I am a microbiologist and molecular biologist turned tenured biotech faculty turned bioinformatics scientist turned entrepreneur. My passion is developing instructional materials for 21st century biology (
DigitalBio RSS feed
When can a really bad virus be used to do something good? 
When we can use it to learn.
The human immunodeficiency virus, cause of AIDS, scourge of countries, and recent focus of ScienceBlogs; like humans, evolves. As one of my fellow ScienceBloggers noted, few biological systems demonstrate evolution as clearly as HIV. In this series, I'm going to guide you through some experiments on HIV evolution that you can do yourself. You won't even have to put on any special clothing (unless you want to), wash glassware or find an autoclave. And, you don't need to any UNIX commands or borrow a fancy computer; you can use a PC with Windows or a Mac (with OS X). As long as you can run a web browser and of course, have an Internet connection that should be enough.
I'm also going to break this experiment up into four parts, so that you can work on solving the puzzle, too, before I give you the answers. (I'm really not a sadist, I just think this is more fun.)
Part I. Today, I'll introduce the experiment and give a link to a short flash movie on HIV.
Part II. Instructions for the experiment.
Part III. Look at the sequence results.
Part IV. Look at protein structures and see if we can explain why the experiment worked the way it did.
First, we'll set the ground rules. Evolution occurs through multiple mechanisms, but we're going to focus on natural selection in this article. I like to tackle problems one at a time.
Second, here are the ideas that we're going to test:
Natural selection is a process where populations change over time.Natural selection occurs because:
1. Populations are genetically diverse.
2. Every population has individuals who reproduce and individuals who do not.
3. Traits can only be inherited from the individuals that reproduce.
4. Over time, the characteristics of a population change.
These statements allow us to make predictions that can be tested. Statement 1, for example, that populations are genetically diverse, can be tested by looking at some feature in a set of individuals from a population. If HIV is genetically diverse, we should see this reflected in samples from the HIV virus.
The other statements can be examined by looking at genetic changes that have occurred in a set of HIV samples and have resulted in a new trait. We're going to repeat an experiment that was done earlier by Colonno, and collaborators and see if we can identify any genetic changes that have occurred in a large number of HIV isolates.
Here's our experiment in a nutshell:
We are going to compare a protein sequence from a wild type, drug-sensitive, HIV virus with protein sequences from HIV samples that were isolated from patients who were taking an anti-viral drug (actually a protease inhibitor) called "Atazanavir."
HIV protease is the protein target for that drug. This protease is a protein that HIV requires for replication. Atazanvir inhibits the activity of the protease by binding to it's active site. This isn't a perfect analogy, but if you think about what a protease does, its job is to chew up proteins (like you've seen on detergent commercials) and its active site is where it does the chewing, kind of like your mouth. A protease inhibitor keeps the protein from being able to chew. If your mouth were the active site and your mouth was wired shut, or stuffed full with a giant chewing inhibitor (if you put a sock in it?), you wouldn't be very good at chewing either.
Why do we care?
Some of the proteins that HIV makes are made as polyproteins - that is, some of the proteins are joined together in a longer chain of amino acids. They can't go out and do their jobs unless the HIV protease sets them free by cutting the bonds that join them to the other proteins in the chain. If the protease can't cut up the polyprotein, the virus can't reproduce. There is a nice animation of the HIV life cycle at the Johns Hopkins University, so go check it out, and see a cartoon of the protease in action.
Then come back in a few days, for the next part of the story.
technorati tags: digital biology,
AIDs, bioinformatics, evolution, HIV,
protease,
genetics
Del.icio.us
Digg
StumbleUpon
Reddit
Facebook
Twitter
Comments
Here's another post I wrote on HIV, CSI, and evolution.
Posted by: RPM | August 18, 2006 5:53 PM
Thanks RPM,
That's a great post!
And a great movie. That dentist was definitely plant food.
I remember when the Acer paper was published. I used to use that one as an example when I taught the section on PCR in my biotechnology lab course.
There have been some other good epidemiology cases, too, that are not HIV related, that have been solved by some phylogenetic sleuthing. I have a collection of a few papers on using phylogeny in epidemiology here. There's a surgeon transmitting Hepatitis C to patients, and the link between pathogenic E. coli (O1H57) and organic, unpasteurized juice, and a PubMed search that brings up lots of stuff on oysters and Hepatitis A.
Posted by: Sandra Porter | August 18, 2006 6:36 PM