Path: utzoo!attcan!uunet!lll-winken!lll-tis!ames!ncar!gatech!bloom-beacon!athena.mit.edu!ayermish From: ayermish@athena.mit.edu (Aimee Yermish) Newsgroups: sci.bio Subject: Re: DNA for interstellar messages Keywords: multiple protein coding Message-ID: <6208@bloom-beacon.MIT.EDU> Date: 14 Jul 88 22:54:46 GMT References: <2743@bath63.ux63.bath.ac.uk> <2148@pt.cs.cmu.edu> Sender: daemon@bloom-beacon.MIT.EDU Reply-To: ayermish@athena.mit.edu (Aimee Yermish) Organization: Massachusetts Institute of Technology Lines: 29 Actually, it's pretty common for DNA (especially viral DNA) to have coding sequences which overlap or can be spliced together in different patterns. Bear in mind that not every single amino acid in a given protein is necessary for its folding or its function, and that, in fact, many proteins have chunks taken out of them after they are synthesized. Quite a few organisms use overlapping sequences to regulate the relative amounts of various proteins. As to why, well, if you're a cell, it's a slight advantage to have less DNA, because it doesn't take so long or so much effort to replicate. If you're a virus, you're under a serious constraint: the size of the capsid that will hold the genome. There are some really clever tricks some of those bugs have managed. One I recall offhand is a phage that relies on a frameshift error in translation (a rather uncommon event) to produce very small amounts of lysis protein (what pops the cell open and lets all the progeny phage free) in relationship to the amount of coat protein. In other words, it makes lots of coat protein (and hence lots of progeny phage) and a very little lysis protein (a little dab'll do ya, and you don't want to killl the cell before you're good and ready), all without the messy waste of regulatory proteins and sequences. Very elegant... --Aimee ------------------------------------------------------------------ Aimee Yermish ayermish@athena.mit.edu MIT couldn't care less about anything I say. (as long as I finish that last paper...)