Application of scanning transmission electron microscopy

to the study of biological structure

Andreas Engel and Christian Colliex
University of Basel, Basel, Switzerland, and Universit~ Paris Sud, Orsay, France

The scanning transmission electron microscope provides structural and

chemical information of a specimen at atomic-scale resolution and comple-
ments conventional transmission electron microscopy techniques. Mass
measurements can now be performed routinely on a wide range of molecular
and supramolecular structures using elastically scattered electrons. Recent
progress in the acquisition and analysis of electron energy-loss spectroscopy
data indicates that the scanning transmission electron microscope is an
efficient tool for mapping the chemical composition of biological samples.

Current Opinion in Biotechnology 1993, 4:403-411

Introduction freeze-dried biological macromolecules at low dose.

In this review, w e discuss h o w this combination of
An electron microscope visualizes the structural infor- scattering chamber and electron microscope can be
mation carried b y scattered electrons. In a fixed-beam applied to extract quantitative information from a wide
transmission electron microscope, scattered electrons range of biological structures.
emerging from the irradiated sample are collected over
a narrow solid angle and focused by the objective lens
onto the image plane. Here, elastically scattered elec- Quadrupoles Magnetic spectrometer
trons interfere with the unscattered electrons to pro-
duce a phase-contrast image, whereas the inelastically Parallel
scattered electrons generate a diffuse background im- EELS
detector
age that can, in some microscopes, be eliminated by
an energy filter. In the scanning transmission electron
microscope (STEM), which was first introduced in the
late sixties (for an early and illuminating review sum-
marizing the capabilities of STEM, see [1]), the objective
lens focuseslhe b e a m onto an atomic-scale sample vol-
ume. All scattered electrons can then be collected by a
variety of detectors placed behind the specimen (Fig.
1) and their information exploited to the fullest extent.
An image is generated simply by stepping the focused
beam along a sampling grid. Therefore, the image may
be considered as a large collection of individual scat- aperture
tering experiments. As summarized in Table 1, various Objective
lens
types of signals that can be discriminated by scattering
angle and energy-loss transfer convey different struc-
tural and chemical information and may be captured Electron source
simultaneously in different channels. This sequential
and controlled acquisition of information lends itself Fig. 1. Schematic diagram of the dedicated scanning transmission
to quantitative analyses that are difficult to realize with electron microscope. A field-emissiongun emits the electron beam
other instruments. In addition, as there is no limita- that is focused by the objective lens into a sub-nanometer probe on
tion on the solid angle and the energy-loss interval the specimen. Elasticallyscatteredelectrons are collected with one
over which the scattered electrons may be collected, (or several) annular detectors. A magnetic spectrometer analyzes
the velocity of the forward-scatteredelectrons and quadrupolesvary
60-100% of them contribute to the image. This pro- the dispersion of the electron energy-lossspectrum (EELS)measured
vides a unique opportunity to image beam-sensitive on a diode array.

Abbreviations
EELS--electron energy-lossspectrum; PEELS--parallelelectron energy-lossspectrum;
STEM--scanning transmissionelectron microscope/microscopy; TMV--tobacco mosaic virus.

© Current Biology Ltd ISSN 0958-1669 403

404 Protein engineering

Table 1. Types of electron scattering events and information collected by the scanning transmission electron microscope

Type of Relevant Nature of Level of Scattering probability

scattering detector extracted information specificity for a 50 nm organic layer Resolution

Elastic Annular detectors Topographical, struc- Z3/2 20-30% High

tural, projected mass dependence [69] (0.2-0.4nm)

Inelastic (total, i.e. mostly Post-spectrometer Topographical, struc- Z 1/2 30-40% Medium
valence electron excitations) detectors tural, projected mass dependence [70] (1-3 nm)

Inelastic (core-loss edges) Post-spectrometer Chemical Elemental 0.01-0.1% High*

detectors dependence [53] (0.3-0.5 rim)

*For inelastic core-loss edges the resolution is limited by radiation damage as a result of the weakness of this signal.

Imaging with elastically scattered electrons dark-field micrograph of a negatively stained actin fil-
ament displayed in Fig. 2(a). An astonishing level of
The ray-inversion theorem suggests reciprocity be- reproducible fine structure is readily seen in this im-
tween the fixed-beam transmission microscope source age without any processing. The helical reconstruction
and the STEM detectors (Fig. 1). Therefore, the wide- (Fig. 2b) from this projection reveals a close fit with the
angle annular detector of the STEM is equivalent to a atomic-scale model calculated from X-ray data, which
wide angle electron source of the fixed-beam micro- is attributed to the transfer characteristics of the annu-
scope, an incoherent illumination system that abol- lar STEM dark-field mode (Fig. 2c; A Bremer et al.,
ishes phase contrast. Indeed, a rigorous numerical unpublished data). Two novel chaperonins are fur-
simulation demonstrates that the STEM annular de- ther examples: STEM dark-field images of negatively
tector collects an image that is linearly related to the stained preparations have given a clear picture of their
scattering power of the sample and is free of Fresnel structure [4,51.
fringes characteristic for coherent axial bright-field im- Unrivaled images of single heavy atoms have also fos-
ages [2]. The lack of phase contrast, the linearity, and tered the development of heavy-atom labeling tech-
the high 'signal to noise' ratio provided by the annu- niques. A successful search for increasingly powerful
lar detector STEM dark-field m o d e lead to images that heavy-atom clusters suitable for specific labeling (Fig.
are not only easy to interpret [3] but also facilitate dig- 3; [6,7,8"]) has brought progress also with unstained,
ital image processing. This is illustrated by the STEM vitrified samples [9].

