The 2nd Annual BME Symposium in Quantitative Biology and Physiology
Friday, October 21, 2005
Life Science and Engineering Building (24 Cummington Street) Rm. B01
Program
- 10:15 am – Breakfast
- 10:40 am – Ken Lutchen (BME Chair)
Opening Remarks - 10:50 am – Andrew Kreuger
Probing Biological Networks using Chemical Combinations - 11:10 am – Jane Lin
Characterizing Bidirectional Promoters in the Human Genome - 11:30 am – Kyle Lillis
Postsynaptic Involvement in Short-Term Plasticity - 11:50 am – Corin Williams
Quantifying the Contractile and Synthetic States of Vascular Smooth Muscle Cells - 12:10 pm – Hemali Patel
Transcriptional Profile of E. Coli in Stringent Response - 12:30 pm – Julien Penders
A Metabolic Model for Shewanella Oneidensis: From Gene Annotation to Flux Balance - 12:50 pm – Derek Affonce
New Perspectives: Mechanical Basis for Airway Hyperreactivity - 1:10 pm – Lunch
- 2:00 pm – Keynote – Dr. Raimond Winslow, Johns Hopkins University
Multi-Scale Modeling of Cardiac Electrical Function - 3:00 pm – Dan Freeman
Light and Contrast Sensitive Gain Controls in the Retina - 3:20 pm – Jie Wu
Extracting Higher Order Information from Large-Scale Protein Networks - 3:40 pm – Wynter Duncanson
Development of a Method to Quantify Ultrasound Contrast Agent Performance of Individual Microparticles
Abstracts
10:50 am – Andrew Kreuger
Probing Biological Networks using Chemical Combinations
Molecular insights into human disease are rapidly emerging. Understanding how molecules interact in normal and abnormal states will lead to sophisticated treatments. The distribution and dynamics of these interactions can be described with network models of the biological system. Network models are usually constructed from connectivity data (eg, protein-protein binding or protein-gene binding) and refined using perturbation constraints (eg, responses to environmental or genetic modifications). However a large set of diverse constraints is necessary to explain thousands of intracellular components and conditions.
Here we present a method that produces a unique set of data by applying combination chemical probes to a biological system. Both chemicals are administered in dilution steps, providing a detailed phenotypic response. Because chemical combinations act at the protein level, they can provide constraints which are more immediate to many cellular functions.
We investigated the utility of chemical combinations through numerical simulations of metabolic pathways in conjunction with a small yeast screen of combined antifungal treatments targeting the sterol and other pathways. The simulations produced distinct response surface shapes for differing target configurations. The small yeast screen confirmed theoretical predictions for the known topology and regulation of the sterol pathway in yeast. This work confirms that chemical combinations provide sensitive new constraints on the existence and nature of functional connections between targets, which can be used for systems biology and chemical genetics applications.
We are encouraged by our initial results, and will scale up both the experimental and theoretical aspects of the project. A much larger experimental screen, with a more thorough sampling of network topology, is in progress. We will compare these results to larger scale models, such as flux balance models.
11:10 am – Jane Lin
Characterizing Bidirectional Promoters in the Human Genome
Divergent gene pairs are less than 1000bp apart oriented in opposite directions and have been shown to make up as much as 10% of the genes in the human genome. They are centered around a shared bidirectional promoter, which has the unique ability to drive transcription of both divergent genes. We are interested in the features of the bidirectional promoter that set it apart from its unidirectional counterpart. In particular, we discover that the sequence motif for GABP (aka Nrf-2) is over-represented in bidirectional promoters. An incredible 70% of binding sites predicted by our GABP PWM (position weight matrix) are bound in vivo, verified using chromatin immuno-precipitation assay. Further analysis of the in vivio binding sites also revealed a strong co-occurence with the SP-1 sequence motif, also found to be over-represented in bidirectional promoters. Here, we propose five transcription factors (GABP, SP-1, Nrf-1, NF-Y, and STAT1) as key regulators of the bidirectional promoter based on sequence analysis, and further validate in vivo binding of one (GABP) in particular.
11:30 am – Kyle Lillis
Postsynaptic Involvement in Short-Term Plasticity
Connections between neurons (synapses) in the central nervous system are extremely flexible. Synaptic plasticity, the capacity to dynamically change the strength of synapses, is widely believed to underlie learning, memory, and complex signal processing. In this work, we describe a novel form of short term plasticity (lasting <1min), which we call rate dependent efficacy (RDE). While most forms of short-term plasticity depend solely on presynaptic mechanisms (e.g. neurotransmitter vesicle depletion), RDE requires both pre- and post-synaptic activity. When pre-synaptic action potentials (APs) are paired with post-synaptic APs at a rate greater than 5Hz, synaptic strength increases by ~20% over the change elicited by pre-synaptic stimulation alone. Post-synaptic action potentials alone were not sufficient to change synaptic strength. In the presence of glycine receptor antagonists strychnine or picrotoxin, the boosting effect of pairing on synaptic strength is occluded, indicating that glycine is involved in RDE. Furthermore, amoxapine, which blocks glycine transport to increase the concentration of ambient glycine, also eliminates RDE. These data reveal a phenomenon by which pairing of pre- and postsynaptic action potentials induces a short-lived change in synaptic strength. Pharmacological experiments suggest that pairing acts through some unknown mechanism to reduce glycinergic activity and express RDE.
