Deep Learning Insights into Lanthanides Complexation Chemistry.

Modern structure-property models are widely used in chemistry; however, in many cases, they are still a kind of a "black box" where there is no clear path from molecule structure to target property. Here we present an example of deep learning usage not only to build a model but also to determine key structural fragments of ligands influencing metal complexation. We have a series of chemically similar lanthanide ions, and we have collected data on complexes' stability, built models, predicting stability constants and decoded the models to obtain key fragments responsible for complexation efficiency. The results are in good correlation with the experimental ones, as well as modern theories of complexation. It was shown that the main influence on the constants had a mutual location of the binding centers.

Investigation of the effect of dietary intake of omega-3 polyunsaturated fatty acids on trauma-induced white matter injury with quantitative diffusion MRI in mice.

Previous studies suggest that long-term supplementation and dietary intake of omega-3 polyunsaturated fatty acids (PUFAs) may have neuroprotective effects following brain injury. The objective of this study was to investigate potential neuroprotective effects of omega-3 PUFAs on white matter following closed-head trauma. The closed-head injury model of engineered rotational acceleration (CHIMERA) produces a reproducible injury in the optic tract and brachium of the superior colliculus in mice. Damage is detectable using diffusion tensor imaging (DTI) metrics, particularly fractional anisotropy (FA), with sensitivity comparable to histology. We acquired in vivo (n = 38) and ex vivo (n = 41) DTI data in mice divided into sham and CHIMERA groups with two dietary groups: one deficient in omega-3 PUFAs and one adequate in omega-3 PUFAs. We examined injury effects (reduction in FA) and neuroprotection (FA reduction modulated by diet) in the optic tract and brachium. We verified that diet did not affect FA in sham animals. In injured animals, we found significantly reduced FA in the optic tract and brachium (~10% reduction, p < 0.001), and Bayes factor analysis showed strong evidence to reject the null hypothesis. However, Bayes factor analysis showed substantial evidence to accept the null hypothesis of no diet-related FA differences in injured animals in the in vivo and ex vivo samples. Our results indicate no neuroprotective effect from adequate dietary omega-3 PUFA intake on white matter damage following traumatic brain injury. Since damage from CHIMERA mainly affects white matter, our results do not necessarily contradict previous findings showing omega-3 PUFA-mediated neuroprotection in gray matter.

Graph Convolutional Neural Networks as "General-Purpose" Property Predictors: The Universality and Limits of Applicability.

Nowadays the development of new functional materials/chemical compounds using machine learning (ML) techniques is a hot topic and includes several crucial steps, one of which is the choice of chemical structure representation. The classical approach of rigorous feature engineering in ML typically improves the performance of the predictive model, but at the same time, it narrows down the scope of applicability and decreases the physical interpretability of predicted results. In this study, we present graph convolutional neural networks (GCNNs) as an architecture that allows for successfully predicting the properties of compounds from diverse domains of chemical space, using a minimal set of meaningful descriptors. The applicability of GCNN models has been demonstrated by a wide range of chemical domain-specific properties. Their performance is comparable to state-of-the-art techniques; however, this architecture exempts from the need to carry out precise feature engineering.

Comparing and Validating Machine Learning Models for Mycobacterium tuberculosis Drug Discovery.

Tuberculosis is a global health dilemma. In 2016, the WHO reported 10.4 million incidences and 1.7 million deaths. The need to develop new treatments for those infected with Mycobacterium tuberculosis ( Mtb) has led to many large-scale phenotypic screens and many thousands of new active compounds identified in vitro. However, with limited funding, efforts to discover new active molecules against Mtb needs to be more efficient. Several computational machine learning approaches have been shown to have good enrichment and hit rates. We have curated small molecule Mtb data and developed new models with a total of 18,886 molecules with activity cutoffs of 10 µM, 1 µM, and 100 nM. These data sets were used to evaluate different machine learning methods (including deep learning) and metrics and to generate predictions for additional molecules published in 2017. One Mtb model, a combined in vitro and in vivo data Bayesian model at a 100 nM activity yielded the following metrics for 5-fold cross validation: accuracy = 0.88, precision = 0.22, recall = 0.91, specificity = 0.88, kappa = 0.31, and MCC = 0.41. We have also curated an evaluation set ( n = 153 compounds) published in 2017, and when used to test our model, it showed the comparable statistics (accuracy = 0.83, precision = 0.27, recall = 1.00, specificity = 0.81, kappa = 0.36, and MCC = 0.47). We have also compared these models with additional machine learning algorithms showing Bayesian machine learning models constructed with literature Mtb data generated by different laboratories generally were equivalent to or outperformed deep neural networks with external test sets. Finally, we have also compared our training and test sets to show they were suitably diverse and different in order to represent useful evaluation sets. Such Mtb machine learning models could help prioritize compounds for testing in vitro and in vivo.

