fornia, San Francisco) in all aspects of the investiga- 
tions of chromosome structure. The primary aim of 
this collaborative effort is to provide a physical basis 
for understanding chromosome behavior and function 
through direct determination of the three-dimensional 
structure of eukaryotic chromosomes as a function of 
both transcriptional state and cell cycle stage. 
Chromosomes are extremely complex, dynamic 
structures that have defied definitive analysis for 
over 100 years. The general approach has been to 
disrupt the structure, then to try to infer what the 
intact structure v^^as like. Not surprisingly, this ap- 
proach has been largely unsuccessful. The Agard 
and Sedat laboratories, however, have concentrated 
on applying high-resolution electron microscope 
(EM) tomography to overcome the extraordinary 
complexity of intact chromosomes. From these stud- 
ies, it has been learned that both the radial-loop and 
sequential helical coiling models of chromosome 
structure are gross oversimplifications. In reality, 
mitotic chromosomes are built from a fundamental 
nucleosomal fiber of ~ 110 A in diameter, which is 
organized into higher-order structures measuring 
1,300 A in diameter. 
During the past year the group has continued to 
concentrate on the structure of this 1,300-A fiber. 
Although much work remains, current data support 
a new model in which the nucleosomal fiber is folded 
back and forth upon itself to form a local domain that 
is then supertwisted about the 1,300-A fiber axis. Ef- 
forts have been focused on improving sample preser- 
vation, staining methods, and resolution of the three- 
dimensional reconstructions. Experiments in the use 
of high-pressure freezing, cryo-embedding, and a new 
DNA-specific stain are now under way. 
A key breakthrough in the past year has been the 
development of fully automated EM tomographic 
data collection. Not only does this greatly simplify 
the arduous task of collecting the 120 tilted views 
required for a reconstruction, but it limits the elec- 
tron beam exposure by 100-fold. When coupled 
with cryotechniques, this approach should effec- 
tively eliminate beam damage. The quality of the 
reconstructions should be greatly improved and 
beam-sensitive materials opened to tomographic 
study. Furthermore, the laboratory is making great 
progress in automating the rest of the reconstruction 
process, paving the way for this powerful technique 
to be useful to the entire cell biology community. 
(The project described above was supported in part 
by a grant from the National Institutes of Health.) 
Structure of Apolipoprotein E 
Apolipoprotein E (apo-E) is an important protein 
in human cholesterol metabolism. In specifically 
binding to the low-density lipoprotein (LDL) recep- 
tor, apo-E mediates cellular uptake of high-density 
lipoprotein (HDL), very low density lipoprotein 
(VLDL), and chylomicrons. Previously the Agard 
group had solved the structure of the 22-kDa recep- 
tor-binding domain of apo-E to 2.25 A. This work 
reveals that the protein is organized as an unusually 
long four-helix bundle. Although the surface is 
highly polar, most of the charged groups participate 
in salt bridges. 
Recently the Agard laboratory solved the structure 
of the two most common human mutant forms of 
this domain. One of these was known to knock out 
receptor binding, yet was not located in the putative 
receptor-binding region. Structural analysis indi- 
cates that this mutation causes a dramatic reorgani- 
zation of the salt bridges, so as to recruit a key 
receptor-binding residue into a new salt bridge. This 
represents a novel mechanism for action at a dis- 
tance. Current work focuses on producing soluble 
fragments of the LDL receptor to permit structural 
analysis of complexes with apo-E. 
Dr. Agard is also Associate Professor of Bio- 
chemistry and Pharmaceutical Chemistry at the 
University of California, San Francisco. 
Books and Chapters of Books 
Chen, H., Cly borne, W., Sedat, J. W., and Agard, 
D. 1992. PRIISM: an integrated system for dis- 
play and analysis of 3-D microscope images. 
In Biomedical Image Processing and Three- 
Dimensional Microscopy (Acharya, R.S., Cogs- 
well, C.J., and Goldgof, D.B., Eds.). Bellingham, 
WA: International Society for Optical Engineer- 
ing, vol 1660, pp 784-790. 
Kam, Z., Chen, H., Sedat, J.W., and Agard, D. 
1992. Analysis of three-dimensional image data: 
display and feature tracking. In Electron Tomog- 
raphy: Three- Dimensional Imaging with the 
Transmission Electron Microscope (Frank, J., 
Ed.). New York: Plenum, pp 237-256. 
Articles 
Baker, D., Silen, J.L., and Agard, D.A. 1992. Pro- 
tease pro region required for folding is a potent 
inhibitor of the mature enzyme. Proteins 
12:339-344. 
Baker, D., Sohl, J.L., and Agard, D.A. 1992. A pro- 
tein-folding reaction under kinetic control. Na- 
ture 356:263-265. 
Bone, R. , and Agard, D.A. 1991. Mutational remod- 
eling of enzyme specificity. Methods Enzymol 
202:643-671. 
462 