11:50 am – Corin Williams
Quantifying the Contractile and Synthetic States of Vascular Smooth Muscle Cells
Controlling vascular smooth muscle cell (VSMC) behavior has been a major challenge to the development of functional tissue-based vascular grafts. In standard culture conditions, VSMCs lose their native organization and contractile function as they switch to a proliferative phenotype; however, micropatterned substrata can be used to control cell organization, morphology, and function. Although there is evidence suggesting that cultured VSMCs can revert to the contractile state, the role of microtopographical cues on VSMC functionality has not been thoroughly investigated.
Recently, we found that VSMCs constrained by cell-resistant comb polymer barriers regained morphological features that are reminiscent of native contractile VSMCs. Patterned VSMCs were more elongated and had smaller projected areas as lane width decreased, while unpatterned cells were large and highly spread. Currently, we are further quantifying the contractile and synthetic states by investigating unpatterned and patterned VSMC contractile responses. We will determine forces generated by VSMCs in response to KCl using traction force microscopy. Preliminary results suggest that micropatterned VSMCs display organized contractile responses due to the physical constraints of micropatterned lanes while unpatterned VSMCs have weak and disorganized responses. We expect that the mechanisms of response may differ between micropatterned and unpatterned VSMCs.
12:10 pm – Hemali Patel
Transcriptional Profile of E. Coli in Stringent Response
Stringent response is a complex regulatory pathway in bacteria that is triggered by nutrient limitation and involves a global reprogramming of gene transcription. During stringent response, cell growth is arrested and many major pathways are affected, including: stress responses, nucleotide and macromolecule biosynthesis, DNA-damage responses, and cell protection. The induction of stringent response has also been correlated with increased persistence whereby cells develop tolerance to antibiotics. Transcriptional changes seen during stringent response are believed to be a result of direct modulation of RNA polymerase by an alarmone molecule, ppGpp. However, given that half the genome shows transcriptional changes, we propose that modulation of transcription is also mediated by transcription factors (TFs). Using high-density, whole-genome, oligonucleotide microarrays, we studied gene expression of 4345 Escherichia coli (E. coli) genes before and after induction of stringent response by amino acid starvation. After normalization, we found 2369 genes that show statistically significant (False Discovery Rate < 0.66%) changes in expression: 1101 genes involved in cell protection, promotion of biofilm formation and global regulatory functions were upregulated; 1268 genes involved in energy-consuming pathways including aerobic metabolism, macromolecule biosynthesis, active transport, and cell envelope synthesis were downregulated. The involvement of 75% of currently known TFs in this response suggests that TFs do indeed play a significant role in orchestrating far-reaching transcriptional changes during stringent response.
12:30 pm – Julien Penders
A Metabolic Model for Shewanella Oneidensis: From Gene Annotation to Flux Balance
Shewanella oneidensis exhibits remarkable and uncommon anaerobic properties and is known to have applications in both bioremediation and fuel cell technology. Engineering Shewanella for these applications begins with the development of a model for its metabolism. The recent genome annotation establishes the basis for such a project. In this context, the first step is to build a metabolic map from gene annotations. A primary level of mapping is automatically generated using the PathoLogic (Pathway Tools) software. Manual curation and homology analysis are then performed to further improve the metabolic mapping. The final metabolic map contains 1015 enzymatic reactions.
Going from metabolic map to Flux Balance modeling is the second step of the process. The relative decoupling of catabolism and anabolism suggests that we model these two parts of the metabolism separately. Focusing on the catabolic part, we start from a core-metabolism model including 64 reactions. Thermodynamics of biochemical reactions is used to attribute a rigorous directionality to the reactions included in the model. Up to now, a second-level (accounting for regulated pH and ionic strength in cells) calculation of the free energies is included in the model but we hope that experimental data could bring the model to a third-level, including the effect of important cofactors concentrations (ATP, NAD(P)H).
Preliminary results regarding growth on pyruvate and lactate seem promising. Once available, information about biomass composition, cofactor concentrations, and transport reactions will easily be added to the model. Future work will focus on both an extension of the model to include degradation pathways and an experimental validation of model predictions at a qualitative and then quantitative level.
