How Plants Use Caterpillar Spit for Protection

How do plants protect themselves from the bugs that chew on their leaves?  In the case of the wild tobacco Nicotiana attenuata, when tobacco hornworm (manduca sexta) caterpillars feed on the leaves a collection of molecules called Green Leaf Volatiles (GLV's) is released by the plant.  GLV's are released any time a leaf is damaged, but the interesting thing is that when the damage is done by chewing caterpillars, a different form of the GLV's are produced which attracts Big-Eyed Bugs (Geocoris spp) - a predator for the caterpillars.

Image via Wikipedia

Plants emit two main types of volatile molecules: terpenoids and Green Leaf Volatiles.  The terpenoids are emitted from the whole plant and usually after a delay - maybe as much as a day after the damage.  The green leaf volatiles are more specific - they are emitted from the damaged leaf itself and it looks like they are produced at the same time as the damage.

Green Leaf Volatiles are typically 6-carbon alcohols, aldehydes or esters.  In the case of Nicotiana Attenuata they seem to mostly consist of hexenal, hexenol and simple esters of hexenol.  The interesting bit is the alkene portion of these molecules.  Alkenes can have one of two basic geometries around the double bond: the Z (or cis) isomer is locked into a u-turn shape and the E (or trans) isomer is locked into a zigzag-like orientation.

Normally, Nicotiana attenuata produces mostly the Z isomer of these molecules and a relatively small amount of the E isomer.  However something unusual happens when the damage is caused by caterpillars chewing on the leaves:  in this case the plant produces roughly equal amounts of the Z isomer and the E isomer.  You and I would probably not notice a difference in the smell of the leaves, but apparently there are bugs that can.  When more E isomer is produced, more Big-Eyed Bugs are attracted to the plants.  And the big-eyed bug eats caterpillars and their eggs.  The E isomer GLV's are a plant distress call and the big-eyed bugs are the cavalry.

How exactly does the plant "decide" which GLV isomers to make?  After testing a variety of possible candidates, it looks as though there is an enzyme in the caterpillars' saliva that causes the Z isomers to isomerize to the corresponding E isomers.  It is the caterpillar spit that produces the distress call.

If you look closely at the Z molecules and the E molecules you will notice that there are actually two changes that take place.  First, the geometry around the alkene switches.   In general, the E isomer is more spread-out than the Z isomer and as a result it is lower in energy. Given a choice the alkene will usually adopt the E geometry.  If there is a catalyst available, this change is pretty easy to understand.

The second thing that changes is the location of the alkene, the  alkene moves closer to the oxygen end of the molecule.  Enzymes are very efficient molecules and they are very sensitive to shape.  My guess is that the "real" target for the isomerase in the caterpillar saliva is the aldehyde.  The aldehyde has a carbonyl group as well as the alkene and the most stable arrangement for these two functional groups is the one in hex-2-enal.  When the two double bonds are separated by only one single bond their orbitals are able to interact and form a conjugated system.  The conjugated version is more stable than the one where the two double bonds are farther apart and unable to interact with one another.

If improved conjugation in the product is the reason that the alkene moves from the 3-position to the 2-position, why does the alkene move in the alcohol and ester molecules too?  The alcohol has only one double bond since there is no C=O, so conjugation is not possible in this molecule.  And while the ester does have a C=O, it is too far away to interact with the 2-alkene to form a conjugated system.  What gives?

Enzymes can be very selective about the molecules that they react with, but they can also be forgiving if the structure is not exactly correct.  A lot of drugs affect specific enzymes in the body - the drug isn't exactly the correct shape, but it's close enough to bind to the enzyme.  In the case of the GLV's, the alcohol and ester molecules are close enough to the right shape to bind to the enzyme and react.  In the aldehyde the enzyme causes the alkene to migrate as well as change shape because it forms conjugated molecule.  Even though the alcohol and ester don't benefit from forming a product molecule that has conjugation, the enzyme treats them the same way it treats the aldehyde and the alkene migrates to the 2-position.

