W.K. Keck Center for Computational Biology and Verna and
Marrs McLean Department for Biochemistry, Rm. 326B
Baylor College of Medicine, 1 Baylor Plaza, Houston TX 77030,
U.S.A.
My interest lies currently in three different areas:
Biological macromolecules are generally embedded in ice or a sugar to preserve their native structure during electron microscopic observations. In many cases they are prepared without a carbon support film to avoid structure distortions or preferential orientation. Images of these types of specimens often reveal specimen movement in different directions film breakage upon electron irradiation. A direct effect of specimen irradiation is secondary and Auger electron emission due to inelastic scattering. This emission renders the specimen positively charged. Charging has long been recognized as a problem in conventional fixed-beam and scanning electron microscopy. So far, no simple technique has been proposed for the evaluation of a specimen's susceptibility to charging or the extent of charging.
In collaborating with John Berriman, we have developed procedures to assess charging of biological specimens under electron irradiation in an electron cryomicroscope. Charging can be observed by an expansion of the illuminating beam (Fig. 1), blurring of electron diffraction patterns (Fig. 2) and by beam 'footprints' on the specimen (Fig. 3). These 'footprints' or annuli can easily be observed after irradiating the specimen to an electron dose as low as 0.1 e/Å2 (Fig. 4). Discharging can also be seen in the defocused electron diffraction mode. We investigated the influence of a variety of factors on the magnitude and visibility of charging. A reduction of charging is noticed when part of the adjacent carbon film is included in the irradiated specimen area.
In collaborating with Heinz Gross, thin conductive films were deposited onto catalase crystals prepared across holes using ion-beam sputtering and thermal evaporation and evaluated for the effectiveness of charge reduction. Deposits applied by ion-beam sputtering reduced charging but concurrently resulted in structural damage. Coatings applied by thermal evaporation also reduced charging, and preserved the specimen structure beyond 5 Å resolution as judged from electron diffraction patterns and images of glucose-embedded catalase crystals tilted to 45o in the microscope (Figs. 4 & 5). This study demonstrates for the first time the feasibility of obtaining high resolution data from unstained, unsupported protein crystals with a conductive surface coating.
Fig. 1: Specimen charging as judged from changes in the beam diameter. A
series of frames extracted from a video sequence obtained using the Gatan
wide-angle TV-rate CCD camera demonstrating the change in beam diameter
when the edge of the beam is moved from the holey carbon film (a) into the
hole containing an unsupported layer of ice (b through i). The direction
in which the specimen is moved is indicated by the arrow (a). The beam
diameter has been indicated by the arrowheads. As the beam is moved from
the holey carbon onto the ice layer (b), the beam diameter increases.
This increase is ~20 % as indicated in (h). As the beam returns onto the
holey carbon, now on the opposite side of the hole (j), the beam diameter
reduces to its original width (k and l). The changes in beam diameter can
occur just before the beam leaves, or just after it moves back onto the
holey carbon. Bar is 1 micron. Animated examples the beam expansion can be
found here.
Fig. 2: Specimen charging as judged from (a) the blurring of Bragg
reflections, (b) the expansion of the undiffracted beam in electron
diffraction patterns acquired from glucose-embedded catalase crystal using
a 2.5 microns diameter beam without irradiating part of the carbon
support. (c) DIFF image acquired immediately afterwards from a larger,
7.5 microns-diameter reveals a dark annulus. (d-f) The EDP, undiffracted
beam and DIFF image from the same specimen but including a carbon film
area show sharp Bragg reflections out to 3 Å, a sharper central beam spike
and a less pronounced footprints, respectively. The bar is 1 micron.
Fig. 3: Footprints of the beam observed on glucose-embedded crotoxin
crystals prepared on continuous carbon film. (a) DIFF image recorded from
a 7.5 microns-diameter beam after acquiring an EDP from a 2.5
microns-diameter area. A footprint indicated by the arrow coincides with
the area irradiated by the electron beam. (b) After spot-scan imaging the
same crystal, a raster of small annuli (arrows) can be seen, each with a
diameter of approximately 2000 Å corresponding to the beam diameter used
for spot-scan imaging. Bar is 1 micron.
