Fringe S2E14 (2010) Part III - "Show and Tell"

Spoiler Level – Low (series); moderate (episode)

The Show – FBI agent and crazy scientists solve crimes, commit crimes against science.

Like I keep saying. I'm actually a fan of the show, but every now and then the science gets too terrible for me to let stand.

The Scene – A deadly toxin is killing only certain genetic groups. It enters the air after being exposed to heat (either from a lit candle or hot tea). Just after the second attack, one of the show's scientists reveals a molecular model of the toxin. He shows that it has two parts: a deadly part that is just chromium trioxide and hydrogen cyanide, and an interchangeable targeting portion that recognizes the DNA sequence of the target genetic group. He also finds that the toxin has a seahorse on the end, an apparent signature.

This toxin will haunt my nightmares. But mostly just because of the terrible aesthetics. From Fringe.

The Science – First of all, the idea that the toxin they are looking at is a mixture of chromium trioxide and hydrogen cyanide is ridiculous. Those are molecules with 4 and 3 atoms, respectively, and the structure shown on the screen is much much bigger than that. Also, while those chemicals are great toxins on their own, they can't just be attached to another molecule and still function normally. If you want something that can enter a cell, recognize DNA, and then carry out a toxic function, you need a protein.

But I really wanted to discuss this scene because I don't think the public is clear on how scientists know the structures of the molecules they are studying. There are many ways to do this, and I hope I can talk about more of these methods later. For small molecules, various types of spectroscopy and mass spectrometry would work. For smaller proteins, nuclear magnetic resonance is an option, but what we're dealing with would have to be a pretty large protein. Electron microscopy will give some structural information but not the atomic-level resolution seen in the model. For that, we must use X-ray crystallography.

X-ray crystallography is a very important technique that was instrumental in the discovery of DNA's structure (see Rosalind Franklin), and is most often used today to look at the structure of various proteins. Basically, the proteins are crystallized, and the crystals are shot with a beam of X-rays. The atoms in the protein crystals cause the X-rays to diffract (or bend), and the diffraction pattern can be interpreted to determine a three-dimensional picture of the protein.

This is even more complicated than I made it sound. To make the crystals, you must begin with a large amount of very pure protein and test a large number of crystallization conditions. Even if you get crystals, there's no guarantee that they will diffract well. And even with good diffraction, it is not easy to interpret the pattern. Determining the structure of a protein from scratch can take years; entire PhD projects have centered around getting a single structure. It's ridiculous to think that the scientists in Fringe could get a crystal structure in a day.

Also, that structure looks terrible. It looks like whoever designed it saw a picture of a molecular structure, and interpreted it as just a random arrangement of orbs and sticks. A collection of orbs and sticks (representing the atoms and bonds of the proteins, respectively) is one of the ways to depict a molecule. But it's not a good choice for a wide view of a whole protein. There are better choices that show less atomic detail but more broad structural details.

Two different depictions of a real toxin. Which one would you show in your TV show? From PDB 1JKY. Made using PyMOL.

The idea that the person who designed the protein could tag it with a chain of carbon atoms arranged as a seahorse is pretty silly as well. While some scientists are doing interesting work with molecular art, getting a chain of carbons to form a seahorse is not really on the horizon. Though I will admit that the line about scientists often leaving a signature on their proteins is not entirely false. I have spelled out my name using the amino acids in one of the proteins I've made.

The cutting edge or molecular art. Made from DNA! From Nature.

Finally, I just need to comment on the idea of getting the protein into the air through heat. That's a fine technique for small molecules that are stable as gases but not for proteins. Proteins are stable in water-based solutions and as solid powders. Exposing proteins to high heat will not cause them to enter the air, but would most likely result in their destruction.

Adding a toxin to someone's drink is a fine way to kill one person. Not a whole room full of people. Unless your victim is really good about sharing, I guess. From Fringe.

Fixing the Scene – First, just say the toxin is a protein and not a mix of two small molecules. This requires tweaking the plot a little since they track the culprit using the toxin. However, proteins must be made from synthesizing the DNA that encodes them. To do that, the culprit would need to either buy a very expensive machine or pay a company to make it for him. In either case, he could be tracked through that purchase.

While there's no way to get a real structure of the toxin in the time frame of an FBI investigation, there are other ways to get a good model. They could figure out the toxin's sequence of amino acid building blocks using mass spectrometry. Then, if their toxin is similar to existing toxins and DNA-binding proteins with known structures, the scientists could make a theoretical model of the toxin based on already existing data. It would not be a perfect representation of the toxin, but it would be close enough. And have someone who actually know what a protein looks like design the model.

Attaching a molecular picture of a seahorse to the end of a protein is just bizarre. Have the scientist sign his name in the amino acid sequence of the protein like a normal scientist would. Or at least like I would.

Finally, candles and tea are just not good for protein dispersal. Later in the show, the villain uses a small device that quickly expels the toxin as a powder over a large area. This technique would be perfectly appropriate and should have been used throughout the episode.

So overall, the general concept of a toxin that only targets certain genetic groups is workable. However, this episode manages to botch an incredible number of details. Which is a shame, because all of the problems are ones that could have been fixed if someone who knew the science took a look at the script.

Next Week – I try and get back to writing shorter posts when discussing Spider-man and getting bitten by genetically engineered spiders.