Selective Imaging of Lung Macrophages Using [¹¹C]PBR28-Based Positron Emission Tomography

Abstract

Purpose:

We tested whether the translocator protein (TSPO)-targeted positron emission tomography (PET) tracer, N-acetyl-N-(2-[¹¹C]methoxybenzyl)-2-phenoxy-5-pyridinamine ([¹¹C]PBR28), could distinguish macrophage dominant from neutrophilic inflammation better than 2-deoxy-2-[¹⁸F]fluoro-D-glucose ([¹⁸F]FDG) in mouse models of lung inflammation and assessed TSPO association with macrophages in lung tissue from the mouse models and in patients with chronic obstructive pulmonary disease (COPD).

Procedures:

MicroPET imaging quantified [¹¹C]PBR28 and [¹⁸F]FDG lung uptake in wild-type (Wt) C57BL/6J or heterozygous transgenic monocyte-deficient Wt/opT mice at 49 days after Sendai virus (SeV) infection, during macrophage-dominant inflammation, and in Wt mice at 3 days after SeV infection or 24 h after endotoxin instillation during neutrophilic inflammation. Immunohistochemical staining for TSPO in macrophages and neutrophils was performed using Mac3 and Ly6G for cell identification in mouse lung sections and CD68 and neutrophil elastase (NE) in human lung sections taken from explanted lungs from patients with COPD undergoing lung transplantation and donor lungs rejected for transplantation. Differences in tracer uptake among SeV-infected, endotoxin-treated, and uninfected/untreated control mice and in TSPO staining between neutrophils and macrophage populations in human lung sections were tested using analysis of variance.

Results:

In Wt mice, [¹¹C]PBR28 uptake (% injected dose/ml lung tissue) increased significantly with macrophage-dominant inflammation at 49 days (D49) after SeV infection compared to controls (p = <0.001) but not at 3 days (D49) after SeV infection (p = 0.167). [¹¹C]PBR28 uptake was unchanged at 24 h after endotoxin instillation (p = 0.958). [¹⁸F]FDG uptake increased to a similar degree in D3 and D49 SeV-infected and endotoxin-treated Wt mice compared to controls with no significant difference in the degree of increase among the tested conditions. [¹¹C]PBR28 but not [¹⁸F]FDG lung uptake at D49 post-SeV infection was attenuated in Wt/opT mice compared to Wt mice. TSPO localized predominantly to macrophages in mouse lung tissue by immunostaining, and TSPO staining intensity was significantly higher in CD68+ cells compared to neutrophils in the human lung sections.

Conclusions:

PET imaging with [¹¹C]PBR28 can specifically detect macrophages versus neutrophils during lung inflammation and may be a useful biomarker of macrophage accumulation in lung disease.

Introduction

Macrophages are important regulators of normal innate and adaptive immune responses in the lungs. However, macrophages of particular phenotypes can also contribute to the pathogenesis of chronic lung diseases [1]. The M1/M2 polarization scheme was initially used to describe macrophages responding acutely to endotoxin or interferon-gamma (M1, originally characterized in the context of pathogen response) and alternatively activated M2 macrophages that promote the resolution of inflammation [2]. M2 macrophages, however, have since been shown to promote the development of chronic airway diseases, including chronic obstructive pulmonary disease and asthma [1, 2].

Given the importance of macrophages in lung disease, methods that can measure the lung macrophage burden, track their recruitment and activation kinetics in the lungs, or even distinguish macrophages of different phenotypes would be useful for studying their contribution to lung disease development. Positron emission tomography, a noninvasive, quantitative imaging modality that utilizes molecularly targeted radiolabeled tracers, is ideally suited for developing such methods. Positron emission tomography (PET) imaging with the glucose analog 2-deoxy-2-[¹⁸F]fluoro-D-glucose ([¹⁸F]FDG) has been used to image neutrophil recruitment during lung inflammation [3–5]. However, [¹⁸F]FDG also accumulates in macrophages [6, 7], lymphocytes [8, 9], eosinophils [10], and other structural cell types in the lungs [11, 12] during inflammation. Therefore, more targeted PET tracers for macrophages would be valuable for studying the macrophage contribution to lung disease development.