Fig. 2 The structure of the F-actin filament. (a) Significant fine structure is discernible on the STEM dark-field micrograph of an uranyl-formate
stained F-actin filament recorded at 2500 electrons nm-2. (b) The high 'signal to noise' ratio enables precise unbending of the filaments
and the extraction of helical information from a single helical repeat. (c) Recent progress in the atomic model of the actin filament allows
the helical reconstruction from the STEM data to be interpreted at atomic-scale resolution (Bremer et al., unpublished data). The scale bar
represents 30 nm. (Provided by courtesy of A Bremer, C Henn and K Goldie.)
 Scanning transmission electron microscopy Engel and Colliex 405

The image-intensity distribution is then directly pro-

portional to the mass distribution of the irradiated
sample [10,11]. Under these conditions, the quantitative
interpretation of the digitally recorded STEM dark-field
images, and thus their use for mass measurement, is
straightforward. Structures seen on these images need
to be contoured, the total number of electrons scattered
from the enclosed areas accumulated, the background
as a result of the supporting film subtracted, and the
result normalized to the recording dose [12"]. The cali-
bration may be calculated from elastic-scattering theory
or determined by the use of a calibration standard, such
Fig. 3. Gallery of STEM micrographs of antibodies labeled with gold as tobacco mosaic virus (TMV). W h e n necessary, mul-
clusters. (a) Shows raw data collected by the annular detector and tiple scattering can be accounted for by Monte Carlo
(b) a gallery of antibodies after noise removal by low-pass filtering.
The scale bars represents 10 nm. (Provided by courtesy of J Hain- simulations [13].
feld.)
This mass-determination method has been developed
into a routine tool that can be applied to a vari-
ety of biological specimens, ranging from relatively
small macromolecules, for example, DNA of mass
Mass measurements in the STEM 2 kDa n m - 1 (Fig. 4a) to supramolecular structures,for
example, mature bacteriophage T4 head of mass
As illustrated in Fig. 1, the annular detector collects 194 MDa (Fig. 4e; [14]). The mass of well defined parti-
elastically scattered electrons diverging at large an- cles [15-31], the mass per length of filaments (Fig. 4a-c)
gles to form the elastic dark-field signal, which is uti- [12", 32-36], and the mass per area of sheet structures
lized for mass determination. Provided the specimen [37,38] have all b e e n determined. From the projected
is thin, that is, of thickness t < A (where A represents mass of cylindrical or spherical particles it is possible
the mean free path between two subsequent scat- to calculate the radial mass profile [39,40], whether the
tering events), multiple scattering may be neglected. sample has been freeze dried or left in a vitrified wa-

Fig. 4. Scanning transmission electron microscope mass measurements can be carried out over masses ranging from a few kilodaltons to
hundreds of megadaltons: (a) double-stranded DNA (2+ 0.7 kDa rim- 1); (b) open double-stranded DNA RecA ATPyS filament (25.6 kDa nm- 1;
by courtesy of JS Wall [35]); (c) compact single-stranded DNA RecA filament (32.0+4.3 kDa nm-1; from [36]); (d) E. coli 50S ribosomal
subunits with mass values in kilodaltons (by courtesy of JS Wall JS); (e) T4 bacteriophage exhibiting a capsid mass of 194+ 2 MDa (from
[14]). Samples (b), (d) and (e) were prepared by freeze drying. Scale bars are 20 nm in (a), 50 nm in (b, d, e) and 100 nm in (c).
406 Proteinengineering

ter layer [41]. In addition, the mass of morphological

domains m a y b e measured [37,42-44,45"].

While the precision and reproducibility of these meas-

urements are comparable with those of the analytical
ultracentrifuge, the possibility to assess mass informa-
tion from domains of supramolecular structures in situ
opens exciting experimental avenues that are not yet
fully exploited. Specific examples are the analysis of
fibrinogen w h e r e the total mass as well as that of
individual domains have b e e n determined [42], the
mass m a p p i n g of the hexagonally packed intermedi-
ate layer of D e i n o c o c c u s radiodurans, which illustrates
that morphological domains of 1.5 kDa can b e m a p p e d
on a polypeptide of 110 kDa [37], and the recent mass
map of the nuclear pore complex, an organelle with a
mass of 125 MDa [44].

To achieve such results, several conditions must be

met. First, the sample solution must be free of non-
volatile salts, detergents, or other small solutes that
easily attach to a biomacromolecule during sample
preparation. Second, low-ionic strength m a y some-
times lead to the disassembly of fragile structures.
This can b e prevented b y glutaraldehyde treatment,
but quantification of the corresponding mass increase
is then required [12",36]. Instead, volatile buffers and
specific preparation protocols are used (see e.g. [42]).
Finally, the instrument must b e well calibrated [12"],
or a mass standard has to be co-prepared with the
Fig. 5. Z-contrast image of septate junctions in the testis of
structure to b e measured. TMV is c o m m o n l y used as Drosophila melanogaster. The regularly packed transmembrane
standard; however, this calibration particle is not free proteins are visible as dark stripes. The sample was aldehyde fixed
from problems if a precision of better than 10 % is to and embedded in Lowicryl HM20 without heavy-metal treatment.
be achieved [12",45"1. The scale bar represents 100 nm. (Provided by courtesy of E Car-
lemalm and E Kellenberger.)