Comparison of Deep Learning With Multiple Machine Learning Methods and Metrics Using Diverse Drug Discovery Data Sets.

Machine learning methods have been applied to many data sets in pharmaceutical research for several decades. The relative ease and availability of fingerprint type molecular descriptors paired with Bayesian methods resulted in the widespread use of this approach for a diverse array of end points relevant to drug discovery. Deep learning is the latest machine learning algorithm attracting attention for many of pharmaceutical applications from docking to virtual screening. Deep learning is based on an artificial neural network with multiple hidden layers and has found considerable traction for many artificial intelligence applications. We have previously suggested the need for a comparison of different machine learning methods with deep learning across an array of varying data sets that is applicable to pharmaceutical research. End points relevant to pharmaceutical research include absorption, distribution, metabolism, excretion, and toxicity (ADME/Tox) properties, as well as activity against pathogens and drug discovery data sets. In this study, we have used data sets for solubility, probe-likeness, hERG, KCNQ1, bubonic plague, Chagas, tuberculosis, and malaria to compare different machine learning methods using FCFP6 fingerprints. These data sets represent whole cell screens, individual proteins, physicochemical properties as well as a data set with a complex end point. Our aim was to assess whether deep learning offered any improvement in testing when assessed using an array of metrics including AUC, F1 score, Cohen's kappa, Matthews correlation coefficient and others. Based on ranked normalized scores for the metrics or data sets Deep Neural Networks (DNN) ranked higher than SVM, which in turn was ranked higher than all the other machine learning methods. Visualizing these properties for training and test sets using radar type plots indicates when models are inferior or perhaps over trained. These results also suggest the need for assessing deep learning further using multiple metrics with much larger scale comparisons, prospective testing as well as assessment of different fingerprints and DNN architectures beyond those used.

Preliminary Promising Findings for Manganese Chloride as a Novel Radiation Countermeasure Against Acute Radiation Syndrome.

INTRODUCTION: Military members and first responders may, at moment's notice, be asked to assist in incidents that may result in radiation exposure such as Operation Tomadachi in which the U.S. Navy provided significant relief for the Fukushima Daiichi Nuclear Reactor accident in Japan after an earthquake and tsunami in 2011. We are also currently facing potential threats from nuclear power plants in the Ukraine should a power disruption to a nuclear plant interfere with cooling or other safety measures. Exposure to high doses of radiation results in acute radiation syndrome (ARS) characterized by symptoms arising from hematopoietic, gastrointestinal, and neurovascular injuries. Although there are mitigators FDA approved to treat ARS, there are currently no FDA-approved prophylactic medical interventions to help protect persons who may need to respond to radiation emergencies. There is strong evidence that manganese (Mn) has radiation protective efficacy as a promising prophylactic countermeasure. MATERIALS AND METHODS: All animal procedures were approved by the Institutional Animal Care and Use Committee. Male and female B6D2F1J mice, 10 to 11 weeks old, were used for neurotoxicity studies and temporal effects of Mn. Four groups were evaluated: (1) vehicle injection, (2) dose of 4.5 mg/kg for 3 days, (3) dose of 13.5 mg/kg, and (4) sham. Irradiated mice were exposed to 9.5 Gy whole body Co60 γ-radiation. MRI was performed with a high dose of manganese chloride (MnCl2) (150 mg/kg) to assess the distribution of the MnCl2. RESULTS: The mice have promising survival curves (highest survival-13.5 mg/kg dose over 3 days of MnCl2 at 80% [87% female, 73% male] P = 0.0004). The complete blood count (CBC) results demonstrated a typical hematopoietic response in all of the irradiated groups, followed by mildly accelerated recovery by day 28 in the treated groups. No difference between groups was measured by Rota Rod, DigiGait, and Y-maze. Histologic evaluation of the bone marrow sections in the group given 13.5 mg/kg dose over 3 days had the best return to cellularity at 80%. MRI showed a systemic distribution of MnCl2. DISCUSSION: The preliminary data suggest that a dose of 13.5 mg/kg of MnCl2 given over 3 days prior to exposure of radiation may have a protective benefit while not exhibiting the neurobehavioral problems. A countermeasure that can prophylactically protect emergency personnel entering an area contaminated with high levels of radiation is needed, especially in light that nuclear accidents are a continued global threat. There is a need for a protective agent with easy long-term storage, easy to transport, easy to administer, and low cost. Histologic evaluation supports the promising effect of MnCl2 in protecting tissue, especially the bone marrow using the dose given over 3 days (4.5 mg/kg per day) of MnCl2. CONCLUSIONS: Initial experiments show that MnCl2 is a promising safe and effective prophylactic countermeasure against ARS. MRI data support the systemic distribution of MnCl2 which is needed in order to protect multiple tissues in the body. The pathology data in bone marrow and the brain support faster recovery from radiation exposure in the treated animals and decreased organ damage.