The translocator protein (TSPO), previously known as the peripheral benzodiazepine receptor (PBR), is upregulated in activated macrophages. N-Butan-2-yl-1-(2-chlorophenyl)-N-methylisoquinoline-[¹¹C]-3-carboxamide ([¹¹C]PK11195), the first tracer developed to image TSPO, has been used in humans to image activated microglia (the macrophages of the brain) in neuroinflammation [13–16] and macrophages associated with vascular disease [17–19]. In rabbit lungs after silica particle challenge, [¹¹C]PK11195 uptake tracked with macrophage recruitment to the lungs and macrophage migration to lymph nodes [20]. Clinically stable individuals with COPD and asthma also have increased [¹¹C]PK11195 activity in the lungs relative to healthy age-matched controls [21]. Therefore, imaging with TSPO-targeted tracers may be a promising approach for quantifying the activated macrophage burden in the lungs.

N-Acetyl-N-(2-[¹¹C]methoxybenzyl)-2-phenoxy-5-pyridinamine ([¹¹C]PBR28) is another TSPO-targeted tracer with 80-fold higher specific binding in the cerebellum than [¹¹C]PK11195 [22]. Additionally, the short half-life of C-11 enabled us to perform comparisons with [¹⁸F]FDG in the same imaging session. Therefore, we chose to evaluate [¹¹C]PBR28 as a marker of macrophage activation in the lungs. As TSPO expression has also been demonstrated in other immune cell types [23, 24], we sought to determine the specificity of [¹¹C]PBR28 for imaging macrophages in vivo using mouse models of acute and chronic lung inflammation. We also assessed cellular TSPO expression by immunostaining to compare with the PET results. These experiments revealed that [¹¹C]PBR28 uptake not only depends on macrophage recruitment but may also depend on macrophage phenotype.

Methods

Reagents and Experimental Groups

The institutional Animal Studies Committee approved all animal studies. All applicable institutional and/or national guidelines for the care and use of animals were followed. The use of banked de-identified human tissues was exempt from Institutional Review Board approval. Wild-type C57BL/6 (Wt) male mice or transgenic Wt/opT male and female mice (5–6 weeks old) were imaged at Day 3 (Wt only) or 49 (Wt and Wt/opT) post-infection with either Sendai virus (SeV, 2 × 10⁵ plaque-forming units) or ultraviolet light-inactivated SeV (UV-SeV) as previously described [25, 26] or at 24 h after intranasal instillation of lipopolysaccharide (LPS, Escherichia coli O26:B6, Sigma Chemical, 10 μg in 30 μl of phosphate buffered saline, Wt mice only). Subject numbers in each group were as follows: N = 3 in D3 Wt control, N = 4 each in D3 and D49 Wt-infected, LPS-treated and untreated control groups, N = 5 in D49 Wt control group, and N = 6 each in the Wt/opT SeV-infected and control groups. Immunostaining was performed on lung sections from imaged Wt mice at Days 3 and 49 post-SeV infection (p.i.). Deidentified paraffin-embedded human lung sections (nine control from tissue from lung nodule resections or donated lungs rejected for transplantation, 23 GOLD stage 4 human COPD sections) were processed as previously described [27, 28] and then stained as described below.

MicroPET/CT Image Acquisition

Mice were imaged on either an Inveon microPET/computed tomography (CT) scanner (Siemens) or a Focus 220 dedicated microPET scanner (Siemens/CTI) with CT obtained after the microPET scan. Both scanners were routinely cross-calibrated against a dose calibrator using a 20-ml cylindrical phantom filled with water and 150 uCi of F-18 and imaged for 20 min. Phantom images were reconstructed with the same protocol as the animal studies using CT-based or transmission scan-based attenuation correction as previously described [29]. Three 0.5 ml aliquots from the water phantom were then measured again in a gamma counter to verify the quantitative accuracy of the image data. Mice were anesthetized with 1–3 % isoflurane for two 60-min dynamic microPET scans, acquired after injecting 10.8 ± 1.6 MBq (292 ± 44 μCi) of [¹¹C]PBR28 followed by 11 ± 2.4 MBq (297 ± 66 μCi) of ¹⁸F-FDG at least 1 h and 20 min (four C-11 half-lives) later.

MicroPET Image Analysis

Images were analyzed using Integrated Research Workflow 4.0 (Siemens) as previously described [29]. The lung uptake at 60-min post-tracer injection was expressed as the percent injected dose per milliliter of lung (%ID/ml) or as a fraction of the time-activity curve peak.