To overcome this problem, a detector configuration is

used that collects every incident electron after its scat-
Multisignal imaging tering, single or multiple, b y the specimen. Instrumen-
tal modifications proposed and realized over the past
Another channel for gathering useful information is few years [49,501 have finally permitted the elimination
the inelastic signal, which is measured at the exit of of thickness effects and the acquisition of 'pure' Z-con-
the magnetic spectrometer (Fig. 1). An adjustable slit trast images. O n e key to the success of this method has
may define the energy w i n d o w to b e used for imaging b e e n the quantitative assessment of the detection effi-
and, in particular, to discriminate the no-loss elec- ciency of all data acquisition channels [51"]. Low-dose
trons from those having suffered an inelastic event. 'pure' Z-contrast images, recorded on the Heidelberg
The total inelastic signal is larger than the elastic sig- cryo &TEM, h a v e recently b e e n processed to evaluate
aal for low atomic n u m b e r elements and it has there- the local concentration of DNA in situ for various chro-
lore b e e n used for mass determination [32]. Continuing matins [52].
i m w o r k of Crewe [1], Carlemalm and Kellenb~i~ger [46]
r o v e demonstrated that the ratio of elastic and inelas-
tic signals could provide reproducible micrographs of
~:.nstained e m b e d d e d cellular material in thin sections Electron energy-loss spectroscopy and elemental
Fig. 5). The theoretical b a c k g r o u n d for the extraction mapping
<~f data such as the concentration of various compo-
nents (resin, lipids, proteins, and nucleic acids) has Instead of considering the inelastic signal as a whole,
b e e n subsequently determined b y different authors the spectral distribution of the forward scattered elec-
[13,47,48]. It appears that more c o m p l e x combinations trons can b e analyzed separately. A highly specific
of signals (also involving no-loss a n d incident electron information lies in the characteristic coreqoss edges
currents) are required to improve the validity of this that are s u p e r p o s e d over a monotonously decreasing
method as s o o n as the specimen thickness is compa- background; their energy position corresponds to a
rable with A. given atomic level and therefore identifies a given el-
 Scanning transmission electron microscopy Engel and Colliex 407

ement within the irradiated volume [53]. The detailed

shape of these edges provides access to an extra le-
vel of information, that is, the valence, the molecular
bonding state, the environment of the excited atom, or
the distance between nearest neighbor atoms. Some of
these possibilities were demonstrated in the early stud-
ies of inner-shell excitation spectra of different nucleic
acid bases [54]. The higher the specificity of the signal,
however, the less sensitive it is. Therefore, the number
of primary electrons required to generate very spe-
cific information with sufficient 'signal to noise' ratio
Counts
'40t
(2nd difference) 120
lOO
Cu L3

increases. The implementation of detector arrays that

collect the energy-loss spectrum in parallel (thus, the 60
4O
acronym PEELS for parallel electron energy-loss spec- 20
troscopy) has greatly improved the efficiency of the 0
-20
method'[55-59]. -40
-60
i I I i i i I
EELS can be used in two ways. In the first approach, 910 920 930 940 950 960 970
the energy-loss spectra are recorded with the beam fo- Energy loss (eV)
cused to an atomic-scale volume element. The data
are then processed to both determine the intensity
of a coreqoss edge and to relate it to a number of
Fig. 6. Dark-field scanning transmission electron microscope mi-
atoms [60] with a sensitivity close to the single-atom crographs of two haemocyanin molecules: (a) at a dose of
level [61]. Thus, in addition to its usefulness as an in- 102electronsnm-2; (b) at a dose of 104electronsnm-2; (c) at a
strument for determining structure through the use of dose of 106electronsnm-2. Magnification in all cases is given
in (a); scale bar=50nm. In the graph below, an EELS second
the annular dark-field signal, the STEM can also pro- difference spectrum from a di-decameric haemocyanin molecule
vide a very local elemental analysis using the PEELS is shown containing 320 copper atoms recorded at a dose of
signal. A macromolecule can first be characterized by 1010 electrons nm- 2 with a multiple Cu L23 least-squares fit (black
its shape and mass, and then specifically b o u n d atoms line) superimposed. (Provided by kind permission of R Leapman
and S Andrews.)
can be identified, localized, and quantified on a scale of
a few nanometers. Figure 6 demonstrates that elements
resistant to mass loss can be measured quantitatively: object area. This has b e e n demonstrated at the level
the c o p p e r signal survives long after the internal struc- of a few millimoles per kilogram dry weight by Leap-
ture of the biomacromolecule has been destroyed. man et al. [65"] for the detection of calcium in specific
cellular compartments on freeze-dried cryosections.
The second way of exploiting the EELS is to pro-
duce images representative of elemental distribution
Figure 7 shows compositional images of five elements
by scanning the probe, recording several energy-fil-
and spectra integrated over well defined pixels for an
tered images on both sides of the core-loss edge, and
air-dried T4 bacteriophage. Although the dose per
processing them pixel b y pixel to display maps of the
pixel was approximately 109 electrons n m - 2, and the
resulting characteristic signal [62]. This technique has
total acquisition time was close to 20min, the spec-
been used for the identification of heavy-atom clusters
(terbium and uranium) as small as ten atoms [63]. imen remains very stable and no noticeable change
of the phosphorous and the nitrogen signals is ob-
With the advent of parallel detectors it has b e c o m e served. This suggests that the two elements cova-
possible to collect and store an entire spectrum for lently integrated in the biological structures can be
each image pixel. Even for small 128x 128 images used as indicators for nucleic acids and phospho-
this results in large data arrays, termed spectrum im- lipids (phosphorous) and proteins (nitrogen). These
ages [64], which require large digital storage and pro- elements remain within an area less than 10nm in
cessing capacities. For both compositional imaging and diameter under the dose required for their identifi-
point analysis they offer several important advantages cation. H o w far is it possible to extend this tech-
with respect to the two approaches described above. nique to the localization of labeled c o m p o u n d s used
The elements to be imaged and the spectrum-process- as biochemical probes? A study of the beam-induced
ing parameters do not n e e d to be k n o w n in advance, fluorine loss for five classes of fluorinated molecules
and b a c k g r o u n d removal and other quantitative proce- has shown that the critical dose may vary over five or-
dures can be applied repeatedly a p o s t e r i o r i until the ders of magnitude, 104-109 electrons n m - 2, depending
best fit is obtained. Elements in unexpected locations on the position of the fluorine atoms and o n the tem-
may be found without deciding initially where to locate perature of the specimen [66]. Consequently, it is likely
the probe for data collection. The sensitivity of this that EELS can detect fluorine-labeled pharmacological
method for detecting w e a k elemental concentrations probes in cryosectioned cells at concentrations below
is high as one can add spectra within a segmented 0.1 moles kg dry weight- 1.
408 Proteinengineering