Natural-history Characterization of a Murine Partial-body Irradiation Model System: Establishment of a Multiple-Parameter Based GI-ARS Severity-Scoring System.

The purpose of this investigation was to characterize the natural history of a murine total-abdominal-irradiation exposure model to measure gastrointestinal acute radiation injury. Male CD2F1 mice at 12 to 15 weeks old received total-abdominal irradiation using 4-MV linear accelerator X-rays doses of 0, 11, 13.5, 15, 15.75 and 16.5 Gy (2.75 Gy/min). Daily cage-side (i.e., in the animal housing room) observations of clinical signs and symptoms including body weights on all animals were measured up to 10 days after exposure. Jejunum tissues from cohorts of mice were collected at 1, 3, 7 and 10 days after exposure and radiation injury was assessed by histopathological analyses. Results showed time- and dose-dependent loss of body weight [for example at 7 days: 0.66 (±0.80) % loss for 0 Gy, 6.40 (±0.76) % loss at 11 Gy, 9.43 (±2.06) % loss at 13.5 Gy, 23.53 (± 1.91) % loss at 15 Gy, 29.97 (±1.16) % loss at 15.75 Gy, and 31.79 (±0.76) % loss at 16.5 Gy]. Negligible clinical signs and symptoms, except body weight changes, of radiation injury were observed up to 10 days after irradiation with doses of 11 to 15 Gy. Progressive increases in the severity of clinical signs and symptoms were found after irradiation with doses >15 Gy. Jejunum histology showed a progressive dose-dependent increase in injury. For example, at 7 days postirradiation, the percent of crypts, compared to controls, decreased to 82.3 (±9.5), 69.2 (±12.3), 45.4 (±11.9), 18.0 (±3.4), and 11.5 (± 1.8) with increases in doses from 11 to 16.5 Gy. A mucosal injury scoring system was used that mainly focused on changes in villus morphology damage (i.e., subepithelial spaces near the tips of the villi with capillary congestion, significant epithelial lifting along the length of the villi with a few denuded villus tips). Peak levels of total-abdominal irradiation induced effects on the mucosal injury score were seen 7 days after irradiation for doses ≥15 Gy, with a trend to show a decline after 7 days. A murine multiple-parameter gastrointestinal acute-radiation syndrome severity-scoring system was established based on clinical signs and symptoms that included measures of appearance (i.e., hunched and/or fluffed fur), respiratory rate, general (i.e., decreased mobility) and provoked behavior (i.e., subdued response to stimulation), weight loss, and feces/diarrhea score combined with jejunum mucosal-injury grade score. In summary, the natural-history radio-response for murine partial-body irradiation exposures is important for establishing a well-characterized radiation model system; here we established a multiple-parameter gastrointestinal acute-radiation syndrome severity-scoring system that provides a radiation injury gastrointestinal tissue-based assessment utility.