Immunohistochemical Staining

Mouse lung sections were obtained from the same mice used in the microPET imaging studies, with mice sacrificed after imaging and embedded in paraffin as previously described [30]. Mouse and human lung sections were stained for TSPO (Epitomics, clone EPR5384, catalog #EPR5384). Mouse sections were also costained with either Mac3 (BD Biosciences, clone M3/84, cat #550292) or Ly6G (BD Biosciences, clone 1A8, cat #551459) to identify macrophages or neutrophils, respectively, and assessed qualitatively. Human lung sections were costained with CD68 (Dako, clone KP1, cat# M0814) to identify macrophages or neutrophil elastase (NE, Dako, clone NP57, cat# M0752) to identify neutrophils.

For mouse lung sections, after deparaffinization through xylene and ethanol, epitope retrieval was performed by a 20-min boil in an epitope retrieval solution (IHC-Tek, cat #IW-1100). Slides were allowed to cool for 20 min and rinsed with phosphate buffered saline twice for 5 min and blocked with 5 % goat serum in an antibody diluent (IHC-Tek, cat# IW-1001) for 30 min at room temperature. Staining was performed for TSPO using a 1:1000 antibody dilution for 30 min at room temperature, for Ly6G using a 1:100 dilution overnight at 4 °C, and for Mac3 using 1:10 dilution overnight at 4 °C. Secondary antibodies used were goat anti-rat polyclonal IgG antibody conjugated to Alexa Fluor 555 (Molecular Probes, cat #A21434) and goat anti-rabbit polyclonal IgG antibody conjugated to Alexa Fluor 488 (Cell Signaling, cat #4412S), both at 1:500 dilution for 30 min at room temperature.

Deidentified paraffin-embedded human lung sections were obtained from patients with COPD who had consented to donate their lung explants for research purposes at the time of transplantation. Lung sections were also obtained under the same protocol for donated lungs that were rejected for transplantation. Slides were treated as above for deparrafinization. CD68 and TSPO costained slides were boiled for epitope retrieval. Epitope retrieval was not performed for NE and TSPO costained slides. Slides were blocked with an appropriate serum. Staining was performed for NE using a 1:200 antibody dilution for 30 min at room temperature or CD68 using a 1:100 dilution for 30 min at room temperature along with TSPO using a 1:1000 dilution for 30 min at room temperature. Secondary antibodies used were a goat anti-mouse polyclonal IgG antibody conjugated to Alexa Fluor 488 (Molecular Probes, cat #A11001) and a donkey anti-rabbit polyclonal IgG antibody conjugated to Alexa Fluor 555 (Molecular Probes, cat #A31572), both at a 1:500 dilution for 30 min at room temperature. All slides were mounted in Vectashield mounting medium with DAPI for nuclear staining (Vector Labs, cat #H-1200).

Quantification of Immunohistochemical Staining in Human Lung Sections

TSPO expression was quantified using previously published methods [27, 28, 31]. Images were acquired with a Leica DM 5000B microscope, Retiga 200000R Fast 1395 camera, and Q capture Pro51 software (QImaging, Surrey, BC). Three to four randomly selected regions were imaged from 9 healthy and 23 COPD lung sections. Images were collected in red (TSPO) or green (CD68 or NE) channels. The Image-Pro Plus (Media Cybernetics, Rockville, MD) histogram tool was used to quantify the number of red, green, or red and green pixels above a fixed minimum threshold (defined by the threshold that gave visual separation of positive cells from background on three representative slides from each group) that was applied uniformly on all images. Colocalization was calculated as the percentage of red or green (TSPO⁺ or CD68⁺/NE⁺) pixels that were red and green (TSPO⁺ and CD68⁺/NE⁺).

Statistical Analysis

All data were expressed as the mean ± standard deviation. To compare group differences in the %ID/ml or peak fraction of lung [¹¹C]PBR28 and [¹⁸F]FDG uptake in SeV-infected mice, two-way analysis of variance (ANOVA) (SigmaPlot 12 for Windows, Systat Software, Inc) was performed using infection status and mouse type as factors (i.e. Wt D3, Wt D49, and Wt/opT D49 after SeV or UV-killed SeV) with Tukey post hoc testing. T tests using the Mann-Whitney rank sum test were used for the same measures of lung [¹¹C]PBR28 and [¹⁸F]FDG uptake in mice at 24 h after LPS or PBS. Bonferroni correction for multiple comparisons was also applied, yielding p < 0.0125 as the level for significance for the microPET data. Two-way ANOVA with Tukey post hoc analysis was used to test for differences in TSPO staining in human lung sections using patient condition and cell type as factors, with a p value of < 0.05 determining significance.

