STRUCTURAL BASIS FOR ENZYME SPECIFICITY AND CHROMOSOME STRUCTURE 
David A. Agard, Ph.D., Associate Investigator 
Dr. Agard's research is devoted to structural stud- 
ies of biological problems, in an effort to under- 
stand the fundamental relationships between struc- 
ture and function at the molecular and cellular 
levels. Four areas of investigation are being actively 
pursued: 1) three-dimensional analysis of diploid 
chromosome structure and topology; 2) studies on 
the structural determinants of specificity, using a- 
lytic protease as a model system; 3) functional and 
structural analysis of the role of the precursor in 
proper folding of a-lytic protease; and 4) pursuit of 
the first three-dimensional crystal structure of 
apolipoprotein E (apoE), an important protein in 
human cholesterol metabolism. 
I. Three-dimensional Analysis of Chromosome 
Structure. 
The Agard laboratory is engaged in a close col- 
laboration w^ith Dr. John W Sedat (HHMI, Univer- 
sity of California at San Francisco). The primary aim 
of their research is to provide a physical basis for 
understanding chromosome behavior and function 
by directly determining the three-dimensional 
structure of eukaryotic chromosomes as a function 
of both transcriptional state and the cell cycle 
stage. 
A. Three-dimensional imaging. The Agard and 
Sedat groups have developed the necessary tech- 
nologies (hardware and software) to allow them to 
examine complex noncrystalline specimens in three 
dimensions, using electron microscopy (EM) and 
light microscopy. Three-dimensional EM analysis is 
performed using electron microscopic tomography, 
which allows one to look inside chromosomes (or 
other cellular structures) and examine their inter- 
nal arrangements in three dimensions at —50- 
o 
100 A resolution. The HHMI intermediate voltage 
electron microscope at the University of California 
at San Francisco is equipped with an ultrahigh 
angle tilt stage and a cooled charge-coupled device 
(CCD) imager. This provides ideal imaging capabili- 
ties for EM tomography of specimens up to 0.5 |xm 
thick. Improved software for aligning images, merg- 
ing data in Fourier space, and calculating the recon- 
structions is now under development. Three-di- 
mensional light microscopic reconstructions use 
high-resolution optical sectioning microscopy, a 
cooled CCD detector, and image processing to re- 
move out-of-focus information. Powerful software 
for the display, analysis, and model building that 
preserves all of the three-dimensional gray-level in- 
formation has also been developed and continues 
to be improved. Most exciting is the development 
of the capability to record time-lapse three-dimen- 
sional data on living embryos, thus making possible 
the detailed study of complex dynamic cellular phe- 
nomena. 
Technological advances have been driven by the 
needs posed by the biological problems, and not 
simply by the desire for better technology. This is 
most evident in the balanced developmental effort 
for hardware and software for data collection, data 
processing, display, and analysis, as all of these 
must work together for the biology to succeed. 
B. Higher order chromosome structure. Dr. Agard 
and his colleagues have made significant progress 
in analyzing the details of higher order chro- 
mosome structure. Buffer conditions have been 
o 
found that preserve the 250 A chromatin fibers 
but allow the overall degree of condensation to be 
varied. Careful comparisons of chromosome and 
interphase nuclear morphology, using three- 
dimensional fluorescence microscopy with living 
and permeabilized cells, has indicated the condi- 
tions most similar to the in vivo state. From EM 
studies it has been learned that both the radial- 
loop and sequential helical coiling models of chro- 
mosome structure are gross oversimplifications. Mi- 
totic chromosomes are built from a fundamental 
nucleosomal fiber of —110 A diameter that is organ- 
ized into higher order structures measuring —250, 
600, and 1,300 A. 
The past year has been spent in optimizing con- 
ditions for preserving the highest level of discrete 
chromatin organization: the 1,300 A fiber. There is 
a narrow window in telophase that is particularly 
favorable for viewing this structure. A very high res- 
olution three-dimensional data set comprising 
—150 tilted views over a ±75° range has been col- 
lected, using the newly developed CCD detector 
and ultrahigh tilt stage. Although the reconstruc- 
tion has only recently been obtained, the improve- 
ment in quality compared with previous recon- 
structions using film data is extraordinary. Within 
regions of the reconstruction it should be possible 
o o 
to trace the individual 110 A fibers within the 250 A 
fiber, to trace the path of the 250 A fibers within 
Continued 
565 