CK
Photodiode
counts -

400 - \
. XCa L23

200 ~ P k23 ~ o n phage head

0 on carbon foil

I 1 I I I
200 400 600
E n e r g y loss (eV)

Fig. 7. T4 bacteriophage elemental maps of phosphorous (P), carbon (C), calcium (Ca), nitrogen (N) and oxygen (O). They have been extracted
from the intensities of the corresponding core-loss edges (P L23, C K, Ca L23, N K and O K) in a 64x 64 spectrum image. This collection of
4096 spectra was acquired with 4 nm pixel size and 0.3 s pixel-1 dwell time, and corresponds to a recording dose of 109 electrons nm-2,
The two PEELSspectra (bottom right) correspond to the sum of six pixels on top of the phage head and on the thin supporting carbon foil,
respectively. The scale bar (top left) represents 50 nm and was the same for all other elemental maps. (Provided by courtesy of C Mory and
M Tenc~.)

Conclusions methods (An excellent assessment of DNA sequencing

by imaging techniques is provided in [67].)
STEM instruments were introduced in the early sev-
enties and many important potential capabilities were Twenty years later only a few STEMs are in operation
demonstrated rapidly: it was suggested that the STEM around the world for biological research. Yet they pro-
could be used for DNA sequencing. Although pre- duce invaluable data as demonstrated by the present
liminary experiments were encouraging, the research survey of recently published literature. The most pro-
was neglected as result of rapidly developing chemical ductive approach is that of mass measurement, be-
 Scanning transmissionelectron microscopyEngeland Colliex 409

cause the STEM provides the only method available 7. HAINFELDJF, FURUYA FR: A 1 . 4 - n m G o l d C l u s t e r Cova-
for determining the mass of domains within large l e n t l y A t t a c h e d to A n t i b o d i e s I m p r o v e s I m m u n o l a b e l -
supramolecular aggregates. EELS compositional m a p - ing. J Histochem Cytocbem 1992, 40:177-184.
ping constitutes a further step in quantitative electron 8. HAINFELDJF: Site-Specific C l u s t e r Labels. Ultramicroscopy
microscopy, which n o w begins to emerge at an unri- 1992, 46:135-144.
A compact overview o n heavy-atom labels.
valed scale, for m a n y practical situations. Dose limi-
tations will be circumvented b y working on periodic 9. MILLIGANRA, WHITTAKERM, SAFER D: M o l e c u l a r S t r u c t u r e
arrays or by averaging over m a n y particles, exploiting o f F-actin a n d L o c a t i o n o f S u r f a c e B i n d i n g Sites. Nature
1990, 348:217-221.
the simultaneously recorded annular dark-field signal
for their alignment [68]. One can then imagine n e w 10. ENGELA: M o l e c u l a r Weight D e t e r m i n a t i o n b y S c a n n i n g
classes of single-atom labeling such as europium atoms T r a n s m i s s i o n E l e c t r o n M i c r o s c o p y . Ultramicroscopy
1978, 3:273-281.
within cryptate cages. Finally, the field of molecular
mapping, that is, displaying a two-dimensional m a p of 11. WALLJS: M a s s M e a s u r e m e n t s w i t h t h e E l e c t r o n Micro-
the fine structures present in EELS spectra recorded at s c o p e . In Scanning Electron Microscopy, vol 2. Edited by
Johari O. Chicago: SEM Inc; 1979:291-303.
low dose, seems to lie within the realm of capabilities.
12. MULLER SA, GOLDIE KN, BURKI R, HAmNG R, ENGEL
A: F a c t o r s I n f l u e n c i n g t h e P r e c i s i o n o f Q u a n t i t a t i v e
s c a n n i n g T r a n s m i s s i o n E l e c t r o n M i c r o s c o p y . DTtrami-
croscopy 1992, 46:317-334.
Acknowledgements A complete account of STEM mass-determination technique, instru-
mentation and application.

