Identification of pharmacodynamic biomarker hypotheses through literature analysis with IBM Watson.

BACKGROUND: Pharmacodynamic biomarkers are becoming increasingly valuable for assessing drug activity and target modulation in clinical trials. However, identifying quality biomarkers is challenging due to the increasing volume and heterogeneity of relevant data describing the biological networks that underlie disease mechanisms. A biological pathway network typically includes entities (e.g. genes, proteins and chemicals/drugs) as well as the relationships between these and is typically curated or mined from structured databases and textual co-occurrence data. We propose a hybrid Natural Language Processing and directed relationships-based network analysis approach using IBM Watson for Drug Discovery to rank all human genes and identify potential candidate biomarkers, requiring only an initial determination of a specific target-disease relationship. METHODS: Through natural language processing of scientific literature, Watson for Drug Discovery creates a network of semantic relationships between biological concepts such as genes, drugs, and diseases. Using Bruton's tyrosine kinase as a case study, Watson for Drug Discovery's automatically extracted relationship network was compared with a prominent manually curated physical interaction network. Additionally, potential biomarkers for Bruton's tyrosine kinase inhibition were predicted using a matrix factorization approach and subsequently compared with expert-generated biomarkers. RESULTS: Watson's natural language processing generated a relationship network matching 55 (86%) genes upstream of BTK and 98 (95%) genes downstream of Bruton's tyrosine kinase in a prominent manually curated physical interaction network. Matrix factorization analysis predicted 11 of 13 genes identified by Merck subject matter experts in the top 20% of Watson for Drug Discovery's 13,595 ranked genes, with 7 in the top 5%. CONCLUSION: Taken together, these results suggest that Watson for Drug Discovery's automatic relationship network identifies the majority of upstream and downstream genes in biological pathway networks and can be used to help with the identification and prioritization of pharmacodynamic biomarker evaluation, accelerating the early phases of disease hypothesis generation.

Photodynamic effects of pterin on HeLa cells.

Pterins, heterocyclic compounds widespread in biological systems, participate in relevant biological processes and are able to act as photosensitizers. In the present study, we ascertained that 2-aminopteridin-4(3H)-one, abbreviated as Ptr, is readily incorporated into and/or onto cervical cancer cells (HeLa) and that these cells die upon UV-A irradiation of Ptr. Cell death was assessed using two tests: (1) the Rhodamine 123 fluorescence assay for mitochondrial viability and (2) the Trypan Blue assay for membrane integrity. The data suggest that, for Ptr-dependent photoinitiated cell death, events related to mitochondrial failure precede those associated with the failure of the cell membrane.

The photosensitizing activity of lumazine using 2'-deoxyguanosine 5'-monophosphate and HeLa cells as targets.

Lumazines are an important family of heterocyclic compounds present in biological systems as biosynthetic precursors and/or products of metabolic degradation. Upon UV irradiation, the specific compound called lumazine (pteridine-2,4(1,3H)-dione) is able to generate singlet oxygen (1O2), which is one of the main chemical species responsible for photodynamic effects. To further assess the photosensitizing capability of lumazine (Lum) experiments were performed using the nucleotide 2'-deoxyguanosine 5'-monophosphate (dGMP) and, independently, cervical cancer cells (HeLa cell line) as targets. In the dGMP experiments, the data revealed that dGMP indeed undergoes oxidation/oxygenation photoinduced by Lum. Moreover, dGMP disappearance proceeds through two competing pathways: (1) electron transfer between dGMP and excited-state Lum (Type I process) and (2) reaction of dGMP with 1O2 produced by Lum (Type II process). The multistep processes involved are convoluted and susceptible to changes in experimental conditions. The independent studies with HeLa cells included fluorescence analysis of cell extracts and phototoxicity experiments performed at the single-cell level. Results showed that, upon Lum uptake and irradiation, photodynamic effects occur. In particular, the mitochondria and cell membrane were perturbed, both of which reflect key stages in cell death. The data reported herein illustrate how the irradiation of an endogenous biological compound can have various effects which, depending on the system, can be manifested in different ways.

Photosensitized production of singlet oxygen: spatially-resolved optical studies in single cells.

Singlet molecular oxygen, O(2)(a(1)Delta(g)), can be created in photosensitized experiments with sub-cellular spatial resolution in a single cell. This cytotoxic species can subsequently be detected by its 1270 nm phosphorescence (a(1)Delta(g)--> X(3)Sigma). Cellular responses to the creation of singlet oxygen can be monitored using viability assays. Time- and spatially-resolved optical measurements of both singlet oxygen and its precursor, the excited state sensitizer, reflect the complex and dynamic morphology of the cell. These experiments help elucidate photoinduced, oxygen-dependent events that compromise cell function and ultimately lead to cell death. In this perspective, recent work on the photosensitized production and detection of singlet oxygen in single cells is summarized, highlighting the advantages and current limitations of this unique experimental approach to study an old problem.

