California Institute of Technology. Thanks to our hosts –especially Mel and Joyce – and to Wendy and Kim for tolerating our many, many idiosyncrasies.

California Institute of Technology

California Institute of Technology

Thanks to our hosts –especially Mel and Joyce – and to Wendy and Kim for tolerating our many, many idiosyncrasies.

Set the Stage• Goals of the AfCS: The Six Central Questions

• Accomplishments

People and capabilities

Data and analysis

Products

• Issues and Proposals

The AfCS cell and What are the issues?

What can we accomplish; what do we need to do it?

Central Questions of the AfCS: I

Question 1: How complex is signal processing in cells? The set of ligands for cellular receptors is the potential combinatorial code of inputs. How much of this input complexity can a cell uniquely decode as outputs?

Experiment: Systematic single- and double- (multi?) ligand screens. Classify output responses; determine degree of crosstalk; identify “hotspots” for later quantitative analysis.

New Technologies: Analytic methods to classify and compare multi-dimensional data for different ligand combinations

Central Questions of the AfCS: II

Question 2: What is the structure of the whole signaling network? Is the connectivity sparse or dense?

Experiment: Wholesale mapping of relevant protein-protein and small molecule-protein interactions.

New Technologies: High-throughput assays for intermolecular interactions in vivo, especially in response to ligand stimulation.

Central Questions of the AfCS: III

Question 3: How much does network topology constrain signal processing capability? How much function is specified by the nature of the connections, rather than by the specific biochemical constants of individual activities.

Experiment: Perturbation methods; gain and loss of function, coupled with functional assays.

New Technologies: Perturbations in vivo, singly and in combinations.

Central Questions of the AfCS: IV

Question 4: What are the dynamics of the signaling network? Can we visualize how information propagates through the network and emerges as functional activities?

Question 5: Can functional modules be abstracted mathematically? Can we make physical models and predict input-output relationships

Question 6: Why is the network the way it is? Why have the observed solutions been chosen? What is being optimized?

AfCS People and Capabilities• Refined set of ~35 participating investigators; Steering and System Committee(s); Lab Directors; Bridging Projects

• External Advisory Committee; Joan Brugge, Chair

• Talented set of 25 Ph.D.s, 40 technicians, & 8 programmers in 8 functional laboratories @ 6 sites

• Data analysis teams – work in progress

• Membership of 700 investigators, world-wide

• Effective/established collaborations: Myriad, CST, NPG

• Cell preparation, culture, single/multi ligand incubations

• Analytical capabilities: Ca2+, cyclic AMP, growing set of phosphoproteins (20), transcripts, glycerophospholipids, microscopy (location/translocation),

phosphoprotein/peptide purification, mass spectrometry (protein ID and phosphorylation sites), Y2H interactions, PCA’s

• Web site for display and dissemination of data and products

AfCS Accomplishments

• Data and Analysis

• Products for the Signaling Community

• Check out the 40+ Posters!

AfCS Accomplishments: Data & Analysis

• Single ligand screen in B cells

AfCS Accomplishments: Data & Analysis

• Single ligand screen in B cells

Collect very large data sets

Analyze and merge disparate data sets; cluster/profile ligands

Which ligands can/cannot be distinguished from others – single and double ligand data?

Where is the action? Where will we place our “infometers” to study topology and dynamics?

Gene list

The cycle time between experiment and analysis has been reduced dramatically.

AfCS Accomplishments: Data & Analysis

• Double ligand screen in B cells

AfCS Accomplishments: Data & Analysis

• Double ligand screen in B cells

How complex is the network? What is the density of interactions between inputs?

Do paired responses resolve tight clusters?

Where is the action? Where will we place our “infometers” to study topology and dynamics

AfCS Accomplishments: Data & Analysis

• Protein-Protein Interactions: Yeast Two-Hybrid Screens

AfCS Accomplishments: Data & Analysis

• Identify phosphoproteins and phosphorylation sites

AfCS Accomplishments: Data & Analysis

• Cellular location/translocation of signaling proteins

Products for the Signaling Community

• Signaling Gateway

Products for the Signaling Community

• Molecule page database

Products for the Signaling Community

• Antibody database

Products for the Signaling Community

• Covalent modification database

A cooperative project with the AfCS membership

Products for the Signaling Community

• Reagents: plasmids

Plasmid database

600 fully verified sequences

100 Gateway-ready expression vector backbones

1500 expression constructs

Distribution to ATCC is ongoing; distribution from ATCC this summer

Products for the Signaling Community

• Protocols

60 procedures; 200 ligands & other solutions

Have you ever been at a meeting where some unpleasant or difficult issue was known to everyone and was significantly impacting relationships or standing in the the way of resolving the problem at hand, but no one felt comfortable or even empowered to raise that subject?

Imagine there is a real live moose in the room. It’s huge. It may be mean-spirited. It’s very homely. It has a most unpleasant smell. It’s hard to imagine that everyone would carry on as if the moose simply did not exist. But often they do.

So Let’s Put the Moose On the Table

Moose Herders on the Steering Committee

Elliott Ross

Get that Moose’s Butt On the Damn Table

May I have Chocolate Mousse for

dessert please?