We are indebted to Etienne Delain, Max Haider, Jim Hainfeld, Ed- 13. REICHELTR, ENGEL A: Monte C a r l o C a l c u l a t i o n s o f Elas-
uard Kellenberger, Richard Leapman, Claudie Mory, Shirley Mfiller, tic and Inelastic Electron Scattering i n Biological a n d
Rob Ruigrok, Marcel TencG Jan Tichelaar, a n d Joseph Wall w h o Plastic M a t e r i a l s . Ultramicroscopy 1984, 13:279-294.
all have provided published or u n p u b l i s h e d material and con-
14. BASCHONG W, BASCHONG-PRESCIANOTTO C, ENGEL A,
structive comments. We thank our colleagues Andreas Bremer,
KELLENBERGER E, LUSTIG A, REICHELT R, ZULAUF M, AEBI
Christian H e n n a n d K e n n y Goldie for the dark-held image of
U: M a s s A n a l y s i s o f B a c t e r i o p h a g e T4 P r o h e a d s a n d
a negatively stained F- actin filament a n d its helical reconstruc-
Mature Heads b y S c a n n i n g T r a n s m i s s i o n E l e c t r o n Mi-
tion s h o w n in Fig. 2. The atomic coordinates of an actin trimer
c r o s c o p y a n d H y d r o d y n a m i c Measurements. J Struct
were provided b y Wolfgang Kabsch a n d Ken Holmes. The ex-
Biol 1991, 106:93-101.
pert photographic w o r k by Marlies Zoller and Hedwig Frefel
is gratefully acknowledged. This work w a s supported by Swiss 15. HAMILTON MG, HERSKOVITS TT, FL~CINITI~ PS, WALL JS:
National Science Foundation grant 31-30929.91 to AE, the Mau- Scanning Transmission Electron Microscopic Study of
rice E Mtiller-Foundation of Switzerland a n d the CNRS-Ultimatech M o l l u s c a n H e m o c y a n i n s i n V a r i o u s A g g r e g a t i o n States:
interdisciplinary program. Comparision with Light Scattering Molecular Weights.
J Ultrastruct Mol Struct Res 1989, 102:221-228.
16. HUTCHINSON EG, TICHEIAAR W, HOFHAUS G: Identifica-
tion and Electron Microscopic Analysis of a Chap-
References and recommended reading e r o n i n O l i g o m e r f r o m N e u s o s p o r a crassa M i t o c h o n -
drim EMBO J 1989, 8:1485-1490.
Papers of particular interest, published within the annual period of 17. ARMSTRONGPB, MANGEL WE, WALL JS, HAINFELD JF, VAN
review, have b e e n highlighted as: HOLDE KE, I ~ A, QUIGLEYJP: S t r u c t u r e o f c¢2-Macroglob-
of special interest u l i n f r o m t h e A r t h r o p o d Limulus p o l y p h e m u s . J Biol
•, of outstanding interest Chem 1991, 266:2526-2530.