Brief isoflurane administration as an adjunct treatment to control organophosphate-induced convulsions and neuropathology.

Organophosphate-based chemical agents (OP), including nerve agents and certain pesticides such as paraoxon, are potent acetylcholinesterase inhibitors that cause severe convulsions and seizures, leading to permanent central nervous system (CNS) damage if not treated promptly. The current treatment regimen for OP poisoning is intramuscular injection of atropine sulfate with an oxime such as pralidoxime (2-PAM) to mitigate cholinergic over-activation of the somatic musculature and autonomic nervous system. This treatment does not provide protection against CNS cholinergic overactivation and therefore convulsions require additional medication. Benzodiazepines are the currently accepted treatment for OP-induced convulsions, but the convulsions become refractory to these GABAA agonists and repeated dosing has diminishing effectiveness. As such, adjunct anticonvulsant treatments are needed to provide improved protection against recurrent and prolonged convulsions and the associated excitotoxic CNS damage that results from them. Previously we have shown that brief, 4-min administration of 3%-5% isoflurane in 100% oxygen has profound anticonvulsant and CNS protective effects when administered 30 min after a lethal dose of paraoxon. In this report we provide an extended time course of the effectiveness of 5% isoflurane delivered for 5 min, ranging from 60 to 180 min after a lethal dose of paraoxon in rats. We observed substantial effectiveness in preventing neuronal loss as shown by Fluoro-Jade B staining when isoflurane was administered 1 h after paraoxon, with diminishing effectiveness at 90, 120 and 180 min. In vivo magnetic resonance imaging (MRI) derived T2 and mean diffusivity (MD) values showed that 5-min isoflurane administration at a concentration of 5% prevents brain edema and tissue damage when administered 1 h after a lethal dose of paraoxon. We also observed reduced astrogliosis as shown by GFAP immunohistochemistry. Studies with continuous EEG monitoring are ongoing to demonstrate effectiveness in animal models of soman poisoning.

Gadolinium metallofullerenol nanoparticles inhibit cancer metastasis through matrix metalloproteinase inhibition: imprisoning instead of poisoning cancer cells.

The purpose of this work is to study the antimetastasis activity of gadolinium metallofullerenol nanoparticles (f-NPs) in malignant and invasive human breast cancer models. We demonstrated that f-NPs inhibited the production of matrix metalloproteinase (MMP) enzymes and further interfered with the invasiveness of cancer cells in tissue culture condition. In the tissue invasion animal model, the invasive primary tumor treated with f-NPs showed significantly less metastasis to the ectopic site along with the decreased MMP expression. In the same animal model, we observed the formation of a fibrous cage that may serve as a physical barrier capable of cancer tissue encapsulation that cuts the communication between cancer- and tumor-associated macrophages, which produce MMP enzymes. In another animal model, the blood transfer model, f-NPs potently suppressed the establishment of tumor foci in lung. Based on these data, we conclude that f-NPs have antimetastasis effects and speculate that utilization of f-NPs may provide a new strategy for the treatment of tumor metastasis. FROM THE CLINICAL EDITOR: In this study utilizing metallofullerenol nanoparticles, the authors demonstrate antimetastasis effects and speculate that utilization of these nanoparticles may provide a new strategy in metastatic tumor therapy.

Meningeal and Visual Pathway Magnetic Resonance Imaging Analysis after Single and Repetitive Closed-Head Impact Model of Engineered Rotational Acceleration (CHIMERA)-Induced Disruption in Male and Female Mice.