Results

[¹¹C]PBR28 but not [¹⁸F]FDG Uptake Increases More Prominently with M2 Macrophage-Driven Lung Inflammation

Figure 1 shows representative images (Fig. 1A) and averaged time-activity curves (Fig. 1B) from all mice in each group from the Day 49 post-infection (p.i.) SeV in Wt mice (M2 inflammation, see also Supplemental Figure 1), Day 49 p.i. SeV in Wt/opT mice (attenuated M2 inflammation), and 24 h p.i. LPS in Wt mice (M1 lung inflammation, see also Supplemental Figure 2). [¹¹C]PBR28 uptake increased by the greatest magnitude under M2 conditions in the D49 SeV Wt mice and was partially attenuated in the D49 SeV Wt/opT mice, while the magnitude of increased [¹⁸F]FDG uptake was similar across these same groups. The Wt mice at 24 h after LPS had no increase in [¹¹C]PBR28 uptake but did have increased [¹⁸F]FDG uptake. Wt mice at D3 p.i. of SeV had a small increase in both [¹¹C]PBR28 and [¹⁸F]FDG uptake (images and time-activity curves not shown). CT images also demonstrated increased infiltrates in both lungs in Day 49 p.i. SeV Wt but not in the Wt/opT mice, LPS-treated mice, or Wt mice at D3 p.i. SeV.

Fig. 1.

¹¹C-PBR28 and ¹⁸F-FDG uptake in acute and chronic inflammation. MicroPET images and time-activity curves for Sendai virus ((+) SeV) and ultraviolet-inactivated SeV ((−) SeV) infected wild-type (Wt) mice at Days 49 post-infection (p.i.) or in monocyte-deficient Wt/opT mice at D49 p.i. and Wt mice at 24 h after lipopolysaccharide (LPS) instillation. A. Representative images from the last 10 min of a 60-min dynamic acquisition are shown for Wt and Wt/opT mice at D49 p.i. and Wt mice at 24 h after LPS, after injecting ¹¹C-PBR28 followed by ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG). B. Mean time-activity curves of infected mice and uninfected controls at D49 and D49 after infection. Standard deviation bars shown only in one direction for clarity with some bars not visible outside of the symbol.

Quantification of the microPET data showed significantly increased [¹¹C]PBR28 %ID/ml in the lungs of D49 SeV Wt mice when compared to controls (Fig. 2). This uptake was also significantly increased compared to both D3 SeV mice and D49 SeV in Wt/opT mice. When expressed as the fraction of the peak of the time-activity curve, to account for differences in initial blood flow, [¹¹C]PBR28 was still significantly increased in D49 SeV Wt mice compared to both controls and D3 SeV Wt mice. However, [¹¹C]PBR28 uptake as fraction of the peak was also significantly increased in D49 SeV Wt/opT mice when compared to D49 Wt/opT controls. In contrast, [¹⁸F]FDG %ID/ml, from images acquired in same mice immediately after [¹¹C]PBR28 scans, significantly increased by the same magnitude in all of these groups, regardless of the type of inflammation, with uptake in D3 SeV Wt mice being significantly higher than that observed in D49 SeV Wt/opT mice. [¹¹C]PBR28 uptake did not increase with LPS administration, regardless of quantitative measure used, whereas [¹⁸F]FDG uptake increased consistently, with significance achieved when expressed as fraction of the peak (Fig. 2B).

Fig. 2.

Quantification of microPET data, with uptake expressed as % injected dose per milliliter (%ID/ml, A) or as the fraction of the peak of the time-activity curve (B). MicroPET images were obtained from Sendai virus ((+) SeV) and ultraviolet-inactivated SeV ((−) SeV) infected wild-type (Wt) mice at Days 3 and 49 post-infection (p.i.), monocyte-deficient transgenic Wt/opT mice at D49 p.i., and Wt mice at 24 h after lipopolysaccharide (LPS) instillation. Bars with p values on the graph indicate comparisons of (+) SeV data between groups. Individual symbols above black columns indicate p values compared to (−) SeV controls within the same group as follows: *p < 0.001; ^†p = 0.002; **p = 0.029. N = 3 in D3 Wt control, N = 4 each in D3 and D49 Wt infected, LPS-treated and untreated control groups, N = 5 in D49 Wt control group, and N = 6 each in the Wt/opT SeV-infected and control groups.