Fig. 4: 'Footprints' or annuli visible on glucose-embedded catalase crystals
prepared across holes. (a) Image recorded of a 7.5 micron-diameter area showing
the specimen consisting of catalase crystals suspended in glucose across a holey
carbon film. (b) After irradiating a 2.5 micron-diameter area to 0.1
e/Å2, another image, like the one in (a) was recorded.
This image now shows a clear annulus (see inset). This annulus has a diameter
of 2.5 micron anda width of 2000 Å. (c-e) Repeated acquisition of
images as in (a) without additional electron exposure reveals that the annulus
disappears. (f-j) Recorded as in series (a-e), demonstrating that the annulus is
not related to specimen contamination. Bar is 1 micron.
Fig. 5.Computer processed EDP obtained from glucose-embedded catalase
crystal on holey carbon coated with 30 of carbon on both sides by thermal
evaporation. The crystal was tilted to 45o.
Tilt axis is indicated by lines. Reflections extend out to 3.7 Å
isotropically; some broadening is
present in the direction perpendicular to the tilt axis indicating a small
amount of crystal curvature. The pattern was processed to remove the
radial background due to inelastic scattering and the spike.
Fig. 6. (a) IQ-plot of a spot-scan image obtained from the
glucose-embedded catalase crystal shown in Fig. 5. The crystal was tilted
to 45o. The lattice directions are indicated along with the tilt
axis.
Resolution shells at 5, 4 and 3 Å resolution are indicated by dashed
circles. Shown are reflections with IQ equal to or better than
6 as identified after lattice
refinement and an analysis of the Fourier transform using the program
MMBOX. Reflections are present along the tilt axis well beyond 5 Å
resolution. (b) DIFF image of the catalase crystal after spot-scan
imaging. The image shows a raster of annuli (arrow) from the electron
beam as evidence of a residual amount of charging that results only from
imaging.
References:
Two-dimensional (2-D) protein crystallization on lipid monolayers is an emerging technique with potential application to structure elucidation of biological macromolecules. Aside from ionic interactions utilized for instance in 2d-crystals obtained from RNA polymerase, binding to the lipid monolayer can also occur via specific ligands, as in gangliosides, biotin, or more recently using Ni-chelating lipids. Recently in our group, 2d-crystals were obtained from streptavidin that diffracted to 3.5 Å resolution, enabling visualization of beta-sheets and clusters of amino acid side chains. In collaboration with W. Frey and V. Vogel (at UW, Seattle) 2d-crystals were grown of streptavidin using Cu-chelating lipids that bind to single surface-accessible histidines (Fig. 1). After processing and merging, the 2d-projected structure showed remarkable similarity to the reconstructed projection obtained from ice-embedded streptavidin complexed to biotin as viewed along the R dyad axis (Fig. 2). Current focus for 2d-crystallization is on tetanus toxin using the ganglioside GT1b.
Fig. 1: 2d-crystal of streptavidin complexed to Cu-chelating lipid monolayer
contrasted in uranyl acetate.
Fig. 2: 15 Å projected structure after crystallographically merging
7000 unit cells.
References:
A Gatan (model 679) 1024 x 1024 slow-scan charge-coupled-device (CCD) camera has been interfaced to our JEOL4000EX electron cryomicroscope and explored for its usefulness in the electron crystallographic analysis of thin protein crystals for diffraction and imaging studies.
For diffraction, we used glucose embedded crotoxin complex crystal kept at -125oC as a test specimen. We found that the camera allows for an on-line assessment of the crystals' flastness (Fig. 1), crystallinity (Fig. 2) and thickness. Intensities obtained from electron diffraction patterns acquired with the camera have been statistically analyzed and were found to be consistent with theoretically expected values. A quantitative analysis of the diffraction intensity as function of the accumulated electron dose suggests the possibility of recording up to 250 diffraction patterns with 3.5 Å resolution from a single crotoxin complex crystal 128 Å thick. Tilt series of 125 electron diffraction patterns with 3.5 Å data acquired from a single crystal were found to be practically feasible (Fig. 3). Our study demonstrates for the first time the effectiveness of using a slow-scan CCD camera for electron diffraction data collection from thin protein crystals at near atomic resolution.
