channel mouth. They have found that the net charge 
on the carboxyl-terminal half of the peptide is quite 
important for determining the rate of binding, 
whereas the stability of the bound (inactivated) 
state is predominately determined by hydrophobic 
interactions between residues in the amino-terminal 
half of the peptide and its binding site on the 
channel. 
All Shaker potassium channels contain a consen- 
sus cAMP-dependent phosphorylation site near the 
carboxyl terminal. In addition, the 5^a^er D variant 
has a potential cAMP phosphorylation site in the 
amino-terminal inactivation particle region. To test 
if these sites are important for regulating the proper- 
ties of the channels, Drs. Peter Drain, Adrienne Du- 
bin, and Aldrich applied bovine alkaline phospha- 
tase (BAP) and cAMP-dependent protein kinase 
(PKA), with appropriate substrates and cofactors, to 
the internal face of inside-out patches from oocytes 
expressing Shaker D. An increase in current and a 
decrease in the time constant of macroscopic inacti- 
vation upon application of BAP could often be re- 
versed by PKA. These results demonstrate a poten- 
tial mechanism for regulation of neuronal activity 
by modulation of potassium channel inactivation 
rates. (The project described above was supported 
in part by a grant from the National Institutes of 
Health.) 
Mechanisms of C-Type Inactivation 
The rate of C-type inactivation can vary by two 
orders of magnitude among different carboxyl- 
terminal Shaker variants. In many previous studies 
of Shaker channel inactivation, the distinction be- 
tween the two types of inactivation has not been 
made clear. However, because N-type inactivation 
can be eliminated by mutations in the amino- 
terminal domain, C-type inactivation can be studied 
in isolation. Using amino-terminal mutants, Drs. Ho- 
shi, Zagotta, and Aldrich constructed chimeras from 
Shaker variants with C-inactivation rates differing 
by two orders of magnitude. These and subsequent 
point mutations allowed them to localize, in the 
sixth membrane-spanning region, a single-amino 
acid difference that was responsible for the large 
difference in rates between these two variants. (The 
project described above was supported in part by a 
grant from the National Institutes of Health.) 
Because the substitution that caused the large dif- 
ference in C-type inactivation rates was a change 
between an alanine and a valine, Drs. Hoshi, Za- 
gotta, and Aldrich examined the effects of substitu- 
tion of other nonpolar amino acids at this position. 
Channels with a glycine inactivated slowly, much 
like the alanine-containing channels, whereas chan- 
nels with an isoleucine inactivated with an inter- 
mediate rate. Surprisingly, these mutations also 
changed the single-channel conductance, with 
valine- and isoleucine-containing channels having 
about a 1 .5 and 2 times larger conductance, respec- 
tively, than alanine- or glycine-containing channels. 
Paul Zei, Dr. Hoshi, and Dr. Aldrich found that the 
changes in conductance do not reflect an alteration 
in ion selectivity of potassium relative to sodium, 
rubidium, or ammonium. Drs. Jose Lopez-Barneo, 
Stefan Heinemann, Hoshi, and Aldrich have found 
that the C-type inactivation rate can be markedly 
influenced by external cations in a manner consis- 
tent with the idea that inactivation is impeded when 
the channel is occupied by permeant ions. (The 
project described above was supported in part by a 
grant from the National Institute of Mental Health's 
Silvio Conte Center for Neuroscience Research.) 
Mechanisms of Vohage-Dependent Activation 
The ability to eliminate N-type inactivation and to 
slow C-type inactivation by mutations allows the 
study of single-5^afeer-channel activation in rela- 
tive isolation from inactivation. Dr. Hoshi, Dr. Za- 
gotta, Jeremy Dittman, and Dr. Aldrich have per- 
formed a detailed study of the single-channel, 
macroscopic, and gating currents in these channels 
to determine the minimum number of states re- 
quired, the transition rates between them, and the 
voltage dependence of these rates. The results have 
been used to develop an understanding of wild-type 
channel activation that will provide a basis for inter- 
preting the results of mutagenesis. (The project de- 
scribed above was supported in part by grants from 
the National Institutes of Health and the National 
Institute of Mental Health's Silvio Conte Center for 
Neuroscience Research.) 
Dr. Aldrich, Dr. Catherine Smith-Maxwell, Dr. Za- 
gotta, Max Kanevsky, Jennifer Ledwell, Michael 
Root, and Melinda Przetak have generated and stud- 
ied mutations in regions shown to be involved in 
voltage-dependent activation . A particularly interest- 
ing mutation is one in which a leucine residue in the 
cytoplasmic loop between the fourth and fifth mem- 
brane-spanning regions is changed to a tyrosine. Ala- 
nine, phenylalanine, and tyrosine substitutions at 
this position result in shifts in the conductance- 
voltage relationship to more positive potentials. 
The tyrosine substitution, however, drastically 
changes the macroscopic activation kinetics, in ad- 
dition to the shift in the voltage dependence. The 
wild-type channels activate much faster than the ty- 
rosine mutants. The importance of the hydroxyl 
group may lie in its ability to hydrogen bond to 
solvents. 
NEUROSCIENCE 387 
