Molecular property prediction based on graph structure learning.

MOTIVATION: Molecular property prediction (MPP) is a fundamental but challenging task in the computer-aided drug discovery process. More and more recent works employ different graph-based models for MPP, which have achieved considerable progress in improving prediction performance. However, current models often ignore relationships between molecules, which could be also helpful for MPP. RESULTS: For this sake, in this article we propose a graph structure learning (GSL) based MPP approach, called GSL-MPP. Specifically, we first apply graph neural network (GNN) over molecular graphs to extract molecular representations. Then, with molecular fingerprints, we construct a molecule similarity graph (MSG). Following that, we conduct GSL on the MSG, i.e. molecule-level GSL, to get the final molecular embeddings, which are the results of fuzing both GNN encoded molecular representations and the relationships among molecules. That is, combining both intra-molecule and inter-molecule information. Finally, we use these molecular embeddings to perform MPP. Extensive experiments on 10 various benchmark datasets show that our method could achieve state-of-the-art performance in most cases, especially on classification tasks. Further visualization studies also demonstrate the good molecular representations of our method. AVAILABILITY AND IMPLEMENTATION: Source code is available at https://github.com/zby961104/GSL-MPP.

All-stage targeted red blood cell membrane-coated docetaxel nanocrystals for glioma treatment.

Challenges for glioma treatment with nanomedicines include physio-anatomical barriers (the blood-brain barrier and blood-brain tumor barrier), low drug loading capacity, and limited circulation time. Here, a red blood cell membrane-coated docetaxel drug nanocrystal (pV-RBCm-NC(DTX)), modified with pHA-VAP (pV) for all-stage targeting of glioma, was designed. The NC(DTX) core exhibited a high drug loading capacity but low in vivo stability, and the RBCm coating significantly enhanced the stability and prolonged in vivo circulation. Moreover, the Y-shaped targeting ligand pV was modified by a mild avidin-biotin interaction, which endowed RBCm-NC(DTX) with superior barrier-crossing ability and therapeutic efficacy. The integration of nanocrystal technology, cell membrane coating, and the avidin-biotin insertion method into this active targeting biomimetic formulation represents a promising drug delivery strategy for glioma.

All-stage targeted therapy for the brain metastasis from triple-negative breast cancer.

Brain metastasis is a common and serious complication of breast cancer, which is commonly associated with poor survival and prognosis. In particular, the treatment of brain metastasis from triple-negative breast cancer (BM-TNBC) has to face the distinct therapeutic challenges from tumor heterogeneity, circulating tumor cells (CTCs), blood-brain barrier (BBB) and blood-tumor barrier (BTB), which is in unmet clinical needs. Herein, combining with the advantages of synthetic and natural targeting moieties, we develop a "Y-shaped" peptide pVAP-decorated platelet-hybrid liposome drug delivery system to address the all-stage targeted drug delivery for the whole progression of BM-TNBC. Inherited from the activated platelet, the hybrid liposomes still retain the native affinity toward CTCs. Further, the peptide-mediated targeting to breast cancer cells and transport across BBB/BTB are demonstrated in vitro and in vivo. The resultant delivery platform significantly improves the drug accumulation both in orthotopic breast tumors and brain metastatic lesions, and eventually exhibits an outperformance in the inhibition of BM-TNBC compared with the free drug. Overall, this work provides a promising prospect for the comprehensive treatment of BM-TNBC, which could be generalized to other cell types or used in imaging platforms in the future.

Influence of lung cancer model characteristics on tumor targeting behavior of nanodrugs.

Evidence is mounting that there is a significant gap between the antitumor efficacy of nanodrugs in preclinical mouse tumor models and in clinical human tumors, and that differences in tumor models are likely to be responsible for this gap. Herein, we investigated the enhanced permeability and retention (EPR) effect in mouse lung cancer models with different tumor growth rates, volumes and locations, and analyzed the nanodrug tumor targeting behaviors limited by tumor vascular pathophysiological characteristics in various tumor models. The results showed that the fast-growing tumors were characterized by lower vascular tight junctions, leading to higher vascular paracellular transport activity and nanodrug tumor accumulation. The paracellular transport activity increased with the growth of tumor, but the vascular density and transcellular transport activity decreased, and as a result, the average tumor accumulation of passive targeting nanodrugs decreased. Orthotopic tumors were rich in blood vessels, but had low vascular transcellular and paracellular transport activities, making it difficult for nanodrug accumulation in orthotopic tumors via passive targeting strategies. The antitumor efficacy of passive targeting nanodrugs in various lung cancer-bearing mice validated the aforementioned nanodrug accumulation behavior, and nanodrugs based on the angiogenesis-tumor sequential targeting strategy achieved obviously improved efficacy in orthotopic lung cancer-bearing mice. These results suggest that the EPR effect varies in different tumor models and should not be used as a universal targeting strategy for antitumor nanodrugs. Besides, attention should be paid to the animal tumor models in the evaluation of nanodrugs so as to avoid exaggerating the antitumor efficacy.

