Marker-controlled watershed with deep edge emphasis and optimized H-minima transform for automatic segmentation of densely cultivated 3D cell nuclei.

BACKGROUND: The segmentation of 3D cell nuclei is essential in many tasks, such as targeted molecular radiotherapies (MRT) for metastatic tumours, toxicity screening, and the observation of proliferating cells. In recent years, one popular method for automatic segmentation of nuclei has been deep learning enhanced marker-controlled watershed transform. In this method, convolutional neural networks (CNNs) have been used to create nuclei masks and markers, and the watershed algorithm for the instance segmentation. We studied whether this method could be improved for the segmentation of densely cultivated 3D nuclei via developing multiple system configurations in which we studied the effect of edge emphasizing CNNs, and optimized H-minima transform for mask and marker generation, respectively. RESULTS: The dataset used for training and evaluation consisted of twelve in vitro cultivated densely packed 3D human carcinoma cell spheroids imaged using a confocal microscope. With this dataset, the evaluation was performed using a cross-validation scheme. In addition, four independent datasets were used for evaluation. The datasets were resampled near isotropic for our experiments. The baseline deep learning enhanced marker-controlled watershed obtained an average of 0.69 Panoptic Quality (PQ) and 0.66 Aggregated Jaccard Index (AJI) over the twelve spheroids. Using a system configuration, which was otherwise the same but used 3D-based edge emphasizing CNNs and optimized H-minima transform, the scores increased to 0.76 and 0.77, respectively. When using the independent datasets for evaluation, the best performing system configuration was shown to outperform or equal the baseline and a set of well-known cell segmentation approaches. CONCLUSIONS: The use of edge emphasizing U-Nets and optimized H-minima transform can improve the marker-controlled watershed transform for segmentation of densely cultivated 3D cell nuclei. A novel dataset of twelve spheroids was introduced to the public.

Assembly of Bleomycin Saccharide-Decorated Spherical Nucleic Acids.

Glyco-decorated spherical nucleic acids (SNAs) may be attractive delivery vehicles, emphasizing the sugar-specific effect on the outer sphere of the construct and at the same time hiding unfavorable distribution properties of the loaded oligonucleotides. As examples of such nanoparticles, tripodal sugar constituents of bleomycin were synthesized and conjugated with a fluorescence-labeled antisense oligonucleotide (AONARV7). Successive copper(I)-catalyzed azide-alkyne and strain-promoted alkyne-nitrone cycloadditions (SPANC) were utilized for the synthesis. Then, the glyco-AONARV7 conjugates were hybridized with complementary strands of a C60-based molecular spherical nucleic acid (i.e., a hybridization-mediated carrier). The formation and stability of these assembled glyco-decorated SNAs were evaluated by polyacrylamide gel electrophoresis (PAGE), UV melting profile analysis, and time-resolved fluorescence spectroscopy. Association constants were extracted from time-resolved fluorescence data. Preliminary cellular uptake experiments of the glyco-AONARV7 conjugates (120 nM solutions) and of the corresponding glyco-decorated SNAs (10 nM solutions) with human prostate cancer cells (PC3) showed an efficient uptake in each case. A marked variation in intracellular distribution was observed.

Controlled Monofunctionalization of Molecular Spherical Nucleic Acids on a Buckminster Fullerene Core.

An azide-functionalized 12-armed Buckminster fullerene has been monosubstituted in organic media with a substoichiometric amount of cyclooctyne-modified oligonucleotides. Exposing the intermediate products then to the same reaction (i.e., strain-promoted alkyne-azide cycloaddition, SPAAC) with an excess of slightly different oligonucleotide constituents in an aqueous medium yields molecularly defined monofunctionalized spherical nucleic acids (SNAs). This procedure offers a controlled synthesis scheme in which one oligonucleotide arm can be functionalized with labels or other conjugate groups (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid, DOTA, and Alexa-488 demonstrated), whereas the rest of the 11 arms can be left unmodified or modified by other conjugate groups in order to decorate the SNAs' outer sphere. Extra attention has been paid to the homogeneity and authenticity of the C60-azide scaffold used for the assembly of full-armed SNAs.