12:50 pm – Derek Affonce
New Perspectives: Mechanical Basis for Airway Hyperreactivity
The asthmatic airway wall is characterized by thickening of all three of its layers. Of particular importance is thickening of the airway smooth muscle (ASM) layer. There has been no data to suggest whether or not the thickened ASM in asthmatics results from increases in the amount of contractile proteins and hence an increased ability to generate force. A model was developed to determine what effect thickening of the airway wall has on airway narrowing (JAP 1993:2771). However in that study it was assumed that the thicker ASM in asthmatic airways was inherently stronger, and also that the thicker airway wall was not stiffer than the airway wall in healthy subjects. The goal of this project was to modify the model set forth by Lambert et. al. to determine whether the ASM of the asthmatic has to be able to generate greater forces to create constrictions seen in vivo and what effect stiffening the airway wall has on luminal narrowing in the asthmatic and non-asthmatic airway. The key feature of the model developed by Lambert et. al. is that it defined the tension ASM must overcome to constrict. These loads are the result of the transpulmonary pressure, parenchymal tethering, and passive pressure-area relationships of the airway. The model also limits the maximal tension the ASM can generate to physiologically realistic values. To modify this model the slope of the pressure-area relationship was modified to increase wall stiffness. Secondly when calculating the stress that the ASM must generate the thickness of the ASM was set to be the thickness of the healthy ASM, this was done to probe the effect thickening of the ASM layer would have if the thickening did not result in more contractile protein. Through this modeling study it was determined for luminal constrictions seen in asthmatics in vivo that ASM in asthmatic airways has to generate more force than the ASM of a healthy person. Secondly we found stiffening the airway wall, promotes airway luminal diameter reduction at baseline, but inhibits airway luminal diameter reduction after stimulation.
2:00 pm – Keynote – Dr. Raimond Winslow, Johns Hopkins University
Multi-Scale Modeling of Cardiac Electrical Function
There is a growing recognition that the emergent, integrative behaviors of biological systems are a result of complex interactions between all the components from which these systems are composed and that knowledge of each system component, however detailed, is not sufficient by itself to understand overall system behavior. Achieving an integrative understanding of molecules, cells, tissues and organs is indeed the next major frontier of biomedical science. Because of the inherent complexity of real biological systems, the development and analysis of computational models based directly on experimental data is necessary to achieve this understanding. To illustrate these ideas, in this talk I will describe how understanding of biological processes which take place over enormously different spatio-temporal scales is crucial to the understanding of global heart function in both health and disease.
3:00 pm – Dan Freeman
Light and Contrast Sensitive Gain Controls in the Retina
The visual system faces the challenge of encoding a stimulus that varies over 10 orders of magnitude in the limited dynamic range of the spiking output of neurons that make up the optic nerve. This limitation is dealt with through adaptation, which alters the sensitivity of the retina to the current range of stimulus intensities in the visual scene. There are two types of adaptive mechanisms; one is sensitive to light intensity and the other to deviations of light from the mean intensity, referred to as contrast. Classical light adaptation is the primary gain control of the retina, acting to reduce the sensitivity of the retina as the ambient light intensity is increased. There are also multiple forms of contrast sensitive adaptation, but these are not as well understood in terms where they act in the retina and on what timescales they respond. Our research is aimed deciphering these contrast adaptation mechanisms in terms of their spatial and temporal properties in order to better understand the physiology underlying the adaptation. Also, we are interested in how light and contrast adaptation are related and weather or not competing light and dark adaptation could be confused for contrast adaptation.
3:20 pm – Jie Wu
Extracting Higher Order Information from Large-Scale Protein Networks
The increasing availability of fully sequenced genomes from diverse species has provided new possibilities of understanding gene functions at a system level. High throughput techniques such as those that explore genomic context or analyze the genome wide gene expression data have been developed to identify protein linkage networks and consequently to shed light on the functionality of genes and the organization of the cell. While pair-wise linkage is useful in constructing networks, it misses correlations in higher order groups: triplets, quadruplets, and so on. Moreover, such simple binary representation cannot adequately describe the complexity of cellular networks that involve branching, parallel and alternate pathways.
Here we describe a statistical framework to extract higher order information from large-scale protein networks. Specifically, we assess the probability of observing co-occurrence patterns of 3 binary profiles by chance and show that this probability is asymptotically the same as the multi information in the three profiles. Using phylogenetic profiling data, we demonstrate the utility of our metrics in detecting overly represented triplets of orthologous proteins which could not be detected using pair-wise profiles. These triplets serve as small building blocks, i.e. motifs in protein networks; they allow us to infer the function of uncharacterized members, and facilitate analysis of the local structure and global organization of the protein network. Our method is extendable to N-component clusters, and therefore serves as a general tool for high order information extraction in protein networks. We envision the general applicability of the method to the analysis of higher order organization of protein networks and the potential discovery of interesting network motifs that elucidate the functional organization of the cell.
3:40 pm – Wynter Duncanson
Development of a Method to Quantify Ultrasound Contrast Agent Performance of Individual Microparticles
Ultrasound contrast agents are currently used extensively for perfusion imaging. In addition, site-targeted acoustic contrast agents have enormous potential to aid in the early detection of diseases of the vasculature. One approach to quantify the acoustic properties of microbubbles is to directly image individual microbubbles, which have been shown to be independent scatterers of sound waves. The quantitative acoustic response of individual microbubbles can then be used to validate theoretical models that predict acoustic response. We developed and evaluated a system to probe the individual voltage waveforms of particles immobilized in polyacrylamide gels with a scanning acoustic microscope with a high frequency transducer. We used solid micron-sized polystyrene (PS) spheres to validate the system. The PS spheres are also fluorescently-labeled to confirm particle positions via fluorescence microscopy. Our results show that our system can be used to detect, image and quantify the acoustic behavior individual microbubbles.