1. CREWE A: T h e C u r r e n t State o f H i g h R e s o l u t i o n 18. KEDERSHANL, HEUSERJE, CHUGANI DC, ROME KH: Vaults Ill.
S c a n n i n g E l e c t r o n M i c r o s c o p y . Q Rev Biophys 1970, Vault R i b o n u c l e o p r o t e i n P a r t i c l e s O p e n i n t o F l o w e r -
3:137-175. Like S t r u c t u r e s w i t h O c t a g o n a l S y m m e t r y . J Cell Biol
1991, 112:225-235.
2. ENGELA, WIGGINS JW, WOODRUFF DC: A C o m p a r i s o n o f
C a l c u l a t e d I m a g e s b y Six M o d e s o f T r a n s m i s s i o n Elec- 19. KOSTRIKISLG, REISBERG SA, SIMON MN, WALL JS, DAY LA:
t r o n M i c r o s c o p y . J Appl Phys 1974, 45:2739-2747. E x p o r t o f I n f e c t i o u s P a r t i c l e s b y Escberichia coli
T r a n s f e c t e d w i t h t h e RF D N A o f P f l , a Virus o f
3. ENGELA, REICHELTR: I m a g i n g o f Biological S t r u c t u r e s Pseudomonas aeruginosa Strain IL Mol Microbiol 1991,
with the Scanning Transmission Electron Microscope. 5:2641-2647.
J Ultrastruct Mol Struct Res 1984, 88:105-120.
20. MAND1YANV, TUMMINIA S, WALL JS, HAINFELD JF, BOUBLIK
4. TRENTJD, NIMMESGERN E, WALLJS, HARTL FU, HORWICH AL: M: A s s e m b l y o f t h e Escherichta colt 30S R i b o s o m a l
A Molect!!~r C h a p e r o n e f r o m a T h e r m o p h i l i c A r c h a e - Subunit Reveals Protein Dependent Folding of the
b a c t e r i u m is R e l a t e d to t h e E u k a r y o t i c P r o t e i n T - C o m - 16S rRNA D o m a i n s . Proc Natl Acad Sci USA 1991,
p l e x P o l y p e p t i d e - 1 . Nature 1991, 354:490-493. 88:8174-8178.
5. FRYDMAN,J, NIMMESGERN E, ERDJUMENT--BROMAGEH, WALL 21. MASTRANGELOIA, COUREY AJ, WALLJS, JACKSON SP, HOUGH
JS, TEMPST P, HARTL F-U: F t u l c t i o n i n P r o t e i n Folding PVC: DNA L o o p i n g and S p l M u l t i m e r Links: a M e c h -
o f TRiC, a C y t o s o l i e R i n g - C o m p l e x o f TCP-1 and anism for Transcriptional Synergism and Enhance-
S e v e r a l S t r u c t u r a l l y Related S u b u n i t s . EMBO J 1992, m e n t . Proc Natl Acad Sci USA 1991, 88:5670-5674.
11:4767~1778.
22. QABARAN, STERN MS, WALZ DA, CHIU JT, TIMKOVICH R,
6. HAINFELDJF, SPmNZL M, MANDIYAN V, TUMMINIASJ, BOUBLIK WALLJS, KAPP OH, VINOGRADOV SN: H i e r a r c h y o f G l o b i n
M: L o c a l i z a t i o n o f a Specific N u c l e o t i d e i n Y e a s t tRNA C o m p l e x e s - - t h e Q u a t e r n a r y S t r u c t u r e o f t h e Extra-
by Scanning Transmission Electron Microscopy using ceUular C h l o r o c r u o r i n o f Eudistylta-vancouverii. J Mol
a n U n d e c a g o l d Cluster. J Struct Biol 1991, 107:1-5. Biol 1991, 222:1109-1129.
410 Protein engineering