New In Vitro-In Silico Approach for the Prediction of In Vivo Performance of Drug Combinations.

Pharmacokinetic (PK) studies improve the design of dosing regimens in preclinical and clinical settings. In complex diseases like cancer, single-agent approaches are often insufficient for an effective treatment, and drug combination therapies can be implemented. In this work, in silico PK models were developed based on in vitro assays results, with the goal of predicting the in vivo performance of drug combinations in the context of cancer therapy. Combinations of reference drugs for cancer treatment, gemcitabine and 5-fluorouracil (5-FU), and repurposed drugs itraconazole, verapamil or tacrine, were evaluated in vitro. Then, two-compartment PK models were developed based on the previous in vitro studies and on the PK profile reported in the literature for human patients. Considering the quantification parameter area under the dose-response-time curve (AUCeffect) for the combinations effect, itraconazole was the most effective in combination with either reference anticancer drugs. In addition, cell growth inhibition was itraconazole-dose dependent and an increase in effect was predicted if itraconazole administration was continued (24-h dosing interval). This work demonstrates that in silico methods and AUCeffect are powerful tools to study relationships between tissue drug concentration and the percentage of cell growth inhibition over time.

Label-Free Analysis with Multiple Parameters Separates G Protein-Coupled Receptor Signaling Pathways.

Real-time label-free techniques are used to profile G protein-coupled receptor (GPCR) signaling pathways in living cells. However, interpreting the label-free signal responses is challenging, and previously reported methods do not reliably separate pathways from each other. In this study, a continuous angular-scanning surface plasmon resonance (SPR) technique is utilized for measuring label-free GPCR signal profiles. We show how the continuous angular-scanning ability, measuring up to nine real-time label-free parameters simultaneously, results in more information-rich label-free signal profiles for different GPCR pathways, providing a more accurate pathway separation. For this, we measured real-time full-angular SPR response curves for Gs, Gq, and Gi signaling pathways in living cells. By selecting two of the most prominent label-free parameters: the full SPR curve angular and intensity shifts, we present how this analysis approach can separate each of the three signaling pathways in a straightforward single-step analysis setup, without concurrent use of signal inhibitors or other response modulating compounds.

Structure and Dynamics of Thermosensitive pDNA Polyplexes Studied by Time-Resolved Fluorescence Spectroscopy.

Combining multiple stimuli-responsive functionalities into the polymer design is an attractive approach to improve nucleic acid delivery. However, more in-depth fundamental understanding how the multiple functionalities in the polymer structures are influencing polyplex formation and stability is essential for the rational development of such delivery systems. Therefore, in this study the structure and dynamics of thermosensitive polyplexes were investigated by tracking the behavior of labeled plasmid DNA (pDNA) and polymer with time-resolved fluorescence spectroscopy using fluorescence resonance energy transfer (FRET). The successful synthesis of a heterofunctional poly(ethylene glycol) (PEG) macroinitiator containing both an atom transfer radical polymerization (ATRP) and reversible addition-fragmentation chain-transfer (RAFT) initiator is reported. The use of this novel PEG macroinitiator allows for the controlled polymerization of cationic and thermosensitive linear triblock copolymers and labeling of the chain-end with a fluorescent dye by maleimide-thiol chemistry. The polymers consisted of a thermosensitive poly(N-isopropylacrylamide) (PNIPAM, N), hydrophilic PEG (P), and cationic poly(2-(dimethylamino)ethyl methacrylate) (PDMAEMA, D) block, further referred to as NPD. Polymer block D chain-ends were labeled with Cy3, while pDNA was labeled with FITC. The thermosensitive NPD polymers were used to prepare pDNA polyplexes, and the effect of the N/P charge ratio, temperature, and composition of the triblock copolymer on the polyplex properties were investigated, taking nonthermosensitive PD polymers as the control. FRET was observed both at 4 and 37 °C, indicating that the introduction of the thermosensitive PNIPAM block did not compromise the polyplex structure even above the polymer's cloud point. Furthermore, FRET results showed that the NPD- and PD-based polyplexes have a less dense core compared to polyplexes based on cationic homopolymers (such as PEI) as reported before. The polyplexes showed to have a dynamic character meaning that the polymer chains can exchange between the polyplex core and shell. Mobility of the polymers allow their uniform redistribution within the polyplex and this feature has been reported to be favorable in the context of pDNA release and subsequent improved transfection efficiency, compared to nondynamic formulations.

