = 5).C) 89 Zr-mA1 uptake in the infarcted myocardium represented as target-to-blood (top) or target-to-muscle (bottom) ratio, as determined by ex vivo gamma counting (n = 5).ID/g, injected dose per gram tissue.**P < 0.01.Cell populations in the infarcted myocardium as analyzed by CyTOF.A) CyTOF gating strategy of the infarcted myocardium to identify leukocyte subsets and their nat Zr-positive fractions.See also Figure 3H-J.B) Distribution of nat Zr-mA1 in different macrophage subclusters of three independent infarcted myocardium samples.Flow plots are color coded for nat Zr signal.C) Quantification of nat Zr-mA1 uptake in the macrophage subsets as a percentage of nat Zr + cells (n=3).
A) Representative 3D-rendered PET/CT images of tumor-bearing animals scanned at 1 (left, n = 4) and 24 hours (right, n = 11) after tracer injection.Arrows indicate the tumor area, delineated in red.B) Representative PET (left), CT (middle) and fused PET/CT (right) images of tumor-bearing animals scanned 24 hours after tracer injection showing different axis: axial (upper panel), coronal (middle panel) and sagittal (lower panel).Arrows indicate the tumor area, delineated in red.C) Biodistribution of 89 Zr-mA1 in the B16F10 melanoma mouse model 24 hours after injection, determined by ex vivo gamma counting (n = 5).C) 89 Zr-mA1 uptake in the tumor represented as targetto-blood (top) or target-to-muscle (bottom) ratio, as determined by ex vivo gamma counting.ID/g, injected dose per gram tissue.**P < 0.01.

.
Figure 4F.B) Distribution of nat Zr-mA1 in different macrophage subclusters of three independent tumor samples.Plots are color coded for nat Zr signal.C) Quantification of nat Zr-mA1 uptake in the macrophage subsets as a percentage of nat Zr + cells (n=3).

TABLE 2 .
1. PET-derived 89 Zr-mA1 in vivo uptake values in selected tissues at 24 hours post-injection [mean ± SD].Non-radioactive nat Zr-mA1 was used for CyTOF-based investigation of the peptide's in vivo cellular specificity in MI and melanoma mice.The compound was administered at the same dose as in the imaging experiments (200 µg) and quantified by measuring the abundance of 90 Zr.nat Zr-mA1 was allowed to circulate for 24 hours before euthanasia and tissue collection.Infarcted myocardium from MI mice and tumors from melanoma-inoculated mice were minced and digested with enzymatic digestion solution containing DNAse (60 List of antibodies used for the CyTOF analysis. U/mL), collagenase I (450 U/mL), collagenase XI (125 U/mL) and hyaluronidase (60 U/mL) (all Sigma-Aldrich) in PBS.Samples were then washed and resuspended into C10 media (RPMI + 10% FBS).Next, samples were incubated with a 1 µM Rh103 solution (Fluidigm, San Francisco, CA) for 20 minutes at 37 °C for viability staining and blocked with FcX block solution (Biolegend, San Diego, CA).Cells were then stained for 30 minutes at room temperature with monoclonal antibodies as described in the Supplemental Table2.Subsequently, single cell suspensions were fixed with 4% formaldehyde in PBS and incubated with Cell-ID™ 20-Plex Pd Barcoding Kit (Fluidigm, San Francisco, CA) for 30 minutes at room temperature.Prior to data acquisition, samples were washed with Cell Staining buffer and Cell Acquisition solution and resuspended in solution containing EQ normalization beads (Fluidigm, San Francisco, CA).Data was acquired on a Helios Mass Cytometer (Fluidigm, San Francisco, CA).Repeat acquisitions of the same sample were concatenated and normalized and barcoded samples were de-multiplexed using the Fluidigm software.Barcoded samples were de-multiplexed using the Zunder single cell debarcoder.Data were analyzed using Cytobank software (Beckman Coulter, Pasadena, CA) and FlowJo v10.8.0 (BD, Ashland, OR).SUPPLEMENTAL