The other curious thing about this is the isomerase enzyme in the caterpillar saliva.  I would bet the reason the caterpillars make this enzyme has nothing to do with attracting big-eyed bugs to come eat the caterpillars, that would be counter productive. The plants probably evolved their GLV's to take advantage of this enzyme that the caterpillars make anyway.  So what is the isomerase "supposed" to do that benefits the caterpillars?

The smell of freshly-cut grass is actually a plant distress call | IO9.COM

Allmann S, & Baldwin IT (2010). Insects betray themselves in nature to predators by rapid isomerization of green leaf volatiles. Science (New York, N.Y.), 329 (5995), 1075-8 PMID: 20798319

Anticancer Compound from a Tumor-Promoter

The title of this article caught my eye because of the irony of designing an anticancer drug by modifying a known tumor promoter. Aplysiatoxin is a tumor promoter, while Compound 1 from the paper is not. In fact, Compound 1 is just a simpler version of Aplysiatoxin: the hemiacetal has become an ether and several side groups have been lost. There are also fewer stereocenters in Compound 1 than in Aplysiatoxin.

First a little background. Both compounds affect cancerous cells the way they do because they bind to Protein Kinase C (PKC). PKC is an enzyme that contributes to a number of signaling pathways within the cell, particularly having to do with cell differentiation, proliferation and apoptosis. PKC's involvement in cellular growth cycles also results in its involvement in carcinogenesis, and it has been a target for developing anti-cancer drugs for this reason.

The curious thing about PKC is that some molecules that bind to PKC activate the enzyme, while others de-activate it. Even stranger, some activators promote tumor formation and other activators do not.

PKC activators have shown some promise for treating diseases such as Alzheimers or AIDS, but their tumor-promoting behavior is a big drawback. Ideally you would want to find a PKC activator that was also non-tumor promoting. Bryostatins fit this description, but the compounds are too complex to be easily made in the laboratory. In nature, bryostatins are made by a coral-like organism but in extremely small amounts. According to Wikipedia, you would need a ton (2000 lb) of bryozoans to obtain just one gram of bryostatin.

Aplysiatoxin binds to PCK as a tumor promoting activator. Compound 1 was designed as a simpler version of aplysiatoxin that might be a PKC activator without also being a tumor promoter. As it turns out, compound 1 shows minimal tumor-promoting activity, and it counteracts the effects of the tumor-promoter 12-O-tetradecanoylphorbol-13 acetate. It's anti-cancer activity as well as it's mode of binding to PKC seems to be comparable to the bryostatins. The authors report that they can make Compound 1 in only 22 steps, which makes it a promising alternative to bryostatins as a potential therapeutic agent.

Nakagawa, Y., Yanagita, R., Hamada, N., Murakami, A., Takahashi, H., Saito, N., Nagai, H., & Irie, K. (2009). A Simple Analogue of Tumor-Promoting Aplysiatoxin Is an Antineoplastic Agent Rather Than a Tumor Promoter: Development of a Synthetically Accessible Protein Kinase C Activator with Bryostatin-like Activity Journal of the American Chemical Society, 131 (22), 7573-7579 DOI: 10.1021/ja808447r

Quorum-Sensing Molecules

I was fascinated by Bonnie Brasler's TED talk on Quorum-Sensing, and being a chemist I wanted to know more about the molecules involved. She did put up a slide with structures during the talk, but I wanted more so I did a search on PubMed and found this Perspective written by Brassler and Michael Federle.

My only experience with the notion of a “quorum” is our Faculty Assembly where we sometimes have difficulty achieving a quorum. In order for the meeting to be “official” and for any votes taken to be valid we need to have a minimum number of faculty present, a “quorum.” For bacteria, quorum sensing is the way the bacteria “count” one another. The bacterium releases a particular molecule, called an autoinducer - if there are lots of the molecules, then there are a lot of bacteria. If there are very few autoinducer molecules, then there are few bacteria present. The bacteria has a protein receptor that binds to the autoinducer molecule – so the bacteria can “sense” the presence or absence of autoinducer molecules depending on whether or not the receptor protein has detected any. In this way the bacteria can change their behavior depending on the number of bacteria present, as measured by the number of autoinducer molecules it finds. As a group, the bacteria behave one way when there is a low density of bacteria present and a different way when there is a high density of bacteria present.

