Targeted delivery of anti-inflammatory osteoarthritis treatments have the potential to significantly decrease undesirable systemic side effects and reduce required therapeutic dosage. Here we present a targeted, non-invasive drug delivery system to decrease inflammation in an osteoarthritis model. Hollow thermoresponsive poly(N-isopropylacrylamide) (pNIPAM) nanoparticles have been synthesized via degradation of a N,N′-bis(acryloyl)cystamine (BAC) cross-linked core out of a non-degradable pNIPAM shell. Sulfated 2-acrylamido-2-methyl-1-propanesulfonic acid (AMPSA) was copolymerized in the shell to increase passive loading of an anti-inflammatory mitogen-activated protein kinase-activated protein kinase 2 (MK2)-inhibiting cell-penetrating peptide (KAFAK). The drug-loaded hollow nanoparticles were effective at delivering a therapeutically active dose of KAFAK to bovine cartilage explants, suppressing pro-inflammatory interleukin-6 (IL-6) expression after interleukin-1 beta (IL-1β) stimulation. This thermosensitive hollow nanoparticle system provides an excellent platform for the delivery of peptide therapeutics into highly proteolytic environments such as osteoarthritis.

Peripheral artery disease is an atherosclerotic stenosis in the peripheral vasculature that is typically treated via percutaneous transluminal angioplasty. Deployment of the angioplasty balloon damages the endothelial layer, exposing the underlying collagen and allowing for the binding and activation of circulating platelets which initiate an inflammatory cascade leading to eventual restenosis. Here, we report on collagen-binding sulfated poly(N-isopropylacrylamide) nanoparticles that are able to target to the denuded endothelium. Once bound, these nanoparticles present a barrier that reduces cellular and platelet adhesion to the collagenous surface by 67% in whole blood and 59% in platelet-rich plasma under biologically relevant shear rates. In vitro studies indicate that the collagen-binding nanoparticles are able to load and release therapeutic quantities of anti-inflammatory peptides, with the particles reducing inflammation in endothelial and smooth muscle cells by 30% and 40% respectively. Once bound to collagen, the nanoparticles increased endothelial migration while avoiding uptake by smooth muscle cells, indicating that they may promote regeneration of the damaged endothelium while remaining anchored to the collagenous matrix and locally releasing anti-inflammatory peptides into the injured area. Combined, these collagen-binding nanoparticles have the potential to reduce inflammation, and the subsequent restenosis, while simultaneously promoting endothelial regeneration following balloon angioplasty.

ABSTRACT. Peripheral artery disease is an atherosclerotic occlusion in the peripheral vasculature that is typically treated via percutaneous transluminal angioplasty. Unfortunately, deployment of the angioplasty balloon damages the endothelial layer, exposing the underlying collagen and allowing for the binding and activation of circulating platelets, which initiate an inflammatory cascade leading to eventual restenosis. Here, we report on the development of poly(NIPAm-MBA-AMPS-AAc) nanoparticles that have a collagen I-binding peptide crosslinked to their surface allowing them to bind to exposed collagen. Once bound, these particles mask the exposed collagen from circulating platelets, effectively reducing collagen-mediated platelet activation. Using collagen I-coated plates, we demonstrate that these particles are able to bind to collagen at concentrations above 0.5 mg/mL. Once bound, these particles inhibit collagen-mediated platelet activation by over 60%. Using light scattering and zeta potential measurements, we investigated the potential of the nanoparticles as a drug delivery platform. We have verified that the collagen-binding nanoparticles retain the temperature sensitivity common to poly(NIPAm)-based nanoparticles while remaining colloidally stable in aqueous environments. We also demonstrate that they are able to passively load and release anti-inflammatory cell penetrating peptides. Combined, we have developed a collagen-binding nanoparticle that has dual therapy potential, preventing collagen-mediated platelet activation while delivering water-soluble therapeutics directly to the damaged area.

Cell penetrating peptides (CPP) have been widely used to increase the cellular delivery of their associated cargo. Multiple modes of uptake have been identified, however they cannot be predicted apriori. Elucidating these mechanisms is important for understanding peptide function as well as further optimizing cellular delivery. We have developed a class of MK2 inhibitor peptides, named FAK and YARA that utilize CPP domains to gain cellular access. In this study, we investigate the mechanism of endocytosis of these MK2 inhibitors by examining the uptake of fluorescently labeled peptide in human monocyte (THP-1) and mesothelial cells, and looking for colocalization with known markers of endocytosis. Our results indicate that uptake of the MK2 inhibitors was minimally enhanced by the addition of the fluorescent label, and that the type of endocytosis used by the inhibitor depends on several factors including concentration, cell type, and which CPP was used. We found that in THP-1 cells uptake of YARA occurred primarily via macropinocytosis, while FAK entered via all three mechanisms of endocytosis examined in this study. In mesothelial cells, uptake of YARA occurred via caveolae-mediated endocytosis, but became less specific at higher concentrations; while uptake of FAK occurred through clathrin-mediated endocytosis. In all cases, the delivery resulted in active inhibition of MK2. In summary, the results support endocytic uptake of fluorescently labeled FAK and YARA in two different cell lines, with the mechanism of uptake dependent on extracellular concentration, cell type, and choice of CPP.

This paper describes a novel application of NLP models to detect denial of service attacks using only social media as evidence. Individual networks are often slow in reporting attacks, so a detection system from public data could better assist a response to a broad attack across multiple services. We explore NLP methods to use social media as an indirect measure of network service status. We describe two learning frameworks for this task: a feed-forward neural network and a partially labeled LDA model. Both models outperform previous work by significant margins (20% F1 score). We further show that the topic-based model enables the first fine-grained analysis of how the public reacts to ongoing network attacks, discovering multiple "stages" of observation. This is the first model that both detects network attacks (with best performance) and provides an analysis of when and how the public interprets service outages. We describe the models, present experiments on the largest twitter DDoS corpus to date, and conclude with an analysis of public reactions based on the learned model's output.