Effects of nanofibrillated cellulose hydrogels on adipose tissue extract and hepatocellular carcinoma cell spheroids in freeze-drying.

The aim of this study was to evaluate the effects of two nanofibrillated cellulose (NFC) hydrogels on two human derivatives during freeze-drying. Native NFC hydrogel is a suitable platform to culture 3D cell spheroids and a hydrogel processed further, called anionic NFC (ANFC) hydrogel, is an excellent platform for controlled release of proteins. Moreover, it has been shown to be compatible with freeze-drying when correct lyoprotectants are implemented. Freeze-drying is a method, where substance is first frozen, and then vacuum dried trough sublimation of water in order to achieve dry matter without the loss of the original three-dimensional structures. The first chosen human derivative was adipose tissue extract (ATE) which is a cell-free growth factor-rich preparation capable of promoting growth of regenerative cells. The release of growth factors from the freeze-dried mixture of ATE and ANFC was compared to that of non-freeze-dried control mixtures. The release profiles remained at the same level after freeze-drying. The second derivative was hepatocellular carcinoma (HepG2) cell spheroids which were evaluated before and after freeze-drying. The 3D structure of the HepG2 cell spheroids was preserved and the spheroids retained 18% of their metabolic activity after rehydration. However, the freeze-dried and rehydrated HepG2 cell spheroids did not proliferate and the cell membrane was damaged by fusion and formation of crystals.

Differences in definitive endoderm induction approaches using growth factors and small molecules.

Definitive endoderm (DE) is the first stage of human pluripotent stem cell (hPSC) differentiation into hepatocyte-like cells. Developing human liver cell models for pharmaceutical applications is highly demanding. Due to the vast number of existing protocols to generate DE cells from hPSCs, we aimed to compare the specificity and efficiency of selected published differentiation conditions. We differentiated two hPSC lines (induced PSC and embryonic stem cell) to DE cells on Matrigel matrix using growth factors (Activin A and Wnt-3a) and small molecules (sodium butyrate and IDE 1) in different combinations. By studying dynamic changes during 6 days in cell morphology and the expression of markers for pluripotency, DE, and other germ layer lineages, we found that Activin A is essential for DE differentiation, while Wnt-3a and sodium butyrate are dispensable. Although sodium butyrate exerted rapid DE differentiation kinetics, it caused massive cell death and could not generate sufficient cells for further differentiation and applications. We further discover that IDE 1 could not induce DE as reported previously. Hereby, we compared different conditions for DE induction and found an effective six day-protocol to obtain DE cells for the further differentiation and applications.

Pectin and Mucin Enhance the Bioadhesion of Drug Loaded Nanofibrillated Cellulose Films.