TSPO Expression Colocalizes to Macrophages but not Neutrophils in Both Mouse and Human Lung Sections

Qualitative assessment of immunohistochemical staining in mouse lung sections at D3 and D49 p.i. SeV in Wt mice confirmed that TSPO expression was predominantly in Mac3⁺ but not in Ly6G⁺ cells (Fig. 3, Supplemental Figures 3 and 4). Nearly all Mac3⁺ cells stained positively for TSPO under all conditions, with Mac3⁺ cells being the most numerous at Day 49 p.i. SeV. Ly6G⁺ cells were more apparent at D3 than at D49 p.i. SeV. Ly6G⁺ cells had very little TSPO colocalization at any time point except in Wt mice at Day 49 p.i. in which Ly6G^lo cells were observed with low level TSPO expression. These Ly6G^loTSPO⁺ cells were most likely dendritic cells.

Fig. 3.

Immunohistochemical assessment of translocator protein (TSPO) expression with and without Sendai virus (SeV) infection in mice imaged by microPET. Representative images demonstrate that TSPO colocalized to macrophages (Mac3) but not neutrophils (Ly6G) in lung sections taken from mice imaged at either Day 3 or Day 49 post-infection (p.i.) with Sendai virus (SeV). Yellow arrow denotes a Ly6g^loTSPO⁺ cell that is most likely a dendritic cell. Blue represents DAPI staining. See Supplement Figures 3 and 4 for single-channel images.

Sections taken from the lungs of subjects without known COPD and from COPD lung explants demonstrated that TSPO staining colocalized to CD68⁺ but not neutrophil elastase positive (NE⁺) cells (Fig. 4, Supplemental Figure 5). Quantification demonstrated significantly more TSPO expression in CD68⁺ cells than in NE⁺ cells. The total amount of TSPO⁺, CD68⁺, and NE⁺ pixels did not differ between the non-COPD and COPD lung sections. Since lung transplants that are not suitable for transplantation are expected to have inflammation present, they provide a non-COPD-related lung inflammation control.

Fig. 4.

Human lung sections costained for the translocator protein (TSPO) and either neutrophil elastase (NE) or CD68 along with DAPI nuclear staining from healthy volunteers or participants with GOLD Stage 4 COPD. A. Representative images of human lung section stains with insets showing magnified view of colocalization of NE or CD68 with TSPO staining. Blue represents DAPI staining. See Supplement Figure 5 for single-channel images. B. Quantification of TSPO within different cell types and percentage of cell types within TSPO-positive cells. ⇀ p < 0.001, N = 32 (sections from all individuals). **p < 0.001, sections from N = 9 non-COPD lungs and N = 23 COPD lungs.

Discussion

TSPO is expressed abundantly in activated macrophages [17, 20] and thus has been an attractive target for PET imaging approaches aimed at quantifying the macrophage burden in various organs, with [¹¹C]PBR28 one of the TSPO tracers demonstrating high in vivo specific binding. Therefore, we designed this study to determine whether changes in [¹¹C]PBR28 uptake and TSPO levels would depend specifically on macrophage recruitment in mouse models of lung inflammation. In a well-characterized model of acute and chronic inflammation after SeV infection, we demonstrated that [¹¹C]PBR28 increased nearly 3-fold at Day 49 post-infection, a time at which M2-polarized macrophages are known to predominate [25]. [¹¹C]PBR28 uptake did not increase by the same magnitude at D49 in SeV-infected Wt/opT mice, in line with the known attenuated chronic inflammatory response in these monocyte-deficient mice [30], or in Wt mice at Day 3 post-infection, when neutrophils are most abundant but are also accompanied by a smaller increase in macrophage recruitment [26]. The immunostaining in the mouse lung sections from these imaged mice confirmed that TSPO expression localized primarily to macrophages. Additionally, [¹¹C]PBR28 uptake did not change at all after endotoxin instillation, which also causes a predominantly neutrophilic, M1 macrophage-driven inflammatory response. [¹⁸F]FDG uptake increased to a similar magnitude under all conditions tested, confirming its known non-specificity for the types of activated immune cells present. These data together suggest that changes in [¹¹C]PBR28 uptake in fact depend on specific recruitment of M2 macrophage.

