Natalia Onishchenko

It is known that nanoparticles functionalized with mixed charge monolayers – that is, monolayers composed of both positively [+] and negatively [-] charged ligands – can induce selective death of cancer cells. Remarkably, however, when the ligand polarities and proportions are kept constant, but the chemical nature of the “minority” [-] component is altered, the nanoparticles exhibit distinct intracellular localization patterns and induce different forms of iron-dependent lipid-peroxidation-driven cell death, which we term “ferroptosis-like” due to concurrent activation of caspases. In particular, changing the end groups on the [-] ligands from carboxylic to sulfonic to phosphonic acid shifts nanoparticle accumulation from lysosomes to autophagic compartments to late endosomes, respectively. These distinct cellular outcomes are accompanied by marked differences in the composition of serum protein coronas adsorbed onto the mixed-charge nanoparticles. Proteomic analysis revealed exceedingly complex corona proteomes, comprising more than 800 different proteins, many enriched hundreds- to thousands-fold compared to blood serum, with numerous proteins selectively enriched on specific nanoparticle types. Together, these findings identify charge-imbalanced constructs as potential modulators of subcellular organelle targeting and ferroptosis-like cell death.

The utility of enzymes in organic synthesis is constrained by the limited ability to predict the scope of small‐molecule substrates that a given enzyme can act upon. The problem has proven challenging to both classical computational‐chemistry approaches and to modern Machine Learning (ML) methods, the latter suffering from the combination of data scarcity (including all‐important negative examples) and the inaccuracy of chemoinformatic vectorization schemes. The current work addresses both problems, deploying affordable chemical robotics to curate a structurally diverse set of both active and inactive substrates, and then using this data to develop a physics‐grounded ML model for substrate scope prediction. This model uses only a handful of features to learn the proper balance between steric and electronic factors even from small datasets. It maintains useful predictive power on external enzyme families and shows improved out‐of‐distribution performance compared with descriptor‐heavy state‐of‐the‐art ML algorithms.

Despite the undisputable role of the protein corona in the biointeractions of liposome drug carriers, the field suffers from a lack of knowledge regarding the patterns of protein deposition on lipid surfaces with different compositions. Here, we investigated the protein coronas formed on liposomes of basic compositions containing combinations of egg phosphatidylcholine (PC), palmitoyloleoyl phosphatidylglycerol (POPG), and cholesterol. Liposome−protein complexes isolated by size-exclusion chromatography were delipidated and analyzed using label-free LC-MS/MS. The addition of the anionic lipid and cholesterol both affected the relative protein abundances (and not the total bound proteins) in the coronas. Highly anionic liposomes, namely those containing 40% POPG, carried corona enriched with cationic proteins (apolipoprotein C1, beta-2-glycoprotein 1, and cathelicidins) and were the least stable in the calcein release assay. Cholesterol improved the liposome stability in the plasma. However, the differences in the corona compositions had little effect on the liposome uptake by endothelial (EA.hy926) and phagocytic cells in the culture (U937) or ex vivo (blood-derived monocytes and neutrophils). The findings emphasize that the effect of protein corona on the performance of the liposomes as drug carriers occurs through compromising particle stability rather than interfering with cellular uptake.

Previously, we showed in the human umbilical vein endothelial cells (HUVECs) model that a liposome formulation of melphalan lipophilic prodrug (MlphDG) decorated with selectin ligand tetrasaccharide Sialyl Lewis X (SiaLe^X) undergoes specific uptake by activated cells and in an in vivo tumor model causes a severe antivascular effect. Here, we cultured HUVECs in a microfluidic chip and then applied the liposome formulations to study their interactions with the cells in situ under hydrodynamic conditions close to capillary blood flow using confocal fluorescent microscopy. The incorporation of 5 to 10% SiaLe^X conjugate in the bilayer of MlphDG liposomes increased their consumption exclusively by activated endotheliocytes. The increase of serum concentration from 20 to 100% in the flow resulted in lower liposome uptake by the cells. To elucidate the possible roles of plasma proteins in the liposome–cell interactions, liposome protein coronas were isolated and analyzed by shotgun proteomics and immunoblotting of selected proteins. Proteomic analysis showed that a gradual increase in SiaLe^X content correlated with the overall enrichment of the liposome-associated proteins with several apolipoproteins, including the most positively charged one, ApoC1, and serum amyloid A4, associated with inflammation, on the one hand, and a decrease in the content of bound immunoglobulins, on the other. The article discusses the potential interference of the proteins in the binding of liposomes to selectins of endothelial cells.