Identifying essential proteins from protein-protein interaction networks based on influence maximization.

BACKGROUND: Essential proteins are indispensable to the development and survival of cells. The identification of essential proteins not only is helpful for the understanding of the minimal requirements for cell survival, but also has practical significance in disease diagnosis, drug design and medical treatment. With the rapidly amassing of protein-protein interaction (PPI) data, computationally identifying essential proteins from protein-protein interaction networks (PINs) becomes more and more popular. Up to now, a number of various approaches for essential protein identification based on PINs have been developed. RESULTS: In this paper, we propose a new and effective approach called iMEPP to identify essential proteins from PINs by fusing multiple types of biological data and applying the influence maximization mechanism to the PINs. Concretely, we first integrate PPI data, gene expression data and Gene Ontology to construct weighted PINs, to alleviate the impact of high false-positives in the raw PPI data. Then, we define the influence scores of nodes in PINs with both orthological data and PIN topological information. Finally, we develop an influence discount algorithm to identify essential proteins based on the influence maximization mechanism. CONCLUSIONS: We applied our method to identifying essential proteins from saccharomyces cerevisiae PIN. Experiments show that our iMEPP method outperforms the existing methods, which validates its effectiveness and advantage.

In vivo self-assembled drug nanocrystals for metastatic breast cancer all-stage targeted therapy.

Chemotherapy is still the mainstay treatment for metastatic triple-negative breast cancers (TNBC) currently in clinical practice. The unmet needs of chemotherapy for metastatic TNBC are mainly from the insufficient drug delivery and unavailable targeting strategy that thwart the whole progression of metastatic TNBC. The in vivo ligands-mediated active targeting efficiency is usually affected by protein corona. While, the protein corona-bridged natural targeting, in turn, provides a new way for specific drug delivery. Herein, we develop a novel metastatic progression-oriented in vivo self-assembled Cabazitaxel nanocrystals (CNC) delivery system (PC/CNC) through the CNC automatically absorbing functional plasma proteins (transferrin, apolipoprotein A-IV and apolipoprotein E) in vivo, aiming to achieve the simultaneously targeted delivery to primary tumors, circulating tumor cells and metastatic lesions. With the unique advantages of superhigh drug-loading and protein corona empowered active targeting properties to tumor cells, HUVECs, active-platelets and blood-brain barrier/blood-tumor barrier, the PC/CNC exhibits a significantly improved therapeutic effect in metastatic TNBC therapy compared with free drug and CNC-loaded liposomes.

All-stage targeted therapy for glioblastoma based on lipid membrane coated cabazitaxel nanocrystals.

Glioblastoma (GBM) is the most aggressive brain tumor with poor prognosis and frequent recurrence. The blood-brain barrier (BBB), blood-brain tumor barrier (BBTB) hinder the entry of therapeutics into the glioma region. Vasculogenic mimicry (VM) formed by invasive glioma cells is also related to recurrence of GBM. VAP is a D-peptide ligand of GRP78 protein overexpressed on BBTB, VM, and glioma cells but not on normal tissues. Besides, p-hydroxybenzoic acid (pHA) can effectively traverse the BBB. Herein we developed an all-stage glioma-targeted cabazitaxel (CBZ) nanocrystal loaded liposome modified with a "Y" shaped targeting ligand composed of pHA and VAP (pV-Lip/cNC). The pure drug nanocrystal core provided high drug loading, while lipid membrane promoted the stability and circulation time. pV-Lip/cNC exhibited excellent glioma homing, barriers crossing, and tumor spheroid penetrating capability in vitro. Treatment of pV-Lip/cNC displayed enhanced CBZ accumulation in glioma and anti-glioma effect with a median survival time (53 days) significantly longer than that of cNC loaded liposomes modified with either single ligand (42 days for VAP and 45 days for pHA) in the murine orthotopic GBM model. These results indicated pV-Lip/cNC could traverse the BBB and BBTB, destruct VM, and finally kill glioma cells to realize all-stage glioma therapy.

Protein-protein interaction prediction based on ordinal regression and recurrent convolutional neural networks.