Prior human studies with TSPO tracers also suggest an association of tracer uptake with macrophage phenotype. Though studies in subjects with COPD or asthma showed increased lung uptake of [¹¹C]PK11195, subjects with scleroderma fibrosing alveolitis have decreased uptake [21, 32]. [³H]PK11195 binding of alveolar macrophages obtained by bronchoalveolar lavage (BAL) from patients with various interstitial lung diseases, including scleroderma fibrosing alveolitis, sarcoidosis, and nonspecific and usual interstitial pneumonia, was also lower compared to those obtained from healthy volunteers [33]. More recent studies in patients with rheumatoid arthritis demonstrate that activated macrophages from the synovium have increased TSPO expression compared to quiescent non-activated macrophages [34]. Furthermore, in vitro studies of M1 and M2 polarization in healthy volunteer peripheral blood monocytes demonstrated downregulation of TSPO in M1 conditions but demonstrated no change under M2 conditions [35]. When applied to our mouse model data, these findings suggest that downregulation of TSPO in recruited M1 macrophages may result in a net-zero change in [¹¹C]PBR28 uptake in M1 conditions after endotoxin instillation, while maintenance of TSPO expression in M2 macrophages could explain increased [¹¹C]PBR28 uptake in chronic inflammation after SeV infection due to increased recruitment of M2 macrophages, as has been shown previously in this model [25, 30].

Interestingly, in the D3 SeV-infected Wt mice, a time point at which M1 conditions also predominate, we observed significantly increased [¹¹C]PBR28 but not [¹⁸F]FDG uptake. Two prior studies have also demonstrated that TSPO plays varying roles in viral replication for the human immunodeficiency virus-1 and myxoma poxvirus [36, 37]. Therefore, we hypothesize that our observations reflect the acute murine lung response to viral infection. However, given that [¹⁸F]FDG uptake did not increase, we would further hypothesize that the use of both of these tracers could help distinguish acute viral infection from chronic inflammation. This will need to be tested in future studies to determine the validity of these hypotheses.

In contrast to the above human studies and our own results, rodent studies have shown increased uptake of other reported TSPO tracers, N-benzyl-N-methyl-2-[7,8-dihydro-7-(2-¹⁸F-fluoroethyl)-8-oxo-2-phenyl-9H-purin-9-yl]acetamide ([¹⁸F]FEDAC) and N,N-diethyl-2-(2-(4-[¹²³I]-methoxyphenyl)-5,7-dimethylpyrazolo(1,5-α)pyrimidin-3-yl)acetamide ([¹²³I]DPA713), under M1 conditions at 24 h after endotoxin instillation with endotoxin doses that were similar to those used in our study [38, 39]. TSPO staining was also shown in that study to be present in both neutrophils and macrophages after endotoxin instillation in these studies, but TSPO co-staining with cell type-identifying markers was not performed [38]. The authors of these reports concluded that imaging TSPO levels would provide a nonspecific measure of lung inflammation. However, our lung section staining results demonstrated greater TSPO positivity associated with monocyte lineage cells in both mouse and human lung sections. Furthermore, the different TSPO ligands that have been developed have varying affinities for the low, medium, and high binding affinity sites. For example, when compared to DPA713, PBR28 has a much greater difference in affinity between the low and high binding sites, suggesting potential differences in in vivo behavior [40]. Our data thus suggest that, despite results seen in other analogous TSPO-targeted tracers, [¹¹C]PBR28 may distinguish macrophages specifically.

Interestingly, in our data, TSPO positivity was associated with both CD68⁺ and CD68⁻ pixels. Alveolar type II cells in the mouse lung sections express TSPO [41, 42] and could have contributed to these TSPO⁺/CD68⁻ pixels. The fact that the number of these TSPO⁺CD68⁻ pixels was approximately the same in both the COPD and non-COPD lung sections suggests that the TSPO expression in non-macrophage cells does not change significantly in the presence of COPD-related inflammation. Furthermore, we also observed that the number of TSPO⁺CD68⁺ pixels also did not vary between the COPD and non-COPD lung sections. As the non-COPD lung sections were taken from donated lung tissue that was rejected for transplant or from normal-appearing lung adjacent to lung nodules, frequently a component of inflammation is present in these lungs. More importantly, our human lung section data clearly demonstrated that TSPO is not associated with neutrophils as there was very little colocalization of NE with TSPO in either group. Therefore, taken together, our staining results further suggest that [¹¹C]PBR28 uptake in the lungs is likely to depend on the recruitment of macrophages. Further studies will be needed to verify the association of [¹¹C]PBR28 lung uptake in humans with pulmonary recruitment of M2 macrophages.

In conclusion, our data suggest that [¹¹C]PBR28 possesses unique properties that enable imaging of macrophages specifically in M2-driven chronic lung inflammation and not during M1-driven acute lung inflammation. These data require further validation in human trials; however, if so validated, this approach would be the first available to distinguish phenotypically different macrophages in patients with inflammatory lung disease.

Supplementary Material

Supplementary Information. The online version contains supplementary material available at https://doi.org/10.1007/s11307-021-01617-w.