Liposomes as drug carriers are usually injected into the systemic circulation where they are instantly exposed to plasma proteins. Liposome–protein interactions can affect both the stability of liposomes and the conformation of the associated protein leading to the altered biodistribution of the carrier. In this work, mutual effects of albumin and liposomal membrane in the course of the protein’s adsorption were examined in terms of quantity of bound protein, its structure, liposome membrane permeability, and changes in physicochemical characteristics of the liposomes. Fluorescence spectroscopy methods and Fourier transform infrared spectroscopy (ATR-FTIR), which provides information about specific groups in lipids involved in interaction with the protein, were used to monitor adsorption of albumin with liposomes based on egg phosphatidylcholine with various additives of negatively charged lipidic components, such as phosphatidylinositol, ganglioside GM₁, or the acidic lipopeptide. Less than a dozen of the protein molecules were tightly bound to a liposome independently of bilayer composition, yet they had a detectable impact on the bilayer. Albumin conformational changes during adsorption were partially related to bilayer microhydrophobicity. Ganglioside GM₁ showed preferable features for evading undesirable structural changes.

To assess the stability and efficiency of liposomes carrying a phospholipase A2-sensitive phospholipid-allocolchicinoid conjugate (aC-PC) in the bilayer, egg phosphatidylcholine and 1-palmitoyl-2-oleoylphosphatidylglycerol-based formulations were tested in plasma protein binding, tubulin polymerization inhibition, and cytotoxicity assays. Liposomes L-aC-PC10 containing 10 mol. % aC-PC in the bilayer bound less plasma proteins and were more stable in 50% plasma within 4 h incubation, according to calcein release and FRET-based assays. Liposomes with 25 mol. % of the prodrug (L-aC-PC25) were characterized by higher storage stability judged by their hydrodynamic radius evolution yet enhanced deposition of blood plasma opsonins on their surface according to SDS-PAGE and immunoblotting. Notably, inhibition of tubulin polymerization was found to require that the prodrug should be hydrolyzed to the parent allocolchicinoid. The L-aC-PC10 and L-aC-PC25 formulations demonstrated similar tubulin polymerization inhibition and cytotoxic activities. The L-aC-PC10 formulation should be beneficial for applications requiring liposome accumulation at tumor or inflammation sites.

Previously, a liposomal formulation of a chemotherapeutic agent melphalan (Mlph) incorporated in a fluid lipid bilayer of natural phospholipids in the form of dioleoylglyceride ester (MlphDG) was developed and the antitumor effect was confirmed in mouse models. The formulation composed of egg phosphatidylcholine (ePC), soybean phosphatidylinositol (PI), and MlphDG (8:1:1, by mol) showed stability in human serum for at least 4–5 h. On the contrary, replacing PI with pegylation of the liposomes, promoted fast dissociation of the components from the bilayer. In this work, interactions of MlphDG-liposomes with the most abundant plasma protein—albumin—in function of the presence of PI in the formulation were explored using Fourier transform infrared spectroscopy. The release of MlphDG from the liposomes was studied by asymmetrical flow field-flow fractionation (AF4) using micelles formed by a polyethylene glycol conjugate with phosphatidylethanolamine to mimic the physiological lipid sink like lipoproteins. Our results show that PI actually protects the membrane of MlphDG-liposomes from the protein penetration, presumably due to pairing between the positively charged MlphDG and negatively charged PI, which compensates for the heterogeneity of the lipid bilayer. The AF4 technique also evidences high stability of the formulation as a drug carrier.