Synthetic heparan sulfate standards and machine learning facilitate the development of solid-state nanopore analysis

The application of solid-state (SS) nanopore devices to single-molecule nucleic acid sequencing has been challenging. Thus, the early successes in applying SS nanopore devices to the more difficult class of biopolymer, glycosaminoglycans (GAGs), have been surprising, motivating us to examine the potential use of an SS nanopore to analyze synthetic heparan sulfate GAG chains of controlled composition and sequence prepared through a promising, recently developed chemoenzymatic route. A minimal representation of the nanopore data, using only signal magnitude and duration, revealed, by eye and image recognition algorithms, clear differences between the signals generated by four synthetic GAGs. By subsequent machine learning, it was possible to determine disaccharide and even monosaccharide composition of these four synthetic GAGs using as few as 500 events, corresponding to a zeptomole of sample. These data suggest that ultrasensitive GAG analysis may be possible using SS nanopore detection and well-characterized molecular training sets.

Glycosaminoglycans (GAGs) are linear anionic polysaccharides found on cell surfaces and in the extracellular matrix in all animals. GAGs comprise an important class of biopolymers that are ubiquitous in nature and exhibit a number of critical functional roles including biological recognition and signaling (1–3). Such processes play critical roles in physiology, such as in development and wound healing, and pathophysiology, such as cancer and infectious disease. Sulfated GAGs result from template-independent synthesis in the Golgi of animal cells (4, 5) and are polydisperse, heteropolysaccharides comprising variable disaccharide repeating units that are classified by these repeating units. Like nucleic acids, sulfated GAGs are made up of repeating units that comprise a linear sequence (Fig. 1). Unlike the nucleic acids, GAGs have far more complicated structures and number of possible sequences and they present severe challenges to both synthesis and characterization. Thus, we undertook to chemoenzymatically synthesize defined GAGs and characterize these using solid-state nanopore analysis.Open in a separate windowFig. 1.Structures of four synthetic GAG samples. Polysaccharide NSH is made up of N-sulfoglucosamine (GlcNS) and glucuronic acid (GlcA),NS2S is made up with GlcNS and 2-O-sulfo-iduronic acid (IdoA2S), NS6S is made up with 6-O-sulfo-N-sulfoheparosan (GlcNS6S) and GlcA, and NS6S2S is made up with GlcNS6S and IdoA2S.Despite their structural complexities, sulfated GAGs often contain well-defined domain structures that are responsible for their diverse biological functions, yet even this level of structural complexity poses a significant general challenge to structural analysis and sequencing. The simple, short-chain, chondroitin sulfate GAG component of bikunin has been sequenced using liquid chromatography–tandem mass spectrometry (LC-MS/MS) (6). While LC-MS/MS is capable of sequencing such simple, short-chain GAGs, it is not yet able to distinguish all of the many isobaric isomers of the variably sulfated saccharide residues and uronic acid epimers commonly encountered in more structurally complex GAGs, such as heparan sulfate (HS) (7). NMR has been applied to determine GAG structures but often requires milligram amounts of samples. HS/heparin is made up of →4)-β-d-glucuronic acid (GlcA) [or α-l-iduronic acid (IdoA)] (1→4)-α-d-glucosamine (GlcN) [1→ repeating units with 2-O-sulfo (S) groups on selected uronic acid residues and 3- and/or 6-O-S and N-S or N-acetyl (Ac) group substitutions on the glucosamine residues] (Fig. 1). GAG structural analysis presents challenges beyond their chemical complexity. There are no amplification methods to detect small numbers of GAG chains, whereas nucleic acid analysis can rely on PCR. Similarly, there are few GAG-specific antibodies or aptamers (8), and no natural GAG chromophores or fluorophores (9), in contrast to the many used for protein sensing. Ultrasensitive (zeptomole) detection methods of modified GAGs, based on fluorescence resonance energy transfer (FRET) (10), DNA bar coding (11), and dye-based nanosensors (12) have been demonstrated, but their application to sequencing is particularly challenging because of the high level of structural complexity of sulfated GAGs.Nanopore single-molecule detection is now routinely applied to DNA (13, 14) and RNA (15–17) biopolymers, and is increasingly applied to protein characterization (18–22). In brief, a nanopore is a nanofluidic channel ∼10 nm long and <100 nm in diameter, serving as the sole fluid connection between two reservoirs of electrolyte separated by an otherwise impermeable membrane (Fig. 2A). On applying a voltage across this nanopore, the passage of supporting electrolyte ions