Time-resolved singlet oxygen phosphorescence measurements from photosensitized experiments in single cells: effects of oxygen diffusion and oxygen concentration.

Time-resolved singlet oxygen, O(2)(a(1)Delta(g)), phosphorescence experiments have been performed in single cells upon pulsed laser irradiation of a photosensitizer incorporated into the cell. Data recorded as a function of the partial pressure of ambient oxygen to which the cell is exposed reflect apparent values for the intracellular oxygen diffusion coefficient and intracellular oxygen concentration that are smaller than those found in neat H(2)O. This conclusion is supported by O(2)(a(1)Delta(g)) phosphorescence data and sensitizer triplet state absorption data recorded in control experiments on sucrose solutions with different viscosities. We recently demonstrated that the intracellular lifetime of O(2)(a(1)Delta(g)) is comparatively long ( approximately 3 micros) and does not differ significantly from that in neat H(2)O ( approximately 3.5 micros). Despite this long lifetime, however, our estimate of an apparent intracellular oxygen diffusion coefficient in the range approximately 2-4 x 10(-6) cm(2) s(-1) means that the spatial domain of intracellular O(2)(a(1)Delta(g)) activity will likely have a spherical radius of approximately 100 nm. This latter point helps reconcile seeming inconsistencies between our direct O(2)(a(1)Delta(g)) lifetime data and results obtained from independent photobleaching experiments that show a limited translational diffusion distance for O(2)(a(1)Delta(g)) within a cell.

Measuring the lifetime of singlet oxygen in a single cell: addressing the issue of cell viability.

Singlet molecular oxygen, O(2)(a(1)Delta(g)), has been detected from single neurons and HeLa cells in time-resolved optical experiments by its 1270 nm phosphorescence (a(1)Delta(g)--> X(3)Sigma(-)(g)) upon irradiation of a photosensitizer incorporated into the cell. The cells were maintained in a buffered medium and their viability was assessed by live/dead assays. To facilitate the detection of singlet oxygen, intracellular H(2)O was replaced with D(2)O by an osmotic de- and rehydration process. The effect of this insult on the cells was likewise assessed. The data indicate that, in the complicated transition from a "live" to "dead" cell, the majority of our cells have the metabolic activity and morphology characteristic of a live cell. Quenching experiments demonstrate that the singlet oxygen lifetime in our cells is principally determined by interactions with intracellular water and not by interactions with other cell constituents. The data indicate that in a viable, metabolically-functioning, and H(2)O-containing cell, the lifetime of singlet oxygen is approximately 3 micros. This is consistent with our previous reports, and confirms that the singlet oxygen lifetime in a cell is much longer than hitherto believed. This implies that, in a cell, singlet oxygen is best characterized as a selective rather than reactive intermediate. This is important when considering roles played by singlet oxygen as a signaling agent and as a component in events that result in cell death. The data reported herein also demonstrate that spatially-resolved optical probes can be used to monitor selected events in the light-induced, singlet-oxygen-mediated death of a single cell.

Listening to the brain: microelectrode biosensors for neurochemicals.

Chemical signalling underlies every function of the nervous system, from those of which we are unaware, for example, control of the heart, to higher cognitive functions, such as emotions, learning and memory. Neurotransmitters and neuromodulators mediate communication between neurons and between neurons and non-neural cells such as glia and muscle. In the past, the means for studying the production and release of these signalling agents directly has been limited in its temporal and spatial resolution relative to the dynamics of chemical signalling and the structures of interest in the brain. Now microelectrode biosensors are becoming available that give unprecedented spatial and temporal resolution, enabling, for the first time, direct measurement in real time of the chemical conversations between cells in the nervous system.

Microelectrode biosensor for real-time measurement of ATP in biological tissue.

The purines ATP, ADP, and adenosine are important extracellular signaling agents. Analysis of purinergic signaling has been slowed by lack of direct methods for measurement of purine release in real-time during physiological activity. We have previously reported microelectrode biosensors for adenosine, but similar sensors for ATP have remained elusive. We now describe an ATP biosensor formed by coating a Pt microelectrode with an ultrathin biolayer containing glycerol kinase and glycerol-3-phosphate oxidase. It responds rapidly (10-90% rise time <10 s) and exhibits a linear response to ATP over the physiologically relevant concentrations of 200 nM-50 microM and is very sensitive approximately 250 mA.M(-1).cm(-2). By including phosphocreatine kinase in the biolayer, we can optionally amplify the ATP signal and also make the sensor sensitive to external ADP. We have used our sensors to make the first demonstration that ATP is released from spinal networks in vivo during locomotor activity.