Nicole King, University of California, Berkeley, The origin of animal multicellularity
- Many of the genes for cell adhesion and cell-cell signaling evolved before the transition to multicellularity.
- Originally involved in capturing prey? Now in holding cells together?
- Colony formation: single cell slow swimmer can become rosette colony, chain colony, fast swimmer (reduces collar size) isolated attached cell.
- colony formation through division. What controls this developmental switch?
- one of 64 isolated bacteria induces rosette formation: Algoriphagus machipongonensis. Want to isolate the molecule that induces colony formation.
- component of cell envelope – sphingolipid from outer membrane. 1.5% of sphingolipid – capnine.
Evgeni Nikolaev, Rutgers, The State University of New Jersey, Mathematical Modeling Describes a Novel Bistable Mechanism of Lipid Body Formation in Macrophages Infected with M. tuberculosis
- (Sontag lab). TB research.
- untreated death rate is 50%. 10% lifetime chance that latent infection will become active.
- Goal: top down model to formalize existing literature and make predictions
- Lipid biology in TB. cholesterol and LDL receptors affect TB outcomes. related to (Nobel Prize in Medicine)
- free cholesterol inhibits cholesterol synthesis. Bad cholesterol accumulates in golgi causes cell lysis.
- lungs high oxygen low free radicals.
- Bistable models somehow. Didn’t see axes. Didn’t get intuition on the feedback.
- TB infected macrophages turn on extra biosynthesis of cholesterol which reduces macrophage viability.
- reducing biosynthesis of cholesterol should reduce lipid body formation.
Rebecca McLennan, Stowers Institute for Medical Research, Elucidating multiscale mechanisms of neural crest migration
- (Kulesa Lab) . Looking at migration from chick R4 into brackeal arch.
- VEGF is a chemo-atractant signal. Don’t migrate if remove receptor.
- in the absence of a gradient how do you migrate (looks to flat-line fast?) How do you know there’s no gradient? mRNA expression? diffusion?
- Looks alright, the modelers put the gradient back in. Cells consume the VEGF as they go.
- problem: later emerging cells don’t have a gradient left to follow. Not follow experimental observation.
- Add trailing cells that follow lead cells. previously observe morphological differences between lead and trailers. trailers more polarized, less proliferated.
- oblate neural crest, ‘lead cells’ spread out, ones in front still look like leaders ones in back look like trailers. Transplant trailing cells + microenvironment to front with lower chemoatractant. model: lead cells stop moving.
- Experiment: trailing cells migrate into arch like lead cells and previous lead stay back and follow. Sounds like cell-cell communication and diffusion.
- Do cells migrate around transplant ? get around. In model, when you put trailers in the front, why do they switch?
- Do trailers depend on cell contact or a chemical gradient?
- How do you specify trailers in the model?
Wallace Marshall, University of California, San Francisco, Biological Noise at the level of Cellular Structure
- Flagellar length control system.
- size relation to fitness. is there some indicator of control? behavior?
- extra length flops against the breast stroke and slows you down. To short just moves back and forth can’t go forward — no assymetric bend. One arm too long go in circles.
- Measure speed vs. length. wildtype flagella lie in high speed sweet spot.
- balanced addition and removal at the tip at equilibrium length.
- Kinesien brings train out very processive no pausing, dynine brings it back.
- Disassembly rate is independent of rate. assembly is rate limited by IFT (trains).
- Time to deliver return grab new cargo scales 1/L?
- Total cargo is independent of length (actually get more trains with less cargo each).
- Better way to test the model looking at measurements of noise.
- Look at variation in L1 vs L2 (2 flagella lengths). Length vs diameter mostly linear, saturates at high diameter.
- Restoring “force” depends on length set-point — should grow back faster and slow. slower scale variation in long armed mutants.
- mutants that increase noise should increase variation. Not all long mutant genes have same effect on noise — both affect variation of on diaganol lengths are correlated.
Stas Shvartsman, Princeton University, Multiscale studies of embryonic development: experiments and theory
- Re-titled: reaction kinetics… (in embryos)
- Networks organized around hubs. One regulator, multiple targets.
- substrates compete for MAPK (Curr Bio) compete with MAPK regulators (Mol Sys Bio), control gene dev (Dev Cell).
- Asymmetric down-regulation in head and posterior. Bcd is a competitive inhibitor of MAPK. So is Hb.
- Remove uniform Gro (also a MAPK target), get uniform increase.
- pt2: change capicua levels — also has an effect. Cic has to dock first, substrates compete with regulators. Adding a substrate out-competes the phosphotase.
- Functional significance. Dorsal patterning system DV axis. Zen expressed in dorsal region and posterior pole. MAPK antagonizes Dorsal mediated repression. Competitive inhibition of MAPK prevents MAPK from repressing dorsal repression of zen.
Stephen Helms, University of Texas Southwestern Medical Center, A Tandem PDZ Domain Redox Switch in InaD
- Organization by scaffold proteins in signaling.
- Drosophila photodetection. InaD scaffold complexes with 60 different proteins
- Dynamic scaffold — only reduced form of InaD free to bind its ligand PDz5, oxidized species won’t.
- Lock in reduced form: cells loose refractory period of photo response, can’t as well in high light.
- InaD is mediating negative feedback.
- Pdz5 multiple types in fast flies, not present in slow flies.
- isolated protein in different biochemical state than in vivo protein (reduced).
- Binding partner Pdz4 shifts redox potentional curve to be more reduced at physiological conditions.
- unbound, oxidized, closed. Bound -> reduced -> open. Need to couple these processes to make a good switch.
- Want to know kinetic rates between coupled pairs converting between confirmation and oxidation state.
- Fit to get allosteric coupling coefficient (1/270).
- Active state pDZ5 pomotes NorpA. Repressive state, NorrpA released from scaffold, lowers its activity. system is slow.
Domitilla Del Vecchio, Massachusetts Institute of Technology, The Impact of Retroactivity on the Signal Processing of Transduction Networks
- Modularity: behavior of a system does not change when you integrate a module into a network. Oops.
- What’s the copy number of your load (Single binding sites or multiple copy plasmids of each). Do you have a different effect if you incorporate these on single copy BACs or bacterial genome?
- retroactivity (like nuclear trapping story).
- Rather than connect clock directly, connect clock through a insulation device.
- Large input amplification and large negative feedback minimize retro-activity.
- Consider covelant modification. X <==> X*, Z modifies forward rate Y modulate backward rate. If X* is required for Y this is negative feedback.
- bandwidth goes with 1/load.
- Increase the amounts of the enzymes in the futile cycle better insulation. Opposite of Goldbeter’s Koshland’s keep these low to get switch like responses. enzyme concentration is often a signal itself.