results in a “baseline,” or open-pore current, i₀. The passage of a biopolymer analyte through this nanopore disrupts the flow of supporting electrolyte ions, often as a current blockage. This temporary reduction in ionic current is called an “event,” and its magnitude (mean blockage ratio over the dwell time, ⟨f_b⟩=⟨i⟩_Td/⟨i₀⟩) and its temporal features [dwell time (Td)] (Fig. 2 B and C) depend on the size and shape of the nanopore, the biopolymer analyte, and the applied voltage and interfacial charge distributions. Indeed, the passage of DNA through engineered protein nanopore devices produces current blockages that can be applied in sequencing, and the widespread use of these commercial protein nanopore DNA sequencing devices is increasing (23, 24). Despite this success with protein nanopores, the potential benefits of (abiotic) solid-state (SS) nanopores have continued to drive development efforts. Such a transition to the freely size-tunable SS platform (25, 26), however, is vital for the application of nanopores to the characterization of branched glycans (27). Yet the use of SS nanopores in even the better-established DNA sensing regime remains challenging. The application of nanopore sensing to glycans, while promising, remains profoundly exploratory using nanopores of any kind. The transition to the SS nanopores is accompanied by significant changes in pore geometry, chemistry, characteristics, and potential analyte–pore interactions and sensing modalities, so that there is a critical need for studies in the realm of nanopore glycomics (27, 28). For example, outcomes of early nanopore studies on a structurally simple unsulfated GAG, hyaluronan (HA, →4)- β -GlcA (1 → 3)- β -GlcNAc (1→), while providing some information on HA size does not provide definitive structural information (29, 30). SS nanopore analysis of two sulfated GAGs, heparin and a heparin contaminant, oversulfated chondroitin sulfate, using a silicon nitride SS nanopore was able to qualitatively identify these GAGs by either the magnitude or duration of characteristic current blockages (28). SS nanopore data on GAGs, analyzed using a machine-learning (ML) algorithm (i.e., a support vector machine [SVM]), distinguished heparin and chondroitin sulfate oligosaccharides and unfractionated heparin and low molecular weight heparin with >90% accuracy (31).Open in a separate windowFig. 2.Nanopore characteristics of four samples. (A) Schematic of the nanopore configuration. Anionic GAGs driven by electrophoresis to and through the pore with a negative applied voltage would be detected if they perturbed the open-pore current. (B) A representative current trace and events from polysaccharide NS6S2S test using an ∼6-nm-diameter nanopore. Measurements were collected using a −150-mV applied-voltage (details in Results and Discussion, and Materials and Methods) (C) Scatter plots of dwell time vs. current blockage ratio for four polysaccharides. To remove the bias of event numbers in human image recognition, all plots contain only the first 2,475 events. (D) PCA visualization of the embedded images from the four unique GAGs. The blue circles and region represent NSH, the red X and region represents NS2S, the green triangle and region represents NS6S, and the brown cross and region represents NS6S2S. The algorithm clusters signals from each GAG based on scatter plot images. Each insert shows one 500-events image from each sample class. All 500-events images are in SI Appendix, Fig. S12.Nanopore studies on GAGs, and glycans more broadly, have been severely limited by the lack of a library of structurally defined standards. The uniformity of sulfated GAGs prepared from animal sources is difficult to control and exhibits significant sequence heterogeneity and polydispersity (32). HS is particularly problematic as even for a small HS hexasaccharide, composed of an IdoA/GlcA:GlcNS/GlcNAc sequence with 12 available sites for random sulfation, there are 32,768 possible sequences. Recently, chemoenzymatic synthesis has made inroads in the preparation of high-purity sulfated HS GAGs from heparosan (→4)- β -GlcA (1→4)- β -GlcNAc (1→) (33). HS GAGs of approximately the same chain length and polydispersity and having a single repeating disaccharide unit (SI Appendix, Table S1) including, NSH (→4)- β -GlcA (1→4)- β -GlcNS (1→), NS2S (→4)-α α -IdoA2S (1 → 4)- β -GlcNS (1→), NS6S (→4)- β -GlcA (1→4)- β -GlcNS6S(1→)), NS6S2S (→4)- α -IdoA2S (1→4)- β -GlcNS6S (1→) have been prepared (see Materials and Methods and ref. 34) (Fig. 1). Here we use our recently developed synthetic technique, which has proven difficult to benchmark, in conjunction with a nanopore technique, which has only just begun to be applied to glycomics and has been severely challenged by the lack of available high-quality samples, to develop a fully integrated approach for the nanopore analysis of complex carbohydrates.     