PURPOSE: Bioadhesion is an important property of biological membranes, that can be utilized in pharmaceutical and biomedical applications. In this study, we have fabricated mucoadhesive drug releasing films with bio-based, non-toxic and biodegradable polymers that do not require chemical modifications. METHODS: Nanofibrillar cellulose and anionic type nanofibrillar cellulose were used as film forming materials with known mucoadhesive components mucin, pectin and chitosan as functional bioadhesion enhancers. Different polymer combinations were investigated to study the adhesiveness, solid state characteristics, film morphology, swelling, mechanical properties, drug release with the model compound metronidazole and in vitro cytotoxicity using TR146 cells to model buccal epithelium. RESULTS: SEM revealed lamellar structures within the films, which had a thickness ranging 40-240 µm depending on the film polymer composition. All bioadhesive components were non-toxic and showed high adhesiveness. Rapid drug release was observed, as 60-80% of the total amount of metronidazole was released in 30 min depending on the film formulation. CONCLUSIONS: The liquid molding used was a straightforward and simple method to produce drug releasing highly mucoadhesive films, which could be utilized in treating local oral diseases, such as periodontitis. All materials used were natural biodegradable polymers from renewable sources, which are generally regarded as safe.

Distinct prostate cancer-related mRNA cargo in extracellular vesicle subsets from prostate cell lines.

BACKGROUND: Multiple types of extracellular vesicles (EVs), including microvesicles (MVs) and exosomes (EXOs), are released by all cells constituting part of the cellular EV secretome. The bioactive cargo of EVs can be shuffled between cells and consists of lipids, metabolites, proteins, and nucleic acids, including multiple RNA species from non-coding RNAs to messenger RNAs (mRNAs). In this study, we hypothesized that the mRNA cargo of EVs could differ based on the EV cellular origin and subpopulation analyzed. METHODS: We isolated MVs and EXOs from PC-3 and LNCaP prostate cancer cells by differential centrifugation and compared them to EVs derived from the benign PNT2 prostate cells. The relative mRNA levels of 84 prostate cancer-related genes were investigated and validated using quantitative reverse transcription PCR arrays. RESULTS: Based on the mRNA abundance, MVs rather than EXOs were enriched in the analyzed transcripts, providing a snapshot of the tumor transcriptome. LNCaP MVs specifically contained significantly increased mRNA levels of NK3 Homeobox 1 (NKX3-1), transmembrane protease serine 2 (TMPRSS2), and tumor protein 53 (TP53) genes, whereas PC-3 MVs carried increased mRNA levels of several genes including, caveolin-2 (CAV2), glutathione S-transferase pi 1 (GSTP1), pescadillo ribosomal biogenesis factor 1 (PES1), calmodulin regulated spectrin associated protein 1 (CAMSAP1), zinc-finger protein 185 (ZNF185), and others compared to PNT2 MVs. Additionally, ETS variant 1 (ETV1) and fatty acid synthase (FASN) mRNAs identified in LNCaP- and PC-3- derived MVs highly correlated with prostate cancer progression. CONCLUSIONS: Our study provides new understandings of the variability of the mRNA cargo of MVs and EXOs from different cell lines despite same cancer origin, which is essential to better understand the the proportion of the cell transcriptome that can be detected within EVs and to evaluate their role in disease diagnosis.

Hepatic differentiation of human pluripotent stem cells on human liver progenitor HepaRG-derived acellular matrix.

Human hepatocytes are extensively needed in drug discovery and development. Stem cell-derived hepatocytes are expected to be an improved and continuous model of human liver to study drug candidates. Generation of endoderm-derived hepatocytes from human pluripotent stem cells (hPSCs), including human embryonic stem cells and induced pluripotent stem cells, is a complex, challenging process requiring specific signals from soluble factors and insoluble matrices at each developmental stage. In this study, we used human liver progenitor HepaRG-derived acellular matrix (ACM) as a hepatic progenitor-specific matrix to induce hepatic commitment of hPSC-derived definitive endoderm (DE) cells. The DE cells showed much better attachment to the HepaRG ACM than other matrices tested and then differentiated towards hepatic cells, which expressed hepatocyte-specific makers. We demonstrate that Matrigel overlay induced hepatocyte phenotype and inhibited biliary epithelial differentiation in two hPSC lines studied. In conclusion, our study demonstrates that the HepaRG ACM, a hepatic progenitor-specific matrix, plays an important role in the hepatic differentiation of hPSCs.