BACKGROUND: Protein protein interactions (PPIs) are essential to most of the biological processes. The prediction of PPIs is beneficial to the understanding of protein functions and thus is helpful to pathological analysis, disease diagnosis and drug design etc. As the amount of protein data is growing fast in the post genomic era, high-throughput experimental methods are expensive and time-consuming for the prediction of PPIs. Thus, computational methods have attracted researcher's attention in recent years. A large number of computational methods have been proposed based on different protein sequence encoders. RESULTS: Notably, the confidence score of a protein sequence pair could be regarded as a kind of measurement to PPIs. The higher the confidence score for one protein pair is, the more likely the protein pair interacts. Thus in this paper, a deep learning framework, called ordinal regression and recurrent convolutional neural network (OR-RCNN) method, is introduced to predict PPIs from the perspective of confidence score. It mainly contains two parts: the encoder part of protein sequence pair and the prediction part of PPIs by confidence score. In the first part, two recurrent convolutional neural networks (RCNNs) with shared parameters are applied to construct two protein sequence embedding vectors, which can automatically extract robust local features and sequential information from the protein pairs. Based on it, the two embedding vectors are encoded into one novel embedding vector by element-wise multiplication. By taking the ordinal information behind confidence score into consideration, ordinal regression is used to construct multiple sub-classifiers in the second part. The results of multiple sub-classifiers are aggregated to obtain the final confidence score. Following that, the existence of PPIs is determined by the confidence score. We set a threshold [Formula: see text], and say the interaction exists between the protein pair if its confidence score is bigger than [Formula: see text]. CONCLUSIONS: We applied our method to predict PPIs on data sets S. cerevisiae and Homo sapiens. Through experimental verification, our method outperforms state-of-the-art PPI prediction models.

Impact of macroporous silica nanoparticles at sub-50nm on bio-behaviors and biosafety in drug-resistant cancer models.

The in vivo bio-behaviors and biosafety of nanoparticles were demonstrated to be closely correlated with particle sizes, which illustrated whether they could be used as the effective drug delivery carriers. Though tumor penetration capabilities of the small pore sized-mesoporous silica nanoparticles (MSNs) were reported to be in a particle size-dependent manner, the size effects of large pore sized-MSNs on the safe and effective cancer resistance treatment, especially at sub-50 nm, were not explicitly evaluated. In this study, we fabricate the 20 nm and 50 nm MSNs, and aim at investigating their difference in tumor accumulation, penetration, retention and toxicity both in vitro and in vivo. Our results showed that these two particle sized-MSNs possessed the excellent tumor penetration capabilities both in resistant human hepatocellular carcinoma cells-cultured spheroids and in the corresponding xenograft mice models, but the 50 nm MSNs seemed to have the better tumor accumulation and retention effects than the 20 nm MSNs. Moreover, the 50 nm MSNs displayed the lower toxicities than the 20 nm MSNs whatever on resistant cancer cell lines or on zebrafish embryos, indicating the greater systematic biosafety. In a word, our data provide the evidence that selection of the large pore-sized MSNs at the appropriate particle size (not the smaller the better) as bio-macromolecule nanocarriers will play a key role in the safe and effective treatment against cancer resistance.

Cancer-targeted and intracellular delivery of Bcl-2-converting peptide with functional macroporous silica nanoparticles for biosafe treatment.

Therapeutic peptide, NuBCP-9 (N9) as a Bcl-2 functional converter, has been demonstrated to have the remarkable anticancer efficiency in Bcl-2-abundant cancer. However, it faced technical challenges in clinical use, such as the low bioavailability, the easily-destroyed bio-stability, and the insusceptibility to cellular interior. With the potential of mesoporous silica nanoparticles (MSNs) as the promising delivery vehicle of therapeutic macromolecules, we developed a kind of MSNs with the surface coating of folic acid (FA) for cancer cell targeting and with the macropore loading of N9 peptide for cancer therapy. Our results showed that the functional MSNs had the relatively greater biosafety than the naked MSNs in zebrafish models, leading to less than 30% embryo of death at 200 µg/ml, which could further specifically target the folate receptor (FR)-overexpressed cervical cancer HeLa cells instead of FR-negative normal embryonic kidney HEK 293T cells in a FA-competitive manner. N9 peptide with the delivery of functional MSNs could be internalized by HeLa cells, and co-localized with mitochondria in a Bcl-2-dependent manner. Moreover, N9 peptide delivered by FA-modified MSNs displayed the excellent anticancer efficiency with great selectivity, inducing approximately 52% HeLa cells into apoptosis. In summary, our results illustrated the potential of functional MSNs with large pore size as an efficient nanocarrier for the intracellular delivery of peptide drugs with targeting proteins to realize cancer therapy.