Proapoptotic N-truncated BCL-xL protein activates endogenous mitochondrial channels in living synaptic terminals

Neuronal death is often preceded by functional alterations at nerve terminals. Anti- and proapoptotic BCL-2 family proteins not only regulate the neuronal death pathway but also affect excitability of healthy neurons. We found that exposure of squid stellate ganglia to hypoxia, a death stimulus for neurons, causes a cysteine protease-dependent loss of full-length antiapoptotic BCL-xL, similar to previous findings in mammalian cells. Therefore, to determine the direct effect of the naturally occurring proapoptotic cleavage product of BCL-xL on mitochondria, recombinant N-truncated BCL-xL was applied to mitochondria inside the squid presynaptic terminal and to purified mitochondria isolated from yeast. N-truncated BCL-xL rapidly induced large multi-conductance channels with a maximal conductance significantly larger than those produced by full-length BCL-xL. This activity required the hydrophobic C terminus and the BH3 domain of BCL-xL. Moreover, N-truncated BCL-xL failed to produce any channel activity when applied to plasma membranes, suggesting that a component of the mitochondrial membrane is necessary for its actions. Consistent with this idea, the large channels induced by N-truncated BCL-xL are inhibited by NADH and require the presence of VDAC, a voltage-dependent anion channel present in the outer mitochondrial membrane. These observations suggest that the mitochondrial channels specific to full-length and N-truncated BCL-xL contribute to their opposite effects on synaptic transmission, and are consistent with their opposite effects on the cell death pathway.     

Self-reported comorbidities,perceived needs,and sources for usual care for older and younger homeless adults

BACKGROUND: While older individuals who are homeless tend to be in poorer health, it is less clear how they view their health care needs and whether their self-reported patterns for accessing health services differ from younger homeless counterparts. METHODS: Cross-sectional, community-based survey of homeless adults in Pittsburgh and Philadelphia using face-to-face interviews from population proportionate sampling of sites and random sampling of subjects. Survey questions included physical and mental health comorbidities, self-reported health care, social services and personal needs, means of economic support, and sources for usual health care. For analysis purposes, respondents were grouped by age 18 to 49 years old and 50 years old or older. RESULTS: Overall, 531 adults were interviewed, with 74 respondents 50 years old or older (13.9%). Older homeless persons were 3.6 times more likely to report a chronic medical condition, 2.8 times more likely to have health insurance, and 2.4 times more likely to be dependent on heroin than homeless persons less than 50 years old. However, they also tended to use shelter-based clinics and street outreach teams more commonly as their source of usual care (20.9% vs 10.6%, P=.02) and were significantly less likely to report a need for substance abuse treatment despite high rates of abuse. CONCLUSION: Older homeless adults have a greater disease burden than their younger counterparts. However, it is unclear whether these needs are being appropriately identified and met. There is a need for specific and targeted outreach to connect them to appropriate services.     

