main that is a potent inhibitor of budding. This in- 
hibitory effect is reproduced with pure Sec 1 2p cyto- 
solic fragment protein. 
Finally, the last known cytosolic component, 
Secl3p, has been purified using a strategy that in- 
volves a functional 5i?C73-dihydrofolate reductase 
(DHFR) fusion gene. Cytosol from a yeast strain that 
harbors the hybrid gene as its sole copy of SEC 13 is 
specifically depleted of the DHFR hybrid protein by 
adsorption to a methotrexate affinity column. De- 
pleted cytosol plus a methotrexate-eluted fraction 
restores vesicle budding to a urea-washed mem- 
brane fraction. Pure Secl3p isolates as a large com- 
plex and includes a new 1 50-kDa protein that is also 
required for budding. Mixture of pure Sec23/24p, 
Sarlp, and Seel 3/p 150 restores vesicle budding to 
urea-washed membranes. Hence Dr. Schekman and 
his co-workers are now in a position to examine the 
mechanics of the budding reaction with a set of pure 
proteins and an isolated membrane fraction. 
The first important clue to regulation of the bud- 
ding event has come with the detection of a cycle of 
GTP hydrolysis mediated by pure Sec proteins. 
Sarlp performs slow GTP hydrolysis and nucleotide 
exchange that require detergent or phospholipid. 
GTP hydrolysis by Sar 1 p is stimulated 1 0-fold by the 
Sec23p subunit of the 23/24p complex. GTP-GDP 
nucleotide exchange on Sarlp is stimulated fivefold 
by the cytosolic domain of Secl2p. These require- 
ments confirm and extend the observation that a 
nonhydrolyzable analogue of GTP (GTP7S) retards 
vesicle budding. The signal that triggers nucleotide 
exchange, the target of Sarlp-GTP, and the mecha- 
nistic coupling to vesicle budding are open 
questions. 
Protein Translocation from the Cytosol 
Into the Endoplasmic Reticulum 
With support by a grant from the National Insti- 
tutes of Health, Dr. Schekman's laboratory has iden- 
tified four SEC genes (SEC61, SEC62, SEC63, and 
SEC65) that are required for polypeptide transloca- 
tion into the ER. Three of the gene products are inte- 
gral proteins localized to the ER membrane. Sec6lp 
is very hydrophobic, with from five to eight poten- 
tial membrane-spanning domains. Sec62p and 
Sec63p each have two membrane-spanning domains 
and significant soluble domains that are oriented to- 
ward the cytosol. Antibodies directed against the 
cytosolic domains of Sec62p or Sec63p precipi- 
tate a complex of five polypeptides from detergent- 
solubilized membrane fractions. In addition to the 
three identified Sec membrane proteins, the com- 
plex includes two new proteins: a 31.5-kDa glyco- 
protein and a 23-kDa nonglycosylated polypeptide. 
Clones of both genes have been obtained. 
A firm connection between the three Sec mem- 
brane proteins and the translocation event has been 
established by chemical crosslinking of a translocat- 
ing polypeptide to the Sec protein complex. A trans- 
location substrate protein, yeast a-factor precursor, 
has been altered to contain a carboxyl-terminal cys- 
teine residue, which allows covalent coupling via a 
heterobifunctional crosslinking reagent to the egg 
white protein, avidin. When diluted out of a urea 
solution, the a-factor precursor portion of the con- 
jugate penetrates the ER membrane but becomes 
stuck because avidin, which is not denatured by the 
urea treatment, remains on the cytosolic face of the 
ER. Addition of a different cleavable crosslinking 
agent to the inhibited membranes produces a macro- 
molecular complex that may be isolated from a de- 
tergent-solubilized fraction by precipitation with 
Sec6lp antibody. Formation of the complex re- 
quires hydrolyzable ATP and is blocked when mem- 
branes isolated from the translocation-defective mu- 
tants are used. Thus the Sec protein complex makes 
intimate contact with a translocating polypeptide 
and may represent the elusive translocation pore or 
channel that has been sought for more than 20 years. 
Current effort is devoted to the isolation of a func- 
tional translocation complex. Yeast membranes are 
solubilized with a mixture of octylglucoside and 
phospholipid and dialyzed to remove the detergent, 
and proteins are integrated into preformed phospho- 
lipid vesicles by sonication. Reconstituted vesicles 
are competent for translocation and signal peptide 
processing of a-factor precursor in a reaction that is 
dependent on ATP and stimulated by a crude cytosol 
fraction. Fractionation of the detergent-soluble ma- 
terial has allowed the isolation of a functional com- 
plex that includes Sec63p, the luminal hsc70 iso- 
zyme BiP, and the two new proteins that were 
identified in crosslinking studies (p23 and gp31 -5). 
The participation of BiP appears to be specific: cyto- 
solic hsc70 does not replace the requirement for BiP 
in reconstituted proteoliposomes formed from BiP 
mutant membranes. An effort to resolve completely 
the membrane components required to reconstitute 
translocation is under way. 
Dr. Schekman is also Professor of Biochemistry 
and Molecular Biology in the Department of Cell 
and Molecular Biology, University of California, 
Berkeley, and Adjunct Professor of Biochemistry 
and Biophysics at the University of California, 
San Francisco. 
CELL BIOLOGY AND REGULATION 107 