In the simplest examples, quorum-sensing allows the bacteria to switch between two different behaviors depending on the number of bacteria present. One example would be the staphylococcus aureus bacteria – at low density they adhere to the surface of the cells of the host organism where they can grow and produce more bacteria. Once they reach a “quorum” there are enough bacteria present to be able to invade the host cells Their metabolism then shifts from producing the proteins that allow attachment to the outside of host cells and starts to produce proteins and toxins that allow the bacteria to enter the host cells. The light-producing bacteria from Bonnie Brassler's TED talk produce light when there are a lot of bacteria present, and stop producing light when there are few bacteria present.

Enough about biology, what about the molecules involved? In this Perspective, two categories of autoinducers are discussed, and one “special case.” Gram negative bacteria produce a type of autoinducer referred to as AHL for Acyl Homocysteine Lactone. Different types of bacteria will have different acyl groups attached to the homocysteine, and only recognize their own type of AHL. Gram positive bacteria do not use AHL's, instead they produce specialized proteins called AIP for AutoInducing Peptides, which consist of a string of 5 to 17 amino acids, some of which may be modified. The two types of autoinducer (AI) are detected by the bacteria when the AI binds to a receptor molecule in the bacteria. The details differ, but when enough AI's are around to bind to their receptors, the receptor causes a change in gene expression in the bacteria, which leads to a different behavior by the bacteria. 

The AHL's and AIP's are species specific: each type of bacteria produces only one AI and only recognizes it's own AI. The third type of molecule discussed is an unusual boron-containing molecule that may have a role for communication between different species of bacteria. The light-producing bacterium vibrio harveyi produces two different autoinducer molecules. The first is referred to as AI-1. AI-1 is an AHL molecule used for communication only among the V. harveyi bacteria. The other autoinducer is AI-2 which, on the other hand, may have a role in allowing different species of bacteria to communicate with one another.

AI-2 is synthesized by the bacteria in three steps from S-adenosyl methionine. The enzyme for the final step in this synthesis is called LuxS and as it turns out the gene for LuxS is found in many different bacteria, which all seem to both make and respond to the presence of AI-2. The implication of this is that perhaps AI-2 serves as some sort of generic autoinducer that allows bacteria to sense not only their own species, but also all other species of bacteria that produce AI-2.

The really interesting thing is that if we understand how bacteria communicate, we can find ways to short-circuit that communication. Many pathogens use quorum-sensing to regulate their virulence. In the example I mentioned earlier about S. aureus, the bacteria depend on reaching a “quorum” before they begin to “invade” the host cells. If their ability to sense one another is prevented, then perhaps their ability to invade the host and cause disease could be reduced.

Federle, M. (2003). Interspecies communication in bacteria Journal of Clinical Investigation, 112 (9), 1291-1299 DOI: 10.1172/jci200320195

Structure of a Viral Protein Coat

The June issue of Popular Science has an image of the 3D molecular structure of the protein capsid for the Penicillium stoloniferum virus (PsV-F). This is the protein outer shell of the virus that acts as a container for the genetic information of the virus. The image doesn't appear to be on the Popular Science web site, so I went looking for the protein structure at the Protein Databank.

The image above comes from the PDB entry for structure 3es5 which is the crystal structure of the dimeric capsid protein (CP). The image above shows the biological unit which consists of 60 copies of the CP dimer in a roughly spherical arrangement. In Atomic Structure Reveals the Unique Capsid Organization of a dsRNA Virus, Pan et. al. used electron cryomicroscopy to compute a 3D reconstruction of the entire 120-mer of the protein capsid. The image in PopSci and the images in the original paper are much cooler than the one above, but this gives you an idea of what it would look like.

If you can't find a copy of Popular Science, you can see the original article at PNAS or Pubmed Central in September, since PNAS makes articles available for free 6 months after print publication.

ATP-Synthase Modeled in Glass

The Medical Museion at the University of Copenhagen has this interesting sculpture of ATP-Synthase entirely in glass. Take a look at their site. Scroll down to the comments where the artist, Colin Rennie, shares some of his thoughts on the scuplture.  It's difficult to really see what it looks like from the photos. I would love to got there to see it myself.