23. THOMASDJ, WALLJS, HA~NFELDJF, KACZOREKM, BOOY FP, i n Equine Herpesvirus 1: Scanning T r a n s m i s s i o n Elec-
TRUS BL, EISERLING FA, STEVENAC: Gp160, t h e Envelope t r o n Microscopic Study. J Virol 1989, 63:3777-3783.
Glycoprotein o f HIV-1 is a Dimer o f 123 kDa Sub-
units Stabilized t h r o u g h Interactions b e t w e e n their 41. TRACHTENBERGS, LEONARDKR, TICHELAARW: Radial Mass
GP41 Domains. Virology 1991, 65:37~7-3803. Density Functions o f Vitrified Helical S p e c i m e n s De-
t e r m i n e d b y Scanning Transmission Electron Mi-
24. TUMMINIASJ, MANDIYANLV, WALLJS, BOUBLIK M: Hetero- c r o s c o p y - - their Potential Use as Substitutes for
geneity o f E s c b e r i c h i a c o U Ribosomes Established by Equatorial Data. Ultramicroscopy 1992, 45:307-321.
Scanning Transmission E l e c t r o n Microscopy. Biocbimie
1991, 73:919-925. 42. MOSESSONMW, HAINFELDJS, WALLJ, HASHEMEYERRH: Iden-
tification and Mass Analysis of H u m a n Fibrinogen
25. VINOGgADOVSN, SHARMPK, QABARAN, WALLJS, WESTRICK Molecules a n d their D o m a i n s b y Scanning Transmis-
JA, SIMMONSJH, GILL SJ: A D o d e c a m e r o f Globin Chains sion Electron Microscopy. J Mol Biol 1981, 153:695-718.
is t h e Principal Functional Subunit o f t h e Extracelluiar
Hemoglobin o f L u m b r i c u s t e r r e s t r i s . J Biol Chem 1991, 43. WALLJS, HAINFELDJF: Mass Mapping w i t h t h e Scanning
266:13091-13096. T r a n s m i s s i o n Electron Microscope. A n n u Rev Biophys
Biophys Chem 1986, 15:355-376.
26. ENGELHARDTH, FORCHHAMMERK, MULLER S, GOLDIE KN,
B6CK A: Structtzre of Selenocysteine S y n t h a s e f r o m 44. REICHELTR, I-IOLZENBURGA, BUHLE EL JR, JARNIK M, ENGEL
E s c h e r i c h i a c o l i and Location o f t h e tRNA in the A, AEBI U: Correlation b e t w e e n Structure a n d Mass
Seryl-tRNAsec-Enzyme C o m p l e x . Mol Microbiol 1992, Distribution o f the Nuclear Pore C o m p l e x and o f
6:3461-3467. Distinct P o r e C o m p l e x C o m p o n e n t s . J Cell Biol 1990,
27. PAREDESAM, SIMONMN, BROWNDT: The Mass o f the Sind- 110:883-894.
bis Virus Nucleocapsid Suggests it has T = 4 Icosahedral 45. SOSINSKYGE, FRANCISNR, DEROSlERDJ, WALLJS, SIMONMN,
Symmetry. Virology 1992, 187:329-332. HAINFELD J: Mass D e t e r m i n a t i o n and Estimation o f
28. PETERSJM, HARRISJR, LUSTIG A, MULLERS, ENGELA, VOLKER Subunit Stoichiometry o f the Bacterial Hook-Basal
S, FRANKEWW: Ubiquitous Soluble Mg2+-ATPase Com- Body Flagellar C o m p l e x o f Salmonella-Typhimurium
p l e x - - a Structural Study. J Mol Biol 1992, 223:557-571. by Scanning T r a n s m i s s i o n Electron Microscopy. Proc
Natl A c a d Sci USA 1992, 89:48014805.
29. P~tLER G, WEINKAUF S, BACHMANN L, M~LLER S, ENGEL A careful analysis of the basal body mass distribution that also ad-
A, HEGERL R, BAUMEISTER W: Subunit Stoichiome- dresses the problem of determining the absolute-mass scale.
try and T h r e e - D i m e n s i o n a l A r r a n g e m e n t i n Protea-
s o m e s f r o m T h e r m o p l a s m a - a c i d o D b i l u m . E M B O J 1992, 46. CARLEMALME, KELLENBERGER E: The Reproducible Ob-
11:1607-1616. servation o f Unstained E m b e d d e d Cellular Material
i n Thin Sections: Visualisation o f an Integral Mem-
30. TUMMINIASJ, HELLMANNW, WALLJS, BOUBLIK M: Image b r a n e P r o t e i n by a New Mode o f Imaging f o r STEM.
Analysis o f Artemia-Salina Ribosomes b y Scanning EMBO J 1982, 1:63~7.
Transmission Electron Microscopy. J Struct Biol 1992,
109:109-115. 47. EGERTONRF: Thickness D e p e n d e n c e o f t h e STEM Ratio-
image. Ultramicroscopy 1982, 10:297-300.
31. TURNQUISTS, SIMON M, EGELMAN E, ANDERSON D: Super-
coiled DNA Wraps M o u n d t h e Bacteriophage Phi-29 48. COLLIEXC, JEANGUILLAUME C, MORY C: U n c o n v e n t i o n a l
Head-Tall Connector. Proc N a a A c a d Sci USA 1992, Modes f o r STEM I m a g i n g of Biological Structures. J
89:10479-10483. Uhrastruct Mol Struct Res 1984, 88:177-206.
32. LAMVlKMK: Muscle Thick Filament Mass Measured b y 49. COLLIEXC, JEANGUILLAUMEC, MORY C, TENC~ M: Progress
Electron Scattering. J Mol Biol 1978, 122:55-68. in Electron Energy Loss Spectroscopic I m ~ , i n g and
33. CROSSRA, ENGELA: Scanning T r a n s m i s s i o n Electron Mi- Analysing Biological S p e c i m e n s w i t h a Field Emission
croscopic Mass D e t e r m i n a t i o n o f in V i t r o Self-Assem- STEM. In Electron Probe Microanalysis. Applications in Bi-
bled S m o o t h Muscle Myosin Filnments. J Mol Biol 1991, ology a n d Medicine, vol 4. Edited by Zierold K, Hagler HK.
222:455458. Berlin-Heidleburg: Springer-Verlag; 1989:99-112.
34. GONGMF, MAKOWSKIL: Helical Structure o f P-Pill f r o m 50. HAIDERM: Filtered Dark-Field and Pure Z-Contrast:
E s c h e r i c h i a - C o l i - - Evidence f r o m X-Ray Fiber Diffrac- Two Novel Imaging Modes in a Scanning T r a n s m i s s i o n
t i o n and S c a n n i n g T r a n s m i s s i o n Electron Microscopy. Electron Microscope. Ultramicroscopy 1989, 28:240-247.
J Mol Btol 1992, 228:735-742.
51. JEANGUILLAUMEC, COLLIEXC, BALLONGUEP, TENCI~ M: New
35. Yu X, EGELMANEH: Structural Data Suggest t h a t the Ac- STEM Multisignal I m a g i n g Modes, Made Accessible
tive and Inactive Forms o f t h e RecA F i l a m e n t are Not t h r o u g h t h e Evaluation o f Detection Efficiences. ul-
Simply Interconvertible. J Mol Biol 1992, 227:334-346. tr'amtcroscopy 1992, 45:205-217.
Simultaneous acquisition and quantitative evaluation of many differ-
36. RUIGROK RV¢'H, BOHRMANN B, HEWAT E, ENGEL A,
ent signals are demonstrated on thin sections and silicon spheres.
KELLENBERGER E, D1CAPUAE: The Inactive F o r m o f RecA
Protein: the 'Compact' Structure. EMBO J 1993, 12:9-16. 52. BOHRMANNB, I{AIDER M, KELLENBERGERE: C o n c e n t r a t i o n
37. ENGELA, BAUMEISTERW, SAXTON WO: Mass Mapping o f Evaluation o f C h r o m a t i n in Unstained Resin-Embed-
a Protein C o m p l e x w i t h t h e Scanning Transmission ded Sections b y Means o f Low Dose Ratio-Contrast
Electron Microscope. Proc Natl A c a d Sci USA 1982, Imaging i n STEM. Ultramicroscopy 1993, 49:235-251.
79:4050~054. 53. COLLIEXC: Electron E n e r g y Loss S p e c t r o s c o p y in the
38. LAMPEPD, KISTLERJ, HEgFI A, BOND J, MLTLLERS, JOHNSON Electron Microscope. A n n N Y A c a d Sci 1984, 485:
RG, ENGEL A: I n Vitro A s s e m b l y o f Gap Junctions. J 311-325.
Struct Biol 1991, 107:281-290. 54. ISAACSONM: Interaction o f 25 keV Electrons w i t h t h e
39. STEVENAC, HAINFELDJF, TRUS BL, STEINERT PM, WALLJS: Nucleic Acid Bases, Adenine, Thymine a n d Uracil - -
Radial Distributions o f D e n s i t y w i t h i n Macromolecu- Outer a n d I n n e r Shell Excitation and Inelastic Cross
lar Complexes D e t e r m i n e d f r o m Dark-Field Electron Section. J Cbem Pbys 1972, 56:1803-1818.
Mierographs. Proc NatI A c a d Sci USA 1984, 81:6363-6367. 55. SHUMANH, CHANG C-F, BUHLE EL, SOMLYO AP: Electron
40. NEWCOMBW3~, BROWNJC, BOOY FP, STEVEN AC: Nucle- Energy-Loss Spectroscopy: Quantitation a n d Imaging.
ocapsid Mass and C a p s o m e r Protein Stoichiometry A n n N Y A c a d Sci 1986, 483:295-310.
 Scanning transmission electron microscopy Engel and Coltiex 411