Real-Time Label-Free Monitoring of Nanoparticle Cell Uptake.

The surface plasmon resonance technique in combination with whole cell sensing is used for the first time for real-time label-free monitoring of nanoparticle cell uptake. The uptake kinetics of several types of nanoparticles relevant to drug delivery applications into HeLa cells is determined. The cell uptake of the nanoparticles is confirmed by confocal microscopy. The cell uptake of silica nanoparticles and polyethylenimine-plasmid DNA polyplexes is studied as a function of temperature, and the uptake energies are determined by Arrhenius plots. The phase transition temperature of the HeLa cell membrane is detected when monitoring cell uptake of silica nanoparticles at different temperatures. The HeLa cell uptake of the mesoporous silica nanoparticles is energy-independent at temperatures slightly higher than the phase transition temperature of the HeLa cell membrane, while the uptake of polyethylenimine-DNA polyplexes is energy-dependent and linear as a function of temperature with an activation energy of Ea = 62 ± 7 kJ mol-1 = 15 ± 2 kcal mol-1 . The HeLa cell uptake of red blood cell derived extracellular vesicles is also studied as a function of the extracellular vesicle concentration. The results show a concentration dependent behavior reaching a saturation level of the extracellular vesicle uptake by HeLa cells.

New p32/gC1qR Ligands for Targeted Tumor Drug Delivery.

Cell surface p32, the target of LyP-1 homing peptide, is upregulated in tumors and atherosclerotic plaques and has been widely used as a receptor for systemic delivery of payloads. Here, we identified an improved LyP-1-mimicking peptide (TT1, CKRGARSTC). We used this peptide in a fluorescence polarization-based high-throughput screening of a 50,000-compound chemical library and identified a panel of compounds that bind p32 with low micromolar affinity. Among the hits identified in the screen, two compounds were shown to specifically bind to p32 in multiple assays. One of these compounds was chosen for an in vivo study. Nanoparticles surface-functionalized with this compound specifically adhered to surfaces coated with recombinant p32 and, when injected intravenously, homed to p32-expressing breast tumors in mice. This compound provides a lead for the development of p32-targeted affinity ligands that circumvent some of the limitations of peptide-based probes in guided drug delivery.

Fluorescence-suppressed time-resolved Raman spectroscopy of pharmaceuticals using complementary metal-oxide semiconductor (CMOS) single-photon avalanche diode (SPAD) detector.

In this work, we utilize a short-wavelength, 532-nm picosecond pulsed laser coupled with a time-gated complementary metal-oxide semiconductor (CMOS) single-photon avalanche diode (SPAD) detector to acquire Raman spectra of several drugs of interest. With this approach, we are able to reveal previously unseen Raman features and suppress the fluorescence background of these drugs. Compared to traditional Raman setups, the present time-resolved technique has two major improvements. First, it is possible to overcome the strong fluorescence background that usually interferes with the much weaker Raman spectra. Second, using the high photon energy excitation light source, we are able to generate a stronger Raman signal compared to traditional instruments. In addition, observations in the time domain can be performed, thus enabling new capabilities in the field of Raman and fluorescence spectroscopy. With this system, we demonstrate for the first time the possibility of recording fluorescence-suppressed Raman spectra of solid, amorphous and crystalline, and non-photoluminescent and photoluminescent drugs such as caffeine, ranitidine hydrochloride, and indomethacin (amorphous and crystalline forms). The raw data acquired by utilizing only the picosecond pulsed laser and a CMOS SPAD detector could be used for identifying the compounds directly without any data processing. Moreover, to validate the accuracy of this time-resolved technique, we present density functional theory (DFT) calculations for a widely used gastric acid inhibitor, ranitidine hydrochloride. The obtained time-resolved Raman peaks were identified based on the calculations and existing literature. Raman spectra using non-time-resolved setups with continuous-wave 785- and 532-nm excitation lasers were used as reference data. Overall, this demonstration of time-resolved Raman and fluorescence measurements with a CMOS SPAD detector shows promise in diverse areas, including fundamental chemical research, the pharmaceutical setting, process analytical technology (PAT), and the life sciences.