The consequences of forceful rotational acceleration on the central nervous system are not fully understood. While traumatic brain injury (TBI) research primarily has focused on effects related to the brain parenchyma, reports of traumatic meningeal enhancement in TBI patients may possess clinical significance. The objective of this study was to evaluate the meninges and brain for changes in dynamic contrast enhancement (DCE) magnetic resonance imaging (MRI) following closed-head impact model of engineered rotational acceleration (CHIMERA)-induced cerebral insult. Adult male and female mice received one (1 × ; n = 19 CHIMERA, n = 19 Sham) or four (4 × one/day; n = 18 CHIMERA, n = 12 Sham) injuries. Each animal underwent three MRI scans: 1 week before injury, immediately after the final injury, and 1 week post-injury. Compared with baseline readings and measures in sham animals, meningeal DCE in males was increased after single impact and repetitive injury. In female mice, DCE was elevated relative to their baseline level after a single impact. One week after CHIMERA, the meningeal enhancement returned to below baseline for single injured male mice, but compared with uninjured mice remained elevated in both sexes in the multiple impact groups. Pre-DCE meningeal T2-weighted relaxation time was increased only after 1 × CHIMERA in injured mice. Since vision is impaired after CHIMERA, visual pathway regions were analyzed through imaging and glial fibrillary acidic protein (GFAP) histology. Initial DCE in the lateral geniculate nucleus (LGN) and superior colliculus (SC) and T2 increases in the optic tract (OPT) and LGN were observed after injury with decreases in DCE and T2 1 week later. Astrogliosis was apparent in the OPT and SC with increased GFAP staining 7 days post-injury. To our knowledge, this is the first study to examine meningeal integrity after CHIMERA in both male and female rodents. DCE-MRI may serve as a useful approach for pre-clinical models of meningeal injury that will enable further evaluation of the underlying mechanisms.

Synthesis of amphiphilic triblock copolymers as multidentate ligands for biocompatible coating of quantum dots.

One barrier to apply current tri-octylphosphine oxide (TOPO) based quantum dots (QDs) to biomedical imaging is that the TOPO on TOPO-QDs can be replaced by the proteins in living system, which may cause the degradation of QDs and/or deactivation of protein. In order to develop biocompatible optical imaging agents, a novel triblock copolymer, designed as a multidentate ligand, was synthesized to coat quantum dot nanocrystals (QDs). The copolymer consists of a polycarboxylic acid block at one end and a polythiol block at the other end with an intervening cross-linked poly(styrene-co-divinylbenzene) block bridging the ends. The multiple mercapto groups from the polythiol block act as multidentate ligands to stabilize QDs, while the polycarboxylic acid block improves the water solubility of QDs and offers reaction sites for surface modification or conjugation with bimolecules. The cross-linked poly(styrene-co-divinylbenzene) block provides a densely compacted hydrophobic shell. This shell will act as a barrier to inhibit the degradation of QDs by preventing the diffusion of ions and small molecules into the core of QDs. This new multidentate polymer coating facilitates the transfer of QDs from organic solvent into aqueous phase. The QDs directly bound to multidentate mercapto groups instead of TOPO are less likely to be affected by the mercapto or disulfide groups within proteins or other biomolecules. Therefore, this research will provide an alternative coating material instead of TOPO to produce QDs which could be more suitable for in vivo use under complex physiological conditions.

Translationally Relevant Magnetic Resonance Imaging Markers in a Ferret Model of Closed Head Injury.

Pre-clinical models of traumatic brain injury (TBI) have been the primary experimental tool for understanding the potential mechanisms and cellular alterations that follow brain injury, but the human relevance and translational value of these models are often called into question. Efforts to better recapitulate injury biomechanics and the use of non-rodent species with neuroanatomical similarities to humans may address these concerns and promise to advance experimental studies toward clinical impact. In addition to improving translational aspects of animal models, it is also advantageous to establish pre-clinical outcomes that can be directly compared with the same outcomes in humans. Non-invasive imaging and particularly MRI is promising for this purpose given that MRI is a primary tool for clinical diagnosis and at the same time increasingly available at the pre-clinical level. The objective of this study was to identify which commonly used radiologic markers of TBI outcomes can be found also in a translationally relevant pre-clinical model of TBI. The ferret was selected as a human relevant species for this study with folded cortical geometry and relatively high white matter content and the closed head injury model of engineered rotation and acceleration (CHIMERA) TBI model was selected for biomechanical similarities to human injury. A comprehensive battery of MRI protocols based on common data elements (CDEs) for human TBI was collected longitudinally for the identification of MRI markers and voxelwise analysis of T2, contrast enhancement and diffusion tensor MRI values. The most prominent MRI findings were consistent with focal hemorrhage and edema in the brain stem region following high severity injury as well as vascular and meningeal injury evident by contrast enhancement. While conventional MRI outcomes were not highly conspicuous in less severe cases, quantitative voxelwise analysis indicated diffusivity and anisotropy alterations in the acute and chronic periods after TBI. The main conclusions of this study support the translational relevance of closed head TBI models in intermediate species and identify brain stem and meningeal vulnerability. Additionally, the MRI findings highlight a subset of CDEs with promise to bridge pre-clinical studies with human TBI outcomes.

