Pharyngula

The difference between biology and chemistry: biologists ask how natural systems and molecules work, while chemistry wants to know how to artificially re-synthesize natural biomolecules. Tsien favors a synthetic approach, asking hwo to build new molecules that will perform amusing and useful functions in biology. He compares it to architecture on a molecular scale that is also a rigorous test of one's understanding of molecular function.

He thinks biology has the most interesting grand questions in all of science (hooray!). He also likes pretty colors. However, he thinks that interdisciplinary barriers prevent most biologists from knowing how to manipulate molecules other than nucleotides and proteins. Anyone who knows how to make non-linear molecules has an advantage.

His first target was intracellular Ca++, which we all know is central to cellular signaling. He gave a quick overview of Ca++ properties and functions, as a rationale for why he began looking at Ca++ indicators like aequorin. He started looking for something new, started with something selective for calcium, the calcium chelator EGTA, which lacks any chromophore. One way would be to add benzene rings (chromophores) to EGTA, which led to BAPTA, basically EGTA with two benzene rings. It has high calcium sensitivity and did exhibit a color shift, but a poor one. This work led to Fura-2 synthesis, in which you can still see the EGTA ancestor, but has a much more complex set of rings attached to it. Fura-2 requires microinjection, which Tsien wasn't good at, so he synthesized a protected form of the molecule with a methyl ester protecting group that would be stripped by enzymes in the cell.

He showed a gorgeous image of the Ca++ wave in sea urchin fertilization, visualized with pseudo-colored images of calcium concentration visualized with Fura-2.

Next step: Tsien wanted to visualize cAMP, an important second messenger in cell processes. They needed a non-destructive way to assay the dynamics of cAMP, which is present at very low concentration in the presence of many other nucleotides. The idea again was to find a specific binding agent and coupling it to a chromophore. He chose to use a specific PKA subunit that binds to cAMP, and use fluorescence resonance with a pair of adjacent chromophores. It worked, and they built proteins with rhodamine and fluorescein that did the job.

They wanted a general means to fluorescently label designated proteins. This led to the discovery of Douglas Prasher's work on cloning GFP.

What was wrong with wild-type GFP? It's main excitation was at 395nm (UV), and only minor excitation peak at 475 (blue). We don't like to zap cells with UV. He puzzled out which components of the molecule were responsible for particular peaks in the spectrum (which he got wrong), but by empirically juggling in different amino acids, he came up with a variant that emphasized suitable peaks in the spectrum. This work was done without the aid of knowing the 3D structure.

They worked out the crystal structure, and then with rational design, made more variants with different colors. The work was rejected by Science for amusing reasons: reviewers didn't understand the significance. They got it published by sending a brief note to Science that announced the imminent publication of the crystal structure of wild-type GFP in Nature.

Tsien wanted to make a red version. A Russion group found a red GFP variant in a coral, which they then analyzed and found the structure. This work has led to lots of fluroescent proteins with colors from blue to red.

Problems: Sometimes FPs are too big, the excitation wavelengths are hard to get through mammalian tissues, and a few others I was too slow to get down. Darn.

They are now developing an infrared fluorescent protein based on biliverdin, and are also studying the role of proteases in cancer. They are using polycationic proteins sequences as tools to transport payload proteins into cells, with some clever tricks to regulate accessibility and make it chemically triggerable by enzymes present in tumors. They have a probe that selectively labels tumor cells, which can be a very useful guide for surgical removal of tumors. He also combines it with in vivo labeling of nerves, which surgeons don't want to cut, making it a way to color-code living tissue for tumor resection, a technique called molecular fluorescence imaging guidance, MFIG.

This was another excellent talk — Tsien has an entertaining sense of humor.