A Natural-language-based Visual Query Approach of Uncertain Human Trajectories.

Visual querying is essential for interactively exploring massive trajectory data. However, the data uncertainty imposes profound challenges to fulfill advanced analytics requirements. On the one hand, many underlying data does not contain accurate geographic coordinates, e.g., positions of a mobile phone only refer to the regions (i.e., mobile cell stations) in which it resides, instead of accurate GPS coordinates. On the other hand, domain experts and general users prefer a natural way, such as using a natural language sentence, to access and analyze massive movement data. In this paper, we propose a visual analytics approach that can extract spatial-temporal constraints from a textual sentence and support an effective query method over uncertain mobile trajectory data. It is built up on encoding massive, spatially uncertain trajectories by the semantic information of the POls and regions covered by them, and then storing the trajectory documents in text database with an effective indexing scheme. The visual interface facilitates query condition specification, situation-aware visualization, and semantic exploration of large trajectory data. Usage scenarios on real-world human mobility datasets demonstrate the effectiveness of our approach.

Macroporous organosilicon nanocomposites co-deliver Bcl2-converting peptide and chemotherapeutic agent for synergistic treatment against multidrug resistant cancer.

Therapeutic biomacromolecules are confronted with in vivo challenges of low bio-stability and poor tumor tissue-penetration. Herein, we report for the first time, our development and characterization of a hybrid nanocomposite for delivering a Bcl-2-converting peptide (NuBCP9, N9 hereafter) and testing its efficacy alone or together with doxorubicin (DOX). The hybrid nanocomposite is composed of the internal large pore sized-mesoporous silica nanoparticles (MSNs) and the external highly-branched polyamidoamine (PAMAM) dendrimers, into which N9 peptide and DOX were encapsulated for the different sub-cellular delivery to treat drug-resistant cancer. The nanocomposite possessed the particle and pore sizes of ~37 nm and ~8 nm, which displayed the superior tumor penetration capacity over naked MSNs both in cultured-3D tumor sphere and in live animal models. Moreover, the dual drug nanocomposite exhibited a great synergistic anticancer effect on Bcl-2-positive cancer cells in vitro and animals with the negligible toxic side effects. The tumor inhibition rate of the nanocomposite (89%) was five times as much as the two drugs combination. This design provides a new effective, safe and versatile strategy to fabricate large pore-sized MSNs with the organic-inorganic hybrid framework to concurrently transport therapeutic peptides and chemotherapeutics to the specific sub-cellular locations for the synergistic cancer therapy and drug resistance reversal, which has significant impact on the development of improved cancer therapeutics.

Dendrimer-like mesoporous silica nanospheres with suitable surface functionality to combat the multidrug resistance.

Multidrug resistance (MDR), as a major obstacle in cancer therapy, has resulted in over 90% of cancer chemotherapeutic failure. Mesoporous silica nanospheres (MSNs) have been demonstrated to be tuned with large pore sizes, mediating the MDR-reversal effects. However, the study that surface functionality of the large pore sized-MSNs affects the MDR-overcoming effects hasn't been extensively studied. In this study, we developed a new dendrimer-like MSNs delivery system based on a rational synthesis strategy and further modified MSNs with various surface functionalities to evaluate their roles in overcoming cancer MDR. Our results showed that the small particle sized-MSNs could be fabricated with dendrimer-like internal structure, resulting in the large pore size of 9 nm. Surface functionality of MSNs, especially hydroxylation and carboxylation, largely improved the intra-nuclear delivery and therapeutic efficiency of DOX for MCF7/ADR cells, which was not up to inhibiting P-gp expression but significantly increasing the intracellular drug accumulation of over 90% even under the strong drug efflux. This study indicates that surface functionality design strategy may display the potential of the large pore sized-MSNs as the efficient chemotherapeutic carriers to combat MDR.

Macroporous silica nanoparticles for delivering Bcl2-function converting peptide to treat multidrug resistant-cancer cells.