Genetic inactivation of SARM1 axon degeneration pathway improves outcome trajectory after experimental traumatic brain injury based on pathological, radiological, and functional measures.

Traumatic brain injury (TBI) causes chronic symptoms and increased risk of neurodegeneration. Axons in white matter tracts, such as the corpus callosum (CC), are critical components of neural circuits and particularly vulnerable to TBI. Treatments are needed to protect axons from traumatic injury and mitigate post-traumatic neurodegeneration. SARM1 protein is a central driver of axon degeneration through a conserved molecular pathway. Sarm1-/- mice with knockout (KO) of the Sarm1 gene enable genetic proof-of-concept testing of the SARM1 pathway as a therapeutic target. We evaluated Sarm1 deletion effects after TBI using a concussive model that causes traumatic axonal injury and progresses to CC atrophy at 10 weeks, indicating post-traumatic neurodegeneration. Sarm1 wild-type (WT) mice developed significant CC atrophy that was reduced in Sarm1 KO mice. Ultrastructural classification of pathology of individual axons, using electron microscopy, demonstrated that Sarm1 KO preserved more intact axons and reduced damaged or demyelinated axons. Longitudinal MRI studies in live mice identified significantly reduced CC volume after TBI in Sarm1 WT mice that was attenuated in Sarm1 KO mice. MR diffusion tensor imaging detected reduced fractional anisotropy in both genotypes while axial diffusivity remained higher in Sarm1 KO mice. Immunohistochemistry revealed significant attenuation of CC atrophy, myelin loss, and neuroinflammation in Sarm1 KO mice after TBI. Functionally, Sarm1 KO mice exhibited beneficial effects in motor learning and sleep behavior. Based on these findings, Sarm1 inactivation can protect axons and white matter tracts to improve translational outcomes associated with CC atrophy and post-traumatic neurodegeneration.

CATMoS: Collaborative Acute Toxicity Modeling Suite.

BACKGROUND: Humans are exposed to tens of thousands of chemical substances that need to be assessed for their potential toxicity. Acute systemic toxicity testing serves as the basis for regulatory hazard classification, labeling, and risk management. However, it is cost- and time-prohibitive to evaluate all new and existing chemicals using traditional rodent acute toxicity tests. In silico models built using existing data facilitate rapid acute toxicity predictions without using animals. OBJECTIVES: The U.S. Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM) Acute Toxicity Workgroup organized an international collaboration to develop in silico models for predicting acute oral toxicity based on five different end points: Lethal Dose 50 (LD50 value, U.S. Environmental Protection Agency hazard (four) categories, Globally Harmonized System for Classification and Labeling hazard (five) categories, very toxic chemicals [LD50 (LD50≤50mg/kg)], and nontoxic chemicals (LD50>2,000mg/kg). METHODS: An acute oral toxicity data inventory for 11,992 chemicals was compiled, split into training and evaluation sets, and made available to 35 participating international research groups that submitted a total of 139 predictive models. Predictions that fell within the applicability domains of the submitted models were evaluated using external validation sets. These were then combined into consensus models to leverage strengths of individual approaches. RESULTS: The resulting consensus predictions, which leverage the collective strengths of each individual model, form the Collaborative Acute Toxicity Modeling Suite (CATMoS). CATMoS demonstrated high performance in terms of accuracy and robustness when compared with in vivo results. DISCUSSION: CATMoS is being evaluated by regulatory agencies for its utility and applicability as a potential replacement for in vivo rat acute oral toxicity studies. CATMoS predictions for more than 800,000 chemicals have been made available via the National Toxicology Program's Integrated Chemical Environment tools and data sets (ice.ntp.niehs.nih.gov). The models are also implemented in a free, standalone, open-source tool, OPERA, which allows predictions of new and untested chemicals to be made. https://doi.org/10.1289/EHP8495.

