The acidity problem in the two-hybrid system

Vladimir Svetlov svetlov at
Tue Nov 25 17:49:52 EST 1997

In article <Pandol-2511972301400001 at>, Pandol at
(Per-Anders Olsson) wrote:

> What is the minimum number of local acidic amino acids, D´s and E´s, that
> can give self activation of a bait in the two-hybrid system, in spite of
> an acceptable average isoelectric point for the whole bait peptide?

There is none. So called acidic activation domains were shown by
mutagenesis or combinatorial analysis to depend on hydrophobic (aromatic
and branched) more than on acidic residues. In other words acidic amino
acids are a good dyagnostic feature of natural activation domains but in
many cases they are dispensable (see Leuther et al., Cell, 1993,
72:575-585; Regier et al., PNAS, 1993, 90:883-887), whereas mutations
affecting hydrophobic amino acids in those domains are usually detrimental,
as sometimes are the internal introduction of helix-breaking residues (such
as P). If you are looking for a possible acidic activation domain in your
bait, you might want to look for putative alpha-helices (those seem to be
statistically prevalent among acidic activators) with both acidic (one or
two might be all you find) and hydrophobic (L, I, F and Y) amino acids.
Whether the bait will behave as an activator depends also on the number of
binding sites in the reporter vector, dimerization domains in the bait and
DNA-binding tag. The rule'o'thumb is "multimerization=stronger activation".
Is that confusing enough <G>?

Vladimir Svetlov
McArdle Lab for Cancer Research
Dept. Oncology
1400 University Ave.
Madison, WI 53706

More information about the Methods mailing list