Dynamic structure of lipid-bound synaptobrevin suggests a nucleation-propagation mechanism for trans-SNARE complex formation

The synaptic vesicle protein synaptobrevin engages with syntaxin and SNAP-25 to form the SNARE complex, which drives membrane fusion in neuronal exocytosis. In the SNARE complex, the SNARE motif of synaptobrevin forms a 55-residue helix, but it has been assumed to be mostly unstructured in its prefusion form. NMR data for full-length synaptobrevin in dodecylphosphocholine micelles reveals two transient helical segments flanked by natively disordered regions and a third more stable helix. Transient helix I comprises the most N-terminal part of the SNARE motif, transient helix II extends the SNARE motif into the juxtamembrane region, and the more stable helix III is the transmembrane domain. These helices may have important consequences for SNARE complex folding and fusion: helix I likely forms a nucleation site, the C-terminal disordered SNARE motif may act as a folding arrest signal, and helix II likely couples SNARE complex folding and fusion.     

Protective role of the apolipoprotein E2 allele in age-related disease traits and survival: evidence from the Long Life Family Study

The apolipoprotein E (apoE) is a classic example of a gene exhibiting pleiotropism. We examine potential pleiotropic associations of the apoE2 allele in three biodemographic cohorts of long-living individuals, offspring, and spouses from the Long Life Family Study, and intermediate mechanisms, which can link this allele with age-related phenotypes. We focused on age-related macular degeneration, bronchitis, asthma, pneumonia, stroke, creatinine, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol, diseases of heart (HD), cancer, and survival. Our analysis detected favorable associations of the ε2 allele with lower LDL-C levels, lower risks of HD, and better survival. The ε2 allele was associated with LDL-C in each gender and biodemographic cohort, including long-living individuals, offspring, and spouses, resulting in highly significant association in the entire sample (β = ?7.1, p = 6.6 × 10^?44). This allele was significantly associated with HD in long-living individuals and offspring (relative risk [RR] = 0.60, p = 3.1 × 10^?6) but this association was not mediated by LDL-C. The protective effect on survival was specific for long-living women but it was not explained by LDL-C and HD in the adjusted model (RR = 0.70, p = 2.1 × 10^?2). These results show that ε2 allele may favorably influence LDL-C, HD, and survival through three mechanisms. Two of them (HD- and survival-related) are pronounced in the long-living parents and their offspring; the survival-related mechanism is also sensitive to gender. The LDL-C-related mechanism appears to be independent of these factors. Insights into mechanisms linking ε2 allele with age-related phenotypes given biodemographic structure of the population studied may benefit translation of genetic discoveries to health care and personalized medicine.     

From the Cover: Convolutional networks for fast,energy-efficient neuromorphic computing

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware’s underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.The human brain is capable of remarkable acts of perception while consuming very little energy. The dream of brain-inspired computing is to build machines that do the same, requiring high-accuracy algorithms and efficient hardware to run those algorithms. On the algorithm front, building on classic work on backpropagation (1), the neocognitron (2), and convolutional networks (3), deep learning has made great strides in achieving human-level performance on a wide range of recognition tasks (4). On the hardware front, building on foundational work on silicon neural systems (5), neuromorphic computing, using novel architectural primitives, has recently demonstrated hardware capable of running 1 million neurons and 256 million synapses for extremely low power (just 70 mW at real-time operation) (6). Bringing these approaches together holds the promise of a new generation of embedded, real-time systems, but first requires reconciling key differences in the structure and operation between contemporary algorithms and hardware. Here, we introduce and demonstrate an approach we call Eedn, energy-efficient deep neuromorphic networks, which creates convolutional networks whose connections, neurons, and weights have been adapted to run inference tasks on neuromorphic hardware.For structure, typical convolutional networks place no constraints on filter sizes, whereas neuromorphic systems can take advantage of blockwise connectivity that limits filter sizes, thereby saving energy because weights can now be stored in local on-chip memory within dedicated neural cores. Here, we present a convolutional network structure that naturally maps to the efficient connection primitives used in contemporary neuromorphic systems. We enforce this connectivity constraint by partitioning filters into multiple groups and yet maintain network integration by interspersing layers whose filter support region is able to cover incoming features from many groups by using a small topographic size (7).For operation, contemporary convolutional networks typically use high precision ( ≥ 32-bit) neurons and synapses to provide continuous derivatives and support small incremental changes to network state, both formally required for backpropagation-based gradient learning. In comparison, neuromorphic designs can use one-bit spikes to provide event-based computation and communication (consuming energy only when necessary) and can use low-precision synapses to colocate memory with computation (keeping data movement local and avoiding off-chip memory bottlenecks). Here, we demonstrate that by introducing two constraints into the learning rule—binary-valued neurons with approximate derivatives and trinary-valued ({−1,0,1}) synapses—it is possible to adapt backpropagation to create networks directly implementable using energy efficient neuromorphic dynamics. This approach draws inspiration from the spiking neurons and low-precision synapses of the brain (8) and builds on work showing that deep learning can create networks with constrained connectivity (9), low-precision synapses (10, 11), low-precision neurons (12–14), or both low-precision synapses and neurons (15, 16). For input data, we use a first layer to transform multivalued, multichannel input into binary channels using convolution filters that are learned via backpropagation (12, 16) and whose output can be sent on chip in the form of spikes. These binary channels, intuitively akin to independent components (17) learned with supervision, provide a parallel distributed representation to carry out high-fidelity computation without the need for high-precision representation.Critically, we demonstrate that bringing the above innovations together allows us to create networks that approach state-of-the-art accuracy performing inference on eight standard datasets, running on a neuromorphic chip at between 1,200 and 2,600 frames/s (FPS), using between 25 and 275 mW. We further explore how our approach scales by simulating multichip configurations. Ease-of-use is achieved using training tools built from existing, optimized deep learning frameworks (18), with learned parameters mapped to hardware using a high-level deployment language (19). Although we choose the IBM TrueNorth chip (6) for our example deployment platform, the essence of our constructions can apply to other emerging neuromorphic approaches (20–23) and may lead to new architectures that incorporate deep learning and efficient hardware primitives from the ground up.     