56. KRIVANEKOL, AHN CC, KEENEY RB: Parallel Detection t i o n e d Cells b y Parallel-EELS Mapping. Ultramicroscopy
E l e c t r o n Spectrometer using Quadrupole Lenses. U/- 1993, 49:225-234.
tramicroscopy 1987, 22:103-116. This work illustrates the practical potential of the spectrum-image
technique for element mapping at low concentration.
57. LEAPMANRD, ANDREWS SB: Biological Electron Energy
Loss Spectroscopy: t h e P r e s e n t a n d t h e Future. Microsc 66. CILIAX BJ, KIRK KL, LEAPMANRD: Radiation Damage o f
Microanalysts Microstructure 1991, 2:387-394. Fluorinated Organic C o m p o u n d s Measured b y Paral-
58. LEAPMANRD, ANDREWS SB: Analysis o f Directly Frozen lel Electron Energy Loss Spectroscopy. Uhramicroscopy
Macromolecules and Tissues in t h e Field-Emission 1993, 48:13-26.
STEM. J Microscopy 1991, 161:3-19.
67. COLE MD, WIGGINSJW, BEER M: Molecular Microscopy o f
59. LEAPMANRE), ANDREWSSB: Characterization o f Biologi- Labeled Polynucleotides: Stability o f O s m i u m Atolns. J
cal Macromolecules b y Combined Mass M a p p i n g and Mol Biol 1977, 117:387--400.
Electron Energy-Loss Spectroscopy. J Microscopy 1992,
165:225-238. 68. REICHELT R, ENGEL A, LEAPMAN R: Low Dose Electron En-
60 LEAPMANRD, ORNBERGRE: Quantitative Electron Energy ergy Loss Spectroscopy by Scanning Transmission
Loss Spectroscopy i n Biology. Ultramicroscopy 1988, Electron Microscopy. In Proceedings of the XItb Inter-
24:251-268. national Congress on Electron Microscopy, Kyoto, vol I.
Edited by Imura T, Maruse S, Suzuki T. Tokyo: Japanese
61. KPdVANEKOL, MORY C, TENCI~ M, COLLIEXC: EELS Quan- Society of Electron Microscopy; 1986:33-34.
tification Near the Single-Atom Detection Level. Microsc
Microanalysis Microstructure 1991, 2:257-267. 69. LANGMORE JP, WALLJS, ISAACSON MS: The Collection o f
Scattered Electrons in Dark Field Electron Microscopy.
62. JEANGUILLAUMEC, TREBBIA P, COLLIEX C: About t h e Use
I. Elastic Scattering. Optik 1973, 38:335-350.
o f E l e c t r o n Energy Loss Spectroscopy for Chemical
M a p p i n g o f Thin Foils w i t h High Spatial Resolution. 70. WALL JS, ISAACSONMS, LANGMOREJP: The Collection o f
Ultramicroscopy 1978, 3:237-242. Scattered Electrons i n Dark Field Electron Microscopy.
63. MORY C, COLLIEX C: Elemental Analysis Near the II. Inelastic Scattering. Optik 1974, 39:359-374.
Single-Atom Detection Level by Processing Sequences
o f Energy-Filtered Images. Uhramicroscopy 1989, 28:
252-257.
64. JEANGUILLAUMEC, COLLIEX C: SpectruLm-Image: t h e Next A Engel, M.E. Mtiller-Institute for High Resolution Electron Mi-
Step in EELS Digital Acquisition and Processing. Ultra- croscopy at the Biocenter, University of Basel, CH 4056 Basel,
microscopy 1989, 28:252-257. Switzerland.
65. LEAPMANRD, HUNTJA, BUCHANANRA, ANDREWSSB: Meas- C Colliex, Laboratoire de Physique des Solides Associ6 au CNRS,
u r e m e n t o f Low Calcium Concentrations i n Cryosec- B~timent 510, Universit6 Paris Sud, 94105 Orsay, Paris, France.