The abundance of B cell lymphoma gene 2 (Bcl-2) is closely correlated with the resistance of cancer cells to chemotherapeutic agents, and a peptide derived from orphan nuclear receptor Nur77 can convert Bcl-2 from a protector to a killer of cancer cells. However, successful application of the Bcl-2-converting peptide to treat drug-resistant cancer cells depends on an efficient delivery carrier. Mesoporous silica nanoparticles (MSNs) have been extensively studied as promising candidates for small molecule drug delivery. However, the effective encapsulation and intracellular delivery of peptides using small pore-sized MSNs still remain a great technical challenge. In this paper, an effective delivery platform for Bcl-2-converting peptide was fabricated by us to treat multidrug resistant-cancer cells via tuning the surface functionality of macroporous silica nanoparticles. The resulting large-sized pore silica nanoparticles, especially those modified with thiol group, exhibited the high Bcl-2-converting peptide-loading efficiency of over 40%. Moreover, the peptide induced MCF7/DOX cells into apoptotic status by penetrating cytomembrane into mitochondria and being bound with Bcl-2 to expose the BH3 domain with the aid of various surface functionalities-decorated MSNs. In particular, amine-modified surface of MSNs caused the greater influence on the cell apoptosis-inducing effects of peptide in comparison with other functionalities-modified ones. Taken together, our study, for the first time, demonstrates a special approach towards pore size and surface functionality-collectively modulated silica-based nanostructural material for effective delivery of bio-macromolecules (e.g., Bcl-2-converting peptide) to treat the multidrug resistant-cancer cells with elevated Bcl-2 levels.

Spectroscopic investigations on the simulated solar light induced photodegradation of 4-nitrophenol by using three novel copper(II) porphyrin-TiO2 photocatalysts.

Three porphyrins containing different functional groups (-OH, C-O2C2H5, -COOH), 5-(4-hydroxy) phenyl-10,15,20-triphenyl porphyrin (1a), 5-(4-ethylacetatatomethoxy) phenyl-10,15,20-triphenyl porphyrin (1b), 5-(4-carboxylatomethoxy) phenyl-10,15,20-triphenyl porphyrin (1c), were synthesized and characterized spectroscopically. The CuPp(2a, 2b, 2c)-TiO2 photocatalysts were then prepared and characterized by X-ray diffraction (XRD), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), UV-vis diffuse reflectance spectroscopy (UV-vis-DRS), Fourier-transform infrared spectroscopy (FT-IR). The photocatalytic activities of the photocatalysts were investigated by carrying out the photodegradation of 4-nitrophenol in aqueous solution under simulated solar irradiation. It was found that the CuPp(2a, 2b, 2c)-TiO2 enhanced the photocatalytic efficiency of bare TiO2 in photodegrading the 4-NP due to the interaction between CuPp(2a, 2b, 2c) and TiO2, resulted in the enhancement of the photogenerated electron-hole separation. The reasons of this enhanced photocatalytic activity were also discussed. Based on the present study, it could be considered as a promising photocatalyst for the further industrial application.

[Studies on the anion recognition of N-aryl-N'-(4-ethyloxybenzoyl) thioureas derivatives by spectroscopy].

The interactions of N-aryl-N'-(4-ethyloxybenzoyl) thioureas derivatives and anions such as F-, Cl-, Br-, I-, CH3COO-, H2POT4-, HSO4- and NO3- in DMSO were investigated by UV-Vis and 1H NMR spectroscopy. The results showed that hydrogen--bonding complexes were formed between receptors and the three anions such as F-, CH3COO- and H2PO4-, and the solutions turned light yellow from colorlessness, but there had no evident binding with Cl-, Br-, I-, HSO4- and NO3-. It was regular that the two receptors had different binding ability with F-, CH3COO- and H2PO4-. For the same anion, the association constants followed the trend: receptor 1a > receptor 1b. The UV-Vis data indicated that a 1 : 1 stoichiometry complex was formed.

[UV spectroscopic characterization of the inclusion interaction of beta-cyclodextrin and sym-diphenyl-thiourea and sym-diphenyl-urea].

The inclusion interaction of beta-cyclodextrin and sym-diphenyl-thiourea and sym-diphenyl-urea was studied by UV spectra. The stoichiometry ratio for the formation of the inclusion complexes was determined by Hildebrand-Benesi equation linear analysis and molar ratio method. The standard molar Gibbs energies, enthalpies, and entropies were derived for the inclusion process by Ks at different temperatrues. The result showed that the host:guest ratio of inclusion complex between the two diphenyl compounds and beta-CD is 2 : 1, the stability constant (Ks) of 2 : 1 inclusion complexes was higher than that of 1 : 1 inclusion complexes due to cooperative binding in the close two hydrophobic cyclodextrin cavities, and the association of the guest molecule with beta-CD was favored by enthalpy changes, proving that the Van der Waals interaction and the dipole-dipole interaction were main binding forces of cyclodextrin inclusion complex.