HIV type 1 CRF13_cpx revisited: identification of a new sequence from Cameroon and signal for subsubtype J2

A nearly full-length genome sequence of an HIV-1 isolate originating from Cameroon, 02CM.3226MN, was found to cluster together with previously reported CRF13 sequences 96CM-4164 and 96CM-1849. Similarity plotting, bootscanning, breakpoint analysis, and phylogenetic trees confirmed similar genomic structures with almost identical breakpoint positions among these three isolates. Thus, CRF13 now fulfills the HIV-1 nomenclature requirements. A X2 analysis across all three genomes simultaneously was applied to more accurately determine breakpoints and address the uncertainty in such estimates. Some fragments were found to be difficult to classify, as indicated by a low branching index (BI), due to limited knowledge about parental and reference subtype sequences. One fragment with low BI association to reference subtype J sequences (BI = 0.27, cut-off for subtype classification >0.55) was found to be closer to J fragments of CRF11 similar to the way that A1-A2 and F1-F2 subsubtypes associate. This suggests that subtype J may need to be reclassified into subsubtypes J1 and J2. The CRF13 genome consists of fragments from subtypes A1, G, and both J1 and J2 as well as CRF01 and one region that was left unclassified.     

Semaphorin 6D regulates the late phase of CD4+ T cell primary immune responses

The semaphorin and plexin family of ligand and receptor proteins provides important axon guidance cues required for development. Recent studies have expanded the role of semaphorins and plexins in the regulation of cardiac, circulatory and immune system function. Within the immune system, semaphorins and plexins regulate cell–cell interactions through a complex network of receptor and ligand pairs. Immune cells at different stages of development often express multiple semaphorins and plexins, leading to multivariate interactions, involving more than one ligand and receptor within each functional group. Because of this complexity, the significance of semaphorin and plexin regulation on individual immune cell types has yet to be fully appreciated. In this work, we examined the regulation of T cells by semaphorin 6D. Both in vitro and in vivo T cell stimulation enhanced semaphorin 6D expression. However, semaphorin 6D was only expressed by a majority of T cells during the late phases of activation. Consequently, the targeted disruption of semaphorin 6D receptor–ligand interactions inhibited T cell proliferation at late but not early phases of activation. This proliferation defect was associated with reduced linker of activated T cells protein phosphorylation, which may reflect semaphorin 6D regulation of c-Abl kinase activity. Semaphorin 6D disruption also inhibited expression of CD127, which is required during the multiphase antigen-presenting cell and T cell interactions leading to selection of long-lived lymphocytes. This work reveals a role for semaphorin 6D as a regulator of the late phase of primary immune responses.     

Use of experimental crystallographic phases to examine the hydration of polar and nonpolar cavities in T4 lysozyme

There is conflicting evidence as to whether cavities in proteins that are nonpolar and large enough to accommodate solvent are empty or are occupied by disordered water molecules. Here, we use multiple-wavelength x-ray data collected from crystals of the selenomethionine-substituted L99A/M102L mutant of T4 lysozyme to obtain a high-resolution electron density map free of bias that is unavoidably associated with conventional model-based structure determination and refinement. The mutant, L99A/M102L, has four cavities, two being polar in character and the other two nonpolar. Cavity 1 (polar, volume 45.2 ų) was expected to contain two well ordered water molecules, and this is confirmed in the experimental electron density map. Likewise, cavity 2 (polar, 16.9 ų) is confirmed to contain a single water molecule. Cavity 3 (nonpolar, 21.4 ų) was seen to be empty in conventional x-ray refinement, and this is confirmed in the experimental map. Unexpectedly, however, cavity 4 (nonpolar, volume 133.5 ų) was seen to contain diffuse electron density equivalent to ≈1.5 water molecules. Although cavity 4 is largely nonpolar, it does have some polar character, and this apparently contributes to the presence of solvent. The cavity is large enough to accommodate four to five water molecules, and it appears that a hydrogen-bonded chain of three or more solvent molecules could occupy the cavity at a given time. The results are consistent with theoretical predictions that cavities in proteins that are strictly nonpolar will not contain solvent until the volume is large enough to permit mutually satisfying water–water hydrogen bonds.