A nanocomplex system as targeted contrast agent delivery vehicle for magnetic resonance imaging dynamic contrast enhancement study.

We have developed and tested a liposomal nanocomplex system, which contains Gd-DTPA as a payload and transferrin on the surface, as a tumor specific targeting MRI contrast agent for studying prostate cancer tumors in mice. In vivo, the probe significantly enhanced the MRI signal. The image contrast between the peripheral region of the tumor and the non-involved muscle was nearly 50% higher two hours after administration of the nanocomplex. The liposomal nanocomplex increased the amount of Gd accumulated in tumors by factor 2.8 compared to that accumulated by using Magnevist alone. Moreover, the heterogeneous MRI image features correlate well with the tumor pathology. The image enhancement patterns can be used for cancer prognosis and non-invasive monitoring of the response to therapy.

Transplantation of induced neural stem cells (iNSCs) into chronically demyelinated corpus callosum ameliorates motor deficits.

Multiple Sclerosis (MS) causes neurologic disability due to inflammation, demyelination, and neurodegeneration. Immunosuppressive treatments can modify the disease course but do not effectively promote remyelination or prevent long term neurodegeneration. As a novel approach to mitigate chronic stage pathology, we tested transplantation of mouse induced neural stem cells (iNSCs) into the chronically demyelinated corpus callosum (CC) in adult mice. Male C57BL/6 mice fed 0.3% cuprizone for 12 weeks exhibited CC atrophy with chronic demyelination, astrogliosis, and microglial activation. Syngeneic iNSCs were transplanted into the CC after ending cuprizone and perfused for neuropathology 2 weeks later. Magnetic resonance imaging (MRI) sequences for magnetization transfer ratio (MTR), diffusion-weighted imaging (T2), and diffusion tensor imaging (DTI) quantified CC pathology in live mice before and after iNSC transplantation. Each MRI technique detected progressive CC pathology. Mice that received iNSCs had normalized DTI radial diffusivity, and reduced astrogliosis post-imaging. A motor skill task that engages the CC is Miss-step wheel running, which demonstrated functional deficits from cuprizone demyelination. Transplantation of iNSCs resulted in marked recovery of running velocity. Neuropathology after wheel running showed that iNSC grafts significantly increased host oligodendrocytes and proliferating oligodendrocyte progenitors, while modulating axon damage. Transplanted iNSCs differentiated along astrocyte and oligodendrocyte lineages, without myelinating, and many remained neural stem cells. Our findings demonstrate the applicability of neuroimaging and functional assessments for pre-clinical interventional trials during chronic demyelination and detect improved function from iNSC transplantation. Directly reprogramming fibroblasts into iNSCs facilitates the future translation towards exogenous autologous cell therapies.

Salvianolic acid B inhibits growth of head and neck squamous cell carcinoma in vitro and in vivo via cyclooxygenase-2 and apoptotic pathways.

Overexpression of cyclooxygenase-2 (COX-2) in oral mucosa has been associated with increased risk of head and neck squamous cell carcinoma (HNSCC). Celecoxib is a nonsteroidal anti-inflammatory drug, which inhibits COX-2 but not COX-1. This selective COX-2 inhibitor holds promise as a cancer preventive agent. Concerns about cardiotoxicity of celecoxib, limits its use in long-term chemoprevention and therapy. Salvianolic acid B (Sal-B) is a leading bioactive component of Salvia miltiorrhiza Bge, which is used for treating neoplastic and chronic inflammatory diseases in China. The purpose of this study was to investigate the mechanisms by which Sal-B inhibits HNSCC growth. Sal-B was isolated from S. miltiorrhiza Bge by solvent extraction followed by 2 chromatographic steps. Pharmacological activity of Sal-B was assessed in HNSCC and other cell lines by estimating COX-2 expression, cell viability and caspase-dependent apoptosis. Sal-B inhibited growth of HNSCC JHU-022 and JHU-013 cells with IC(50) of 18 and 50 microM, respectively. Nude mice with HNSCC solid tumor xenografts were treated with Sal-B (80 mg/kg/day) or celecoxib (5 mg/kg/day) for 25 days to investigate in vivo effects of the COX-2 inhibitors. Tumor volumes in Sal-B treated group were significantly lower than those in celecoxib treated or untreated control groups (p < 0.05). Sal-B inhibited COX-2 expression in cultured HNSCC cells and in HNSCC cells isolated from tumor xenografts. Sal-B also caused dose-dependent inhibition of prostaglandin E(2) synthesis, either with or without lipopolysaccharide stimulation. Taken together, Sal-B shows promise as a COX-2 targeted anticancer agent for HNSCC prevention and treatment.