ATP-Synthase is the enzyme responsible for making most of the ATP formed in cells - the main form of stored cellular "energy."  The sculpture is based on a crystal structure published in 1999 and available at the Protien Data Bank, PDB ID 1QO1.

By way of IO9.com

HIV1 Protease and "Flap" Orientations

Enzymes are the machines of the cell - they make almost all of the chemical reactions that take place within a cell happen in a realistic time scale. They are able to do this because they bind specifically to a target molecule (the substrate) and convert it into a new molecule.

There are two models that are frequently used to describe how this "binding" works. The simplest is the Lock-and-Key model which assumes that the enzyme is a rigid molecule with a hole in it, rather like a lock and its keyhole. If a key has the right size and shape to fit into the keyhole, it might be able to open the lock. The enzyme has an opening called the active site - molecules with the right size and shape can fit into this active site and be modified by the enzyme. A drug (an inhibitor) can be designed that has the right size and shape to fit into the active site, but then it gets stuck. Once the active site is blocked by the inhibitor, the enzyme can no longer convert substrate molecules and it no longer works.

This model is rather limited - enzymes are often quite flexible. The second model, called induced fit, says that the enzyme changes shape when it binds to the substrate or inhibitor. If this happens, it will be important to know not only what the active site is like, but you also need to know how the enzyme changes when it binds to the substrate or inhibitor if you want  to design an effective drug.

Two recent papers examine how binding to an inhibitor may affect the shape of HIV-1 protease. HIV-1 protease (HIV PR) is an essential enzyme in the functioning of HIV and the target of many drugs for treating AIDS. If HIV-1 protease can be inhibited, none of the other proteins needed by HIV will get processed into their active forms.

HIV-1 protease bound to an inhibitor. Image from http://en.wikipedia.org/wiki/File:Hiv-1_pdb_1ebz.png

Drug Pressure Selected Mutations in HIV-1 Protease Alter Flap Conformations looks at mutations in HIV-1 protease.  When HIV is exposed to protease inhibitor drug cocktails they observe mutations in HIV PR, especially in the two loops or flaps that cover the active site.  Mutations that reduce the ability of the drug to bind to the enzyme active site will be resistant to the influence of the drug.  Changes in the active site itself would obviously have an affect on the ability of the drug to bind effectively, but why would the protein develop mutations in the flap region, which is not directly related to the active site?  By looking at the orientation of the flaps when different inhibitors are bound to the mutants, they suggest that the flaps adjust to accomodate the binding of the substrate/inhibitor in order to fine tune the binding strength.

The second paper, Dynamics of “Flap” Structures in Three HIV-1 Protease/Inhibitor Complexes Probed by Total Chemical Synthesis and Pulse-EPR Spectroscopy, also looks at the flaps and inhibitor binding.  They also see evidence that the flaps move in response to the nature of the substrate as it is bound to the enzyme.  The reaction catalyzed by HIV PR involves two distinct chemical steps, so they chose inhibitors that resembled different stages along the reaction sequence:  during the first step, between steps one and two, and during step two.  They conclude that the flaps move to fine-tune the interaction between the enzyme and the substrate as the reaction procedes.

Both papers report evidence of "induced-fit" behavior in the way HIV PR interects with it's substrate.  Understanding what role the flaps play in substrate binding can lead to better drug s for treating Aids.

Luis Galiano, Fangyu Ding, Angelo M. Veloro, Mandy E. Blackburn, Carlos Simmerling, Gail E. Fanucci (2009). Drug Pressure Selected Mutations in HIV-1 Protease Alter Flap Conformations Journal of the American Chemical Society, 131 (2), 430-431 DOI: 10.1021/ja807531v

Vladimir Yu. Torbeev, H. Raghuraman, Kalyaneswar Mandal, Sanjib Senapati, Eduardo Perozo, Stephen B. H. Kent (2009). Dynamics of “Flap” Structures in Three HIV-1 Protease/Inhibitor Complexes Probed by Total Chemical Synthesis and Pulse-EPR Spectroscopy Journal of the American Chemical Society, 131 (3), 884-885 DOI: 10.1021/ja806526z