Targeted imaging of very late antigen-4 for noninvasive assessment of lung inflammation-fibrosis axis

Abstract

Background: The lack of noninvasive methods for assessment of dysregulated inflammation as a major driver of fibrosis (i.e., inflammation-fibrosis axis) has been a major challenge to precision management of fibrotic lung diseases. Here, we determined the potential of very late antigen-4 (VLA-4)-targeted positron emission tomography (PET) to detect inflammation in a mouse model of bleomycin-induced fibrotic lung injury.

Method: Single time-point and longitudinal VLA-4-targeted PET was performed using a high-affinity peptidomimetic radiotracer, ⁶⁴Cu-LLP2A, at weeks 1, 2, and 4 after bleomycin-induced (2.5 units/kg) lung injury in C57BL/6J mice. The severity of fibrosis was determined by measuring the hydroxyproline content of the lungs and expression of markers of extracellular matrix remodeling. Flow cytometry and histology was performed to determine VLA-4 expression across different leukocyte subsets and their spatial distribution.

Results: Lung uptake of ⁶⁴Cu-LLP2A was significantly elevated throughout different stages of the progression of bleomycin-induced injury. High lung uptake of ⁶⁴Cu-LLP2A at week-1 post-bleomycin was a predictor of poor survival over the 4-week follow up, supporting the prognostic potential of ⁶⁴Cu-LLP2A PET during the early stage of the disease. Additionally, the progressive increase in ⁶⁴Cu-LLP2A uptake from week-1 to week-4 post-bleomycin correlated with the ultimate extent of lung fibrosis and ECM remodeling. Flow cytometry revealed that LLP2A binding was restricted to leukocytes. A combination of increased expression of VLA-4 by alveolar macrophages and accumulation of VLA-4-expressing interstitial and monocyte-derived macrophages as well as dendritic cells was noted in bleomycin-injured, compared to control, lungs. Histology confirmed the increased expression of VLA-4 in bleomycin-injured lungs, particularly in inflamed and fibrotic regions.

Conclusions: VLA-4-targeted PET allows for assessment of the inflammation-fibrosis axis and prediction of disease progression in a murine model. The potential of ⁶⁴Cu-LLP2A PET for assessment of the inflammation-fibrosis axis in human fibrotic lung diseases needs to be further investigated.

Keywords: Inflammation; Lung fibrosis; Molecular imaging; PET; Very late antigen-4.

Fig. 1.

⁶⁴Cu-LLP2A PET/CT in bleomycin-induced lung injury. Representative axial CT, PET, and co-registered PET/CT (A) of control versus bleomycin-treated mice acquired ~ 4 h after ⁶⁴Cu-LLP2A administration demonstrate significant increases in the tracer uptake during different stages of fibrotic lung injury. The specificity of tracer uptake is confirmed by blocking ⁶⁴Cu-LLP2A uptake, despite the presence of CT evidence of lung injury, by co-injection of excess non-labeled LLP2A (right panels in A). PET-derived quantification, measured as %ID/mL_mean (B) and %ID/mL_max (C), confirms significant increases in ⁶⁴Cu-LLP2A uptake throughout different stages of the progression of bleomycin-induced lung injury as well its near-complete blockade by co-injection of non-labeled LLP2A. N = 14 (control), 22 (1-week), 10 (2-week), 8 (4-week), and 4 (2-week blocked). PBS = phosphate-buffered saline

Fig. 2

Biodistribution of ⁶⁴Cu-LLP2A. Ex vivo γ-counting (A) confirms increased lung uptake of ⁶⁴Cu-LLP2A at 2 and 4 weeks after bleomycin-induced fibrotic injury, which is blocked by co-injection of excess unlabeled LLP2A (ex vivo γ-counting was not performed at 1-week post-bleomycin as this timepoint was only assessed as part of the longitudinal PET/CT experiments). There are robust correlations between PET-derived (B ID/mL_mean and C %ID/mL_max) and γ-counting-derived measures of ⁶⁴Cu-LLP2A uptake, confirming the accuracy of noninvasive quantification of tracer uptake. γ-counting (D) and the visual assessment of whole-body PET (E) images demonstrate specific uptake of ⁶⁴Cu-LLP2A in organs rich in VLA-4-expressing cells, including spleen, thymus, and bone marrow. However, ⁶⁴Cu-LLP2A uptake in these organs is not significantly affected by bleomycin administration. N = 6 (control), 6 (2-week), 8 (4-week), and 4 (2-week blocked)

Fig. 3

Prognostication of bleomycin-induced lung injury by ⁶⁴Cu-LLP2A PET. High lung uptake of ⁶⁴Cu-LLP2A at 1-week post-bleomycin, defined as the uptake above the median %ID/mL_mean (left panel) or %ID/mL_max (right panel), predicts a significantly worse survival over the course of 4 weeks (A). Similarly, both %ID/mL_mean (left panel) and %ID/mL_max (right panel) of ⁶⁴Cu-LLP2A uptake at 1-week post-bleomycin strongly predict fatal lung injury as assessed by ROC curve analysis (B). The progressive increase in lung uptake of ⁶⁴Cu-LLP2A from week-1 to week-4 post-bleomycin correlates with the severity of fibrosis and ECM remodeling as determined by the hydroxyproline content of the lungs and mRNA expression of markers of matrix remodeling at week-4, including collagen 1 (Col1a1), fibronectin (Fn1), lysyl oxidase (Lox) and lysyl oxidase-like 2 (Loxl2) (C). N = 21 for panels A and B, N = 8 for panel C

Fig. 4

Flow cytometric quantification of LLP2A binding to different lung leukocytes. Representative histograms from flow cytometry of enzymatically dissociated murine lung single cells (A) demonstrate negligible binding of LLP2A-Biotin to CD45-negative cells (top panel) while different levels of binding to LLP2A-Biotin are present in CD45-positive leukocytes (bottom panel). The specificity of LLP2A-Biotin binding is confirmed by near-complete blocking of its binding in cells co-incubated with excess non-fluorescent LLP2A. Cell-type specific analysis of the flow cytometry data demonstrated that LLP2A-Biotin binding nearly doubles in alveolar macrophage (aMφ) of bleomycin-injured lungs compared to control lungs, while is not significantly different compared to control lungs in the remaining leukocyte subsets (B). There is a significant increase in the abundance of interstitial and monocyte-derived macrophage (iMφ & MDMφ) as well as dendritic cells (DC) at 1-week post-bleomycin injury versus controls (C). N = 5 (control group) and 6 (bleomycin group)

Fig. 5

Histological assessment of VLA-4 expression in murine lungs with bleomycin-induced fibrosis. Representative low-magnification (top panels) and high-magnification (bottom panels) images demonstrate a marked increase in the binding of LLP2A-Cy3 (red) in fibrotic lungs of the mice at 4 weeks after bleomycin administration compared to the control lungs. The specificity of LLP2A-Cy3 binding is confirmed by the blocking of LLP2A-Cy3 signal in tissues co-incubated with excess non-fluorescent LLP2A (the insets in the right lower corners of high-magnification images). Tissues are counterstained with DAPI nuclear staining (blue). N = 3 per group