Site-directed serology with synthetic peptides representing the large glycoprotein G of respiratory syncytial virus.

A set of 23 nested 15-amino-acid-long peptides with overlaps of 5 amino acids, representing the complete extramembranous part of the large glycoprotein G of respiratory syncytial (RS) virus, was analyzed in ELISA against different sera containing virus-specific antibodies. Seven of the peptides reacted with rabbit hyperimmune sera against purified virions. In contrast, only one of these seven peptides reacted with murine monoclonal antibodies specific for G. In connection with RS virus infections in humans, increase of antibody titers against three peptides was found in about one-third of the cases. These three peptides were included among those identified by both murine and rabbit antibodies. The present findings may open possibilities for site-directed clinical serology in the case of RS virus infections.     

Predictors,costs, and causes of readmission after surgery for sinonasal cancer: a national perspective

Background

Hospital readmissions are an increasingly scrutinized marker of surgical care delivery and quality. There is a paucity of information in the literature regarding the rate, risk factors, and common causes of readmission after surgery for sinonasal cancer.

Methods

We analyzed the Nationwide Readmissions Database for patients who underwent surgery for a diagnosis of sinonasal cancer between 2010 and 2014. Rates, causes, and patient‐, procedure‐, and hospital‐level risk factors for 30‐day readmission were determined. Multivariate logistic regression was used to identify predictors of 30‐day readmission.

Results

Among the 4173 cases, the 30‐day readmission rate was 11.6%, with an average cost per readmission of $18,403. The most common readmission diagnoses were wound complications (15.3%) and infections (13.4%). On multivariate regression, significant risk factors for readmission were chronic renal failure (odds ratio [OR], 2.95; 95% confidence interval [CI], 1.41‐6.17), involvement of the skull base or orbit (OR, 1.67; 95% CI, 1.11‐2.51), nonelective initial surgical admission (OR, 2.35; 95% CI, 1.42‐3.89), and length of stay ≥7 days (OR, 1.87; 95% CI, 1.14‐3.05).

Conclusion

Through the use of a large national database, we found that approximately 1 in 9 patients undergoing surgery for sinonasal cancer was readmitted within 30 days. Readmissions were most commonly associated with wound complications and infections. Factors related to procedural complexity were more important predictors of readmission than patients’ demographics or comorbidities.

     

Promyelocytic leukemia protein in mesenchymal stem cells is essential for leukemia progression

The dynamic interactions between leukemic cells and cells resident within the bone marrow microenvironment are vital for leukemia progression. The lack of detailed knowledge about the cellular and molecular mechanisms involved in this cross-talk restricts the design of effective treatments. Guarnerio et al. (2018) by using state-of-the-art techniques, including sophisticated Cre/loxP technologies in combination with leukemia mouse models, reveal that mesenchymal stem cells via promyelocytic leukemia protein (Pml) maintain leukemic cells in the bone marrow niche. Strikingly, genetic deletion of Pml in mesenchymal stem cells raised survival of leukemic mice under chemotherapeutic treatment. The emerging knowledge from this research provides a novel target in the bone marrow niche for therapeutic benefit in leukemia.     