Subcutaneous Administration of Angiotensin-(1-7) Improves Recovery after Traumatic Brain Injury in Mice.

Angiotensin II (Ang II)-mediated activation of its type I receptor (AT1R) in the central nervous system promotes glial proliferation, local inflammation, and a decrease of cerebral blood flow. Angiotensin-(1-7) (Ang-(1-7))-an Ang II derivative peptide-signals through the Mas receptor (MasR) in opposition to Ang II/AT1R, promoting anti-inflammatory, vasodilatory, and neuroprotective effects. As our laboratory has previously demonstrated beneficial effects of AT1R inhibition following controlled cortical impact (CCI) in mice, we asked whether activation of Ang-(1-7)/MasR signaling would also be beneficial in this model. Adult male C57BL/6 mice were injured by CCI. Ang-(1-7) or vehicle was administered subcutaneously (S.Q.) at 1 mg/kg/day at 1 or 6 h post-injury, until animals were sacrificed at 3 or 29 days post-injury (dpi). Ang-(1-7) attenuated motor deficits at 3 dpi and improved performance in the Morris Water Maze at 28 dpi. Brain histology or magnetic resonance imaging (MRI) indicated that Ang-(1-7)-treated mice had smaller lesion volumes at 3, 10, 24, and 29 dpi. Pre-treatment with A779, a MasR antagonist, prevented Ang-(1-7) from reducing lesion volume at 3 dpi, suggesting that the benefits of Ang-(1-7) were MasR-dependent. Immunohistochemistry revealed that Ang-(1-7) reduced microgliosis at 3 and 29 dpi, and astrogliosis at 29 dpi. Ang-(1-7) decreased neuronal and capillary loss at 29 dpi. In summary, S.Q. administration of Ang-(1-7) after injury had anti-inflammatory, neuroprotective, and cerebrovascular-protective actions leading to improved functional and pathological recovery in a mouse model of traumatic brain injury (TBI). These data show for the first time that Ang-(1-7) has potential therapeutic use for TBI.

Open-source QSAR models for pKa prediction using multiple machine learning approaches.

BACKGROUND: The logarithmic acid dissociation constant pKa reflects the ionization of a chemical, which affects lipophilicity, solubility, protein binding, and ability to pass through the plasma membrane. Thus, pKa affects chemical absorption, distribution, metabolism, excretion, and toxicity properties. Multiple proprietary software packages exist for the prediction of pKa, but to the best of our knowledge no free and open-source programs exist for this purpose. Using a freely available data set and three machine learning approaches, we developed open-source models for pKa prediction. METHODS: The experimental strongest acidic and strongest basic pKa values in water for 7912 chemicals were obtained from DataWarrior, a freely available software package. Chemical structures were curated and standardized for quantitative structure-activity relationship (QSAR) modeling using KNIME, and a subset comprising 79% of the initial set was used for modeling. To evaluate different approaches to modeling, several datasets were constructed based on different processing of chemical structures with acidic and/or basic pKas. Continuous molecular descriptors, binary fingerprints, and fragment counts were generated using PaDEL, and pKa prediction models were created using three machine learning methods, (1) support vector machines (SVM) combined with k-nearest neighbors (kNN), (2) extreme gradient boosting (XGB) and (3) deep neural networks (DNN). RESULTS: The three methods delivered comparable performances on the training and test sets with a root-mean-squared error (RMSE) around 1.5 and a coefficient of determination (R2) around 0.80. Two commercial pKa predictors from ACD/Labs and ChemAxon were used to benchmark the three best models developed in this work, and performance of our models compared favorably to the commercial products. CONCLUSIONS: This work provides multiple QSAR models to predict the strongest acidic and strongest basic pKas of chemicals, built using publicly available data, and provided as free and open-source software on GitHub.