Effect of differentiation on endocytic profiles of endothelial and epithelial cell culture models.

Understanding mechanisms of endocytosis and trafficking of nanoparticles through endothelial and epithelial barriers leads potentially to improved efficacy of nanoparticulate drug delivery systems. Detailed characterizations of cell models with respect to endocytic pathway expression and activity (endocytic profiling) should facilitate data interpretation. We performed endocytic profiling of CaCo-2 and hCMEC/D3 cell lines, widely used as human intestinal and blood-brain barrier permeability models, respectively, during cell differentiation. Furthermore, we compared endocytic profiles of cell lines with those of primary cells. Expression of genes involved in specific endocytic pathways was analyzed at mRNA levels by quantitative real time PCR. Where possible, the respective protein levels were analyzed by Western blotting, and endocytic activities of cells were analyzed by flow cytometry. We showed that differentiated CaCo-2 cells formed tight, well polarized monolayers with reduced endocytic activity accompanied by reduced mRNA expression of most of the endocytosis-related genes. In contrast, hCMEC/D3 cells formed a leaky, less polarized barrier, and in vitro differentiation had little effect on either the expression of endocytosis-related genes or endocytic activity of these cells. Endocytic profiling of in vitro models and comparison with primary cells is an important measure to avoid misleading conclusions in nanoparticle permeation studies.

Exploring the structure-activity relationships of ABCC2 modulators using a screening approach.

ABCC2 is a transporter with key influence on liver and kidney pharmacokinetics. In order to explore the structure-activity relationships of compounds that modulate ABCC2, and by doing so gain insights into drug-drug interactions, we screened a library of 432 compounds for modulators of radiolabeled ß-estradiol 17-(ß-d-glucuronide) (EG) and fluorescent 5(6)-carboxy-2',7'-dichlorofluorescein transport (CDCF) in membrane vesicles. Following the primary screen at 80µM, dose-response curves were used to investigate in detail 86 compounds, identifying 16 low µM inhibitors and providing data about the structure-activity relationships in four series containing 19, 24, 10, and eight analogues. Measurements with the CDCF probe were consistently more robust than for the EG probe. Only one compound was clearly probe-selective with a 50-fold difference in the IC50s obtained by the two assays. We built 24 classification models using the SVM and fused-XY Kohonen methods, revealing molecular descriptors related to number of rings, solubility and lipophilicity as important to distinguish inhibitors from inactive compounds. This study is to the best of our knowledge the first to provide details about structure-activity relationships in ABCC2 modulation.

Different gDNA content in the subpopulations of prostate cancer extracellular vesicles: apoptotic bodies, microvesicles, and exosomes.

BACKGROUND: Extracellular vesicles (EVs) are cell-derived membrane vesicles. EVs contain several RNAs such as mRNA, microRNAs, and ncRNAs, but less is known of their genomic DNA (gDNA) content. It is also unknown whether the DNA cargo is randomly sorted or if it is systematically packed into specific EV subpopulations. The aim of this study was to analyze whether different prostate cancer (PCa) cell-derived EV subpopulations (apoptotic bodies, microvesicles, and exosomes) carry different gDNA fragments. METHODS: EV subpopulations were isolated from three PCa cell lines (LNCaP, PC-3, and RC92a/hTERT) and the plasma of PCa patients and healthy donors, and characterized by transmission electron microscopy, nanoparticle tracking analysis and total protein content. gDNA fragments of different genes were detected by real time quantitative PCR and confirmed by DNA sequencing. RESULTS: We report that the concentration of EVs was higher in the cancer patients than in the healthy controls. EV subpopulations differed from each other in terms of total protein and DNA content. Analysis of gDNA fragments of MLH1, PTEN, and TP53 genes from the PCa cell-derived EV subpopulations showed that different EVs carried different gDNA content, which could even harbor specific mutations. Altogether, these results suggest that both nucleic acids and proteins are selectively and cell-dependently packed into the EV subtypes. CONCLUSIONS: EVs derived from PCa cell lines and human plasma samples contain double-stranded gDNA fragments which could be used to detect specific mutations, making EVs potential biomarkers for cancer diagnostics and prognostics.