Viral myocarditis and coagulopathy: increased tissue factor expression and plasma thrombogenicity

We investigated the effects of viral infection on Tissue Factor (TF) expression and activity in mice within the myocardium to understand increased thrombosis during myocarditis. Mice were infected with coxsackie virus B3 (CVB3) and the hearts were collected at day 4, 8 and 28 post infection (p.i.). Myocardial TF expression and cellular activity as well as plasma activity were analyzed from CVB3 infected mice by Western blot, chromogenic Factor Xa generation assay, in situ staining for active TF and immunohistochemistry. In addition to TF expression, hemodynamic parameters were measured during the time course of infection. Furthermore, we analyzed myocardial tissues from patients with suspected inflammatory cardiomyopathy. TF protein expression was maximally 5-fold elevated 8 days p.i. in mice and remained increased on day 28 p.i. (P < 0.001 vs. non-infected controls). Alterations in TF expression were associated with fibrin deposits within the myocardium. The TF pathway inhibitor protein expression in the myocardium was not altered during myocarditis. Active cellular TF co-localized with CD3 positive cells and VCAM-1 positive endothelial cells in the myocardium. The TF expression was positively correlated with the amount of infiltrating CD3 and Mac3 positive cells (Spearman-Rho ρ = 0.749 P < 0.0001 for CD3⁺ and ρ = 0.775 P < 0.0001 for Mac3⁺; N = 35). Increased myocardial TF expression was associated with a 2-fold elevated plasma activity (P < 0.05 vs. non-infected controls). In the human hearts, the TF expression correlated postively with an endothelial cell activation marker (ρ = 0.523 P < 0.0001 for CD62E; N = 54). Viral myocarditis is a hypercoagulative state which is associated with increased myocardial TF expression and activity. Upregulation of TF contributes to a systemic activation of the coagulation cascade.     

Symptomatic pulmonary embolism and the risk of recurrent venous thromboembolism

BACKGROUND: In patients with a first symptomatic pulmonary embolism (PE), the risk of recurrence is unknown. We therefore investigated the risk of recurrence among patients with spontaneous symptomatic PE and among those with deep vein thrombosis (DVT) without symptoms of PE. METHODS: After discontinuation of secondary thromboprophylaxis for a first venous thromboembolism (VTE), we prospectively observed 436 patients for an average of 30 months. Patients with secondary VTE, natural inhibitor deficiencies, lupus anticoagulant, cancer, long-term antithrombotic therapy, vena cava filters, or pregnancy were excluded. The study outcome was objectively documented recurrent symptomatic VTE. RESULTS: Recurrent VTE was seen among 28 (17.3%) of 162 patients with symptomatic PE and among 26 (9.5%) of 274 patients with DVT without symptoms of PE. Compared with patients with DVT, the relative risk of recurrent VTE among patients with symptomatic PE was 2.2 (95% confidence interval, 1.3-3.7; P =.005). The relative risk was not affected by age, sex, presence of factor V Leiden or prothrombin G20210A, hyperhomocysteinemia, or high factor VIII levels. Compared with patients with DVT without symptoms of PE, patients with symptomatic PE had an adjusted relative risk of PE at recurrence of 4.0 (95% confidence interval, 1.3-12.3; P =.03). CONCLUSION: Patients with a first symptomatic PE not only have a higher risk of recurrent VTE than those with DVT without symptoms of PE, but are also at high risk of symptomatic PE at recurrence.     

A monokine regulates colony-stimulating activity production by vascular endothelial cells

Human umbilical vein endothelial cells were cultured in supernatants of peripheral blood monocytes that had been cultured for 3 days with and without lactoferrin. Colony-stimulating activity (CSA) was measured in supernatants of the endothelial cell cultures and appropriate control cultures using normal, T-lymphocyte-depleted, phagocyte-depleted, low- density bone marrow cells in colony growth (CFU-GM) assays. Monocyte- conditioned medium contained a nondialyzable, heat labile factor that enhanced 4-15--fold the production of CSA by endothelial cells. The addition of lactoferrin to monocyte cultures reduced the activity of this monokine by 69%. Lactoferrin did not inhibit CSA production by monokine-stimulated endothelial cells. Therefore, vascular endothelial cells are potent sources of CSA, the production of CSA by these cells is regulated by a stimulatory monokine, and the production and/or release of the monokine is inhibited by lactoferrin, a neutrophil- derived putative feedback inhibitor of granulopoiesis. Inasmuch as a similar monokine is known to stimulate CSA production by fibroblasts and T lymphocytes, we suggest that mononuclear phagocytes play a pivotal role in the regulation of granulopoiesis by recruiting a variety of cell types to produce CSA.