Supramolecular Organization Predicts Protein Nanoparticle Delivery to Neutrophils for Acute Lung Inflammation Diagnosis and Treatment

Jacob W. Myerson; Priyal N. Patel; Nahal Habibi; Landis R. Walsh; Yi-Wei Lee; David C. Luther; Laura T. Ferguson; Michael H. Zaleski; Marco E. Zamora; Oscar A. Marcos-Contreras; Patrick M. Glassman; Ian Johnston; Elizabeth D. Hood; Tea Shuvaeva; Jason V. Gregory; Raisa Y. Kiseleva; Jia Nong; Kathryn M. Rubey; Colin F. Greineder; Samir Mitragotri; George S. Worthen; Vincent M. Rotello; Joerg Lahann; Vladimir R. Muzykantov; Jacob S. Brenner

Grouping the 23 nanoparticles and three proteins according to the classes defined in Figure 3 and Supplementary Figures 8-13 , we completed a linear discriminant analysis of the naïve-to-inflamed shift for particle retention in the lungs, blood, liver, and spleen (Supplementary Figure 20a) . Data for particle uptake in each organ was normalized by subtracting and then dividing by the mean uptake over all particles. The first two eigenvectors, dominated by splenic uptake and a combination of liver and lung uptake, respectively, accounted for 96% of variation in the data. The resulting projection of the data along the first two linear discriminant analysis eigenvectors was analyzed by k-means clustering to confirm the classes of nanoparticle with specificity for the inflamed lungs (Supplementary Figure 20b) . Indeed, division of the data into two clusters supported the delineation of the 13 nanoparticles with specificity for inflamed lungs. NAPs, nanoparticles based on protein gelation, crosslinking, and charge association, all aligned in one cluster.