Control of the morphology of lipid layers by substrate surface chemistry.

In this study, surface coatings were used to control the morphology of the deposited lipid layers during vesicle spreading, i.e., to control if liposomes self-assemble on a surface into a supported lipid bilayer or a supported vesicular layer. The influence of the properties of the surface coating on formation of the deposited lipid layer was studied with quartz crystal microbalance and two-wavelength multiparametric surface plasmon resonance techniques. Control of lipid self-assembly on the surface was achieved by two different types of soft substrate materials, i.e., dextran and thiolated polyethylene glycol, functionalized with hydrophobic linkers for capturing the lipid layer. The low-molecular-weight dextran-based surface promoted formation of supported lipid bilayers, while the thiolated polyethylene glycol-based surface promoted supported vesicular layer formation. A silicon dioxide surface was used as a reference surface in both measurement techniques. In addition to promoting supported lipid bilayer formation of known lipid mixtures, the dextran surface also promoted supported lipid bilayer formation of vesicles containing the cell membrane extract of human hepatoblastoma cells. The new dextran-based surface was also capable of protecting the supported lipid bilayer against dehydration when exposed to a constant flow of air. The well-established quartz crystal microbalance technique was effective in determining the morphology of the formed lipid layer, while the two-wavelength surface plasmon resonance analysis enabled further complementary characterization of the adsorbed supported lipid bilayers and supported vesicular layers.

Challenges of using in vitro data for modeling P-glycoprotein efflux in the blood-brain barrier.

The efficacy of central nervous system (CNS) drugs may be limited by their poor ability to cross the bloodbrain barrier (BBB). Transporters, such as p-glycoprotein, may affect the distribution of many drugs into the CNS in conjunction with the restricted paracellular pathway of the BBB. It is therefore important to gain information on unbound drug concentrations in the brain in drug development to ensure sufficient drug exposure from plasma at the target site in the CNS. In vitro methods are routinely used in drug development to study passive permeability and p-glycoprotein efflux of new drugs. This review discusses the challenges in the use of in vitro data as input parameters in physiologically based pharmacokinetic (PBPK) models of CNS drug disposition of p-glycoprotein substrates. Experience with quinidine demonstrates the variability in in vitro parameters of passive permeability and active pglycoprotein efflux. Further work is needed to generate parameter values that are independent of the model and assay. This is a prerequisite for reliable predictions of drug concentrations in the brain in vivo.

SVM classification and CoMSIA modeling of UGT1A6 interacting molecules.

The human UDP-glucuronosyltransferase 1A6 (UGT1A6) plays important roles in elimination of many xenobiotics, including drugs. We have experimentally assessed inhibitory properties of 46 compounds toward UGT1A6 catalyzing the glucuronidation of 1-naphthol and built models for predicting compounds interactions with the enzyme. The tested compounds were divided into a training set (n = 31; evaluated by 10-fold cross-validation) and an external test set (n = 15), both of which yielded similar accuracies (80-81%) and Matthews correlation coefficients (0.61-0.63) when classified using support vector machines. Comparative molecular similarity index analysis (CoMSIA) modeling was conducted for nine methods of compound alignment. The most predictive CoMSIA model was analyzed in the light of a homology modeled UGT1A6 structure, with leave-one-out cross-validation, yielding a q² of 0.62 and r² of 0.91 on the training set and a r²(pred) of 0.82 on the test set. The CoMSIA contour plots highlighted the importance of H-bond donors and electrostatic field interactions, accounting for 28% and 25% contribution of the model, respectively.