We have also evaluated the feasibility and limitations of the CCD camera for imaging in our 400 kV electron cryomicroscope. We used catalase crystals and amorphous carbon film as test specimens. Using catalase crystals, we found that the finite (24 micron) pixel size of the slow-scan CCD camera governs the ultimate resolution in the acquired images. For instance, spot-scan images of ice-embedded catalase crystals showed resolutions of 8 Å and 4 Å at effective magnifications of 67,000x and 132,000x, respectively (Fig. 4). Using an amorphous carbon film, we evaluated the damping effect of the modulation transfer function (MTF) of the slow-scan CCD camera on the specimen's Fourier spectrum relative to that of the photographic film. The MTF of the slow-scan CCD camera fell off more rapidly compared to that of the photographic film and reached the value of 0.2 at the Nyquist frequency (Fig. 5). Despite this attenuation, the signal-to-noise ratio of the CCD data, as determined from reflections of negatively stained catalase crystals, was found to decrease to ~ 50% of that of photographic film data. The phases computed from images of the same negatively stained catalase crystals recorded consecutively on both the slow-scan CCD camera and photographic film were found to be comparable to each other within 12o.
To evaluate the reliability of phases retrieved from 400 kV spot-scan images acquired using our CCD camera, we used glucose-embedded purple membranes as a test specimen (Fig. 6). This specimen was chosen because it represents a broad class of low-contrast radiation-sensitive biological objects and its structure is well established. The amplitudes of computed reflections from these images were strongly damped by the modulation transfer function of the camera. Nevertheless, their phases on average were < 12o different from the reference data of Henderson et al. (1986) up to 8.8 Å resolution, which corresponds to 0.8 of the Nyquist frequency of the camera.
Fig. 1: A low-magnification image of a crotoxin complex crystal recorded in
the JEOL4000EX at 400 kV using the CCD camera.
This CCD camera was purchased with financial support from the
W.M. Keck Foundation.
This glucose-embedded crystal measures over 3 micron on edge and
is ~128 Å thick. They were prepared on holey carbon
grids (200-400 holes/inch) covered with a thin
continuous carbon film.
Fig. 2: An EDP acquired on the CCD camera from a crotoxin
complex crystal as shown in fig. 1. The EDP shows reflections extending to
3 Å resolution.
The electron dose used for this pattern was 0.02 e/Å2.
Fig. 3: First and the 128th EDP from a tilt series consisting of 128 patterns acquired
from a single crotoxin complex crystal using the CCD camera.
The tilt angle was decremented for each subsequent pattern from +43o to
finally -53.8o. A loss in overall diffracting power due to radiation damage
is visible. The crystal was 128 Å thick. The patterns are shown after computer
processing to remove the radial background due to inelasti scattering.
Movie (1.2 Mb) of a tilt series of 125 diffraction patterns recorded from
a single crystal 256 Å thick. The tilt varied from +50o to
-52o.
Each EDP was processed to remove the background due to inelastic scattering.
Fig. 4: 400 keV spot-scan image of ice-embedded catalase crystal (a)
with corresponding IQ plot (b) showing the strong (with signal-to-noise
ratio of 2 and better) reflections up to the edge of the transform. The image was
taken on the slow-scan CCD camera (Gatan 679) at an effective magnification of
67,000x with 6 e/Å2. The IQ plot shows good potential
for slow-scan CCD imaging of biological structures at high and medium resolution.
Fig. 5: Computed amplitude spectra of carbon film images taken at 400 keV
on slow-scan CCD camera (Gatan, model 679 mark I) and on photographic film (Kodak
SO163) (83 Kb) with corresponding radial profiles showing different amplitude
fall-off towards high resolution. Images were taken from the same specimen area with
the same electron dose and the same microscope settings (illumination and focus). The
difference in the fall-off is attributed to the difference in modulation transfer
functions of the media.
Fig. 6: Phase difference of reflections averaged from 12,000 unit cells from images acquired
using the CCD camera at 400 kV from glucose-embedded bacteriorhodopsin. Only four
reflections showed a phase angle more than 45o different from the
reference data set. Average phase difference was 6.5o.
References:
To Jaap Brink's Home
Page.