SummaryDuring morphogenesis of mature HIV-1 cores, the viral capsid (CA) proteins assemble conical or tubular shells around the viral ribonucleoprotein complexes. This assembly step is mimicked in vitro through reactions in which CA proteins oligomerize to form long tubes, and this process can be modeled as consisting of a slow nucleation period followed by a rapid phase of tube growth. We have developed a novel fluorescence microscopy approach to monitor in vitro assembly reactions and have employed it, along with electron microscopy analysis, to characterize the assembly process. Our results indicate that temperature, salt concentration, and pH changes have differential effects on tube nucleation and growth steps. We also demonstrate that assembly can be unidirectional or bidirectional, that growth can be capped, and that proteins can assemble onto the surfaces of tubes, yielding multiwalled or nested structures. Finally, experiments show that a peptide inhibitor of in vitro assembly also can dismantle pre-existing tubes, suggesting that such reagents may possess antiviral effects against both viral assembly and uncoating. Our investigations help establish a basis for understanding the mechanism of mature HIV-1 core assembly, and avenues for antiviral inhibition.

Coexisting liquid phases of model membrane systems are chemically identified using imaging time-of-flight secondary ion mass spectrometry (TOF-SIMS). The systems studied were Langmuir-Blodgett (LB) model membranes of cholesterol (CH) with two different phospholipids, one a major component in the outer plasma membrane bilayer leaflet (dipalmitoylphosphatidylcholine (PC)) and the other a major component in the inner leaflet (dipalmitoylphosphatidylethanolamine (PE)). Binary mixtures of CH with each of the phospholipids were investigated, as well as a ternary system. A single homogeneous phase is evident for PC/CH, whereas both systems containing PE show lateral heterogeneity with phospholipid-rich and CH-rich regions. The interaction between CH and the two phospholipids differs due to the disparity between the phospholipid headgroups. Imaging TOF-SIMS offers a novel opportunity to chemically identify and differentiate the specific membrane locations of CH and phospholipid in membrane regions without the use of fluorescent dyes. This unique imaging method has been used to demonstrate the formation of micrometer-size CH domains in phosphatidylethanolamine-rich systems and is further evidence suggesting that CH may facilitate transport and signaling across the two leaflets of the plasma membrane.

Chemical structure extraction from documents remains a hard problem due to both false positive identification of structures during segmentation and errors in the predicted structures. Current approaches rely on handcrafted rules and subroutines that perform reasonably well generally, but still routinely encounter situations where recognition rates are not yet satisfactory and systematic improvement is challenging. Complications impacting performance of current approaches include the diversity in visual styles used by various software to render structures, the frequent use of ad hoc annotations, and other challenges related to image quality, including resolution and noise. We here present end-to-end deep learning solutions for both segmenting molecular structures from documents and for predicting chemical structures from these segmented images. This deep learning-based approach does not require any handcrafted features, is learned directly from data, and is robust against variations in image quality and style. Using the deep-learning approach described herein we show that it is possible to perform well on both segmentation and prediction of low resolution images containing moderately sized molecules found in journal articles and patents.

The location of each lipid in a palmitoyloleoylphosphatidylcholine/18:0 sphingomyelin/cholesterol monolayer system is laterally resolved using imaging time-of-flight secondary ion mass spectrometry (TOF-SIMS) without the necessity of adding fluorescent labels. This system of coexisting immiscible liquid phases shows cholesterol domains with sizes and shapes comparable to those in the fluorescence microscopy literature. The results show that SM localizes with cholesterol and that palmitoyloleoylphosphatidylcholine is excluded. Moreover, the segregation is not complete, and there is a small amount of both phospholipids distributed throughout.

Understanding the influence of molecular environment on phospholipids is important in time-of-flight secondary ion mass spectrometry (TOF-SIMS) studies of complex systems such as cellular membranes. Varying the molecular environment of model membrane Langmuir-Blodgett (LB) films is shown to affect the TOF-SIMS signal of the phospholipids in the films. The molecular environment of a LB film of dipalmitoylphosphatidylcholine (DPPC) is changed by varying the film density, varying the sample substrate, and the addition of cholesterol. An increase in film density results in a decrease in the headgroup fragment ion signal at a mass-to-charge ratio of 184 (phosphocholine). Varying the sample substrate increases the secondary ion yield of phosphocholine as does the addition of proton-donating molecules such as cholesterol to the DPPC LB film. Switching from a model system of DPPC and cholesterol to one of dipalmitoylphosphatidylethanolamine (DPPE) and cholesterol demonstrates the ability of cholesterol to also mask the phospholipid headgroup ion signal. TOF-SIMS studies of simplistic phospholipid LB model membrane systems demonstrate the potential use of these systems in TOF-SIMS analysis of cells.