Different Styles

Experience with AfCS Cell PreparationsProcedure Splenic B Cell WEHI-231 Cardiomyocyte

Prep/Culture ++ Effort Simple ++++ Effort

Longevity Poor, but Bcl-2 Permanent 24 hr, but BDM

Cyclic AMP OK OK OK

Ca2+ OK Ligand set Hard/abnormal

Phosphoproteins OK OK OK

Transcripts Death response ? OK Poor

T’fect/T’duce No Low efficiency Adenovirus

RNAi No Emphatic no ?No

Microscopy No Less than ideal

Lipids OK

Other issues DNA/Prot Cnt’tile proteins

Proposal

• Change the focus of the AfCS to an experimentally tractable cell

The leading candidate is a transformed (v-Abl) mouse macrophage: RAW 264.7

The properties of these and other cells will be discussed tomorrow by Ron Taussig et al. Initial data will be shown.

What issues does this proposal raise? What are the implications for AfCS function?

Issues

• This is not what you proposed to do.

There are no changes in mission, purpose, intent, scope, or overall strategy – just a change of cell type

Issues

• What administrative and functional changes will be required?

• Very few, because of the lack of change of purpose and the portability/extensibility of our operations

Two System Committees will become one: Bourne, Stull, Berridge, Cantley, Devreotes, Gold, and cell-type experts

Some B cell- and myocyte-centric individuals will seek greener pastures

One laboratory will be re-focused

Issues

• You said you would work with a primary (normal) cell. Is this not a substantial change?

What’s In a Cell?

• There is much to be said for “native” function and complexity, and we tried, but…..

• We are not trying to detail and understand the properties of a specific “normal” cell. We are trying to discover the principles that underlie signaling networks.

• How abnormal can a cell be if it performs many regulated, specialized functions? The repertoire of RAW264.7 includes pinocytosis, phagocytosis, chemotaxis, synthesis and secretion of a hoard of inflammatory mediators, and differentiation.

• We can approach the normal counterpart of RAW by studying bone-marrow macrophages.

• RAW was transformed with v-Abl, which raises the possibility of manipulation with STI-571 (Gleevec). The devil you know….

Issues

• You wasted a lot of time and money. Enough.

There is obviously some truth to this statement, but it is a relatively modest fraction of the total spent to date. We have assembled a rather powerful experimental and analytical machine, and it is fully transferable to another system. Virtually none of the time and cost of assembly, training, and learning and acquisition of reagents, software systems, and databases has been wasted.

Is the AfCS Really Just a Big RO1?

• Certainly not in the details:

35 Collaborating participating investigators; multi-disciplinary

Research in 8 labs distinct from those of the participating investigators

Strategy, data, analysis, protcols, reagents disseminated promptly and publicly

• More importantly, the AfCS is clearly focused on important, broad, overarching questions that cannot be answered by individual laboratories: The Glue Grant Mission

The Glue Grant Mission

“The goal of this program is to enable the solution of major problems in biomedical research and to facilitate the next evolutionary stage of integrative biomedical science. The intention is to make resources available for … scientists to form research teams to solve a complex biological problem that is of central importance to biomedical science… and that would be beyond the means of any one research group.”

Is the AfCS Really Just a Big RO1?

• What does this question really mean? Apparently some feel that a project this big (i.e., expensive) should be at least relatively risk-free.

• Will the answers to complex, overarching questions come without risk? And we share a very substantial fraction of this risk because we are dedicated to the questions.

• We must be RO1-like in that we need the freedom to respond to experimental observations and technological innovation. These are the principles that have made NIH great.

What We Can Do in the Immediate Future (16 months) with a Tractable Cell Line

• Single and double ligand screen: ongoing display and analysis of data. Cyclic AMP, Ca2+, inositol phosphates (?), phosphoproteins (20), transcripts (chips and qrtPCR), lipids, function (phagocytosis, secretion, chemotaxis)

• Detailed parts list

• Location and translocation: FPM and kinase molecule sets with ligand pools

• Protein phosphorylation: FPM and kinase molecule sets with ligand pools

• Protein-protein interactions: Y2H, pulldowns, PCA

• Perturbations: RNAi and dominant negatives, FPM

Short-Term Needs• Parts List Database

Contents (Fruman)

Experimental execution (Simon)

Databasing (Subramaniam)

• Automated analysis of multi-ligand screen data (Ranganathan)

• Rational approach to test >2-way interactions (Ranganathan)

• Protein-protein interactions

Yeast two-hybrid (Fraser)

Pull downs (Sternweis/M. Mumby)

Protein complementation assays (PCA’s and dynamics?) (Michnick)

Short-Term Needs (Cont.)

• Extraction of pathway information from experimental data and the literature

• Complete author interfaces, displays, and query interface for the Molecule Page Database

What We Can Do in the Immediate Future (16 months) with a Tractable Cell Line

• Single and double ligand screen: ongoing display and analysis of data. Cyclic AMP, Ca2+, inositol phosphates (?), phosphoproteins (20), transcripts (chips and qrtPCR), lipids, function (phagocytosis, pinocytosis, chemotaxis)

• Detailed parts list

• Location and translocation: FPM and kinase molecule sets with ligand pools

• Protein phosphorylation: FPM and kinase molecule sets with ligand pools

• Protein-protein interactions: Y2H, pulldowns, PCA

• Perturbations: RNAi and dominant negatives, FPM