Q 2009 by The International Union of Biochemistry and Molecular Biology BIOCHEMISTRY AND MOLECULAR BIOLOGY EDUCATION

Vol. 37, No. 6, pp. 325–332, 2009

Miniseries: Illustrating the Machinery of Life

Escherichia coli*

Received for publication, August 21, 2009, and in revised form, September 15, 2009

David S. Goodsell‡
From the Department of Molecular Biology, The Scripps Research Institute, La Jolla, California

Diverse biological data may be used to create illustrations of molecules in their cellular context. I
describe the scientific results that support a recent textbook illustration of an Escherichia coli cell. The
image magnifies a portion of the bacterium at one million times, showing the location and form of indi-
vidual macromolecules. Results from biochemistry, electron microscopy, and X-ray crystallography were
used to create the image.
Keywords: Cellular biology, molecular biology, molecular visualization, textbook, diagrams.

‘‘A clear picture of the interior of a living cell that highly dependent on the environmental conditions of the
shows the average distribution of molecules at the cell. I settled on a hybrid approach. I took the concentra-
proper scale, the proper concentration and with no miss- tions of macromolecules from the same sources that I
ing parts, seems to me to be central to the understand- used in the 1991 article. This includes the overall value of
ing of the workings of life.’’ This is how I began my 1991 70% water for the cell, as well as the number of proteins,
article that presented several illustrations of Escherichia RNA, lipids, and other molecules. I also used the same
coli [1]. At the time, there was just enough information to values for the concentrations for the major players in
create a convincing picture of the environment inside liv- protein synthesis, transport, and energy production. I
ing cells. Today, most of the important features of this then added information from two large-scale proteomic
cell have been revealed, although many intriguing mys- studies—the Gene-Protein Database [7] and a GeneChip
teries remain. For the new edition of ‘‘The Machinery of approach that measures levels of mRNA [8]—to identify
Life’’ [2], I updated these illustrations. I found that I was other proteins that are commonly present.
able to create an illustration where every molecule is I created two images of Escherichia coli cells for the
specified, unlike the 1991 images where many generic book. The first illustration, shown in Fig. 1, is an image of
proteins were included as placeholders with the the entire cell. The magnification is such that molecules
expected size and oligomeric state. can barely be seen, but the illustration gives a feeling for
Escherichia coli is arguably the most comprehensively the overall ultrastructure of the cell. The second illustra-
studied organism known to science. A truly daunting tion, shown in Fig. 2, enlarges a portion of the cell so
amount of information is available for constructing an that the shape and size of individual macromolecules
illustration of this cell. This includes information that may be more easily seen. A detailed key to Fig. 2 is
spans the entire range from atoms to cells, including a included in Fig. 3.
full description of the genome, many studies of the pro- Of course, there are many assumptions to be made
teome, ultrastructural studies of entire cells, and atomic when creating illustrations like this. For molecules that
structures of most of the major macromolecules [3]. This are present in high copy numbers, such as tRNA or ribo-
has prompted a number of efforts to create models of somes, I tried to give a representative distribution based
the entire cell and its processes [4–6]. on the concentration. However, many other molecules,
When I started gathering materials for this illustration, I such as DNA polymerase or flagellar motors, are found
had hoped that I could use proteomic studies to provide in much lower concentrations and would only be seen by
an exact recipe for the cell. Unfortunately, this did not chance in a randomly chosen view. I have chosen to
prove to be possible. Many proteomic studies are avail- include many of these in the illustration, however, to give
able, but the protein composition that they reveal is a feeling for the diversity of function that is present in the
cell.
Whenever possible, I based the shape and size of indi-
‡ To whom correspondence should be addressed. E-mail: vidual molecules on atomic structures of the protein from
[email protected].
*This work is supported by the RCSB PDB (NSF DBI 03-
Escherichia coli or related bacteria. These structures
12718), grant DUE 06-18688 from the National Science Founda- were taken from the RCSB Protein Data Bank [9] and are
tion, and the Fondation Scientifique Fourmentin-Guilbert. referenced here using the four-character accession
This paper is available on line at http://www.bambed.org 325 DOI 10.1002/bmb.20345
326 BAMBED, Vol. 37, No. 6, pp. 325–332, 2009

FIG. 1. Cross section through an entire Escherichia coli cell. The cell wall is in green, the cytoplasm is in magenta and blue,
and the nucleoid is in yellow and orange. The boxed region is enlarged in Fig. 2. (magnification, 370,000).
 327

FIG. 2. Cross section through an Escherichia coli cell, showing all macromolecules at a magnification of 31,000,000. At
this magnification, individual atoms are too small to resolve (about the size of a grain of salt). The cell wall is at the top, the cyto-
plasm runs through the middle, and the nucleoid is at the bottom.
328 BAMBED, Vol. 37, No. 6, pp. 325–332, 2009

FIG. 3.
 329
codes. I have not included the primary references to interacts with lipoproteins and other proteins in the outer
these structures, as they are easily accessible at the membrane, forming the primary structural support for the
RCSB PDB WWW site (http://www.pdb.org). cell wall.
The periplasm also includes many small soluble pro-
teins. These include periplasmic binding proteins that
CELL WALL gather small molecules and deliver them to transporters
Escherichia coli cells are surrounded by a complex cell in the cytoplasmic membrane. I have included two exam-
wall composed of two concentric lipid bilayers, the outer ples with their associated transporters: the molybdenum
membrane and the cytoplasmic membrane with a peri- transporter ModA and ModBC (2onk) and the vitamin B12
plasmic space in between. This cell wall plays many transporter BtuF and BtuCD (2qi9). Also included are
functional roles in protection, transport, locomotion, several protective enzymes, including beta-lactamase
sensing, detoxification, and energy production. (2bls), lysozyme inhibitor (1gpq), and a monomeric cop-
Many molecules contribute to the structural role played per-zinc superoxide dismutase (1eso). Two chaperone
by the cell wall [10, 11]. The outer membrane is filled proteins are also included: heat shock protease/chaper-
with lipopolysaccharides—molecules that include a long one DegP (1ky9) and proline isomerase FkpA (1q6u).
polysaccharide chain connected to a complex lipid with The inner membrane is filled with proteins of many dif-
several fatty acid tails. I based the lipid core on the ferent functions. I tried to include a variety of examples,
structure of lipopolysaccharide bound to the outer mem- including proteins from energy production, transport,
brane protein FhuA (1fcp). The polysaccharide chains are peptidoglycan synthesis, sensing, and defense, to give a
heterogeneous, composed of 1–40 copies of a pentasac- cross section of the processes that are occurring. These
charide repeat, each about 15 Å long. There are about include several transporters: magnesium transporters
3.5 million per cell [10]. CorA (2bbj) and MgtE (2yvy), sodium-proton antiporter
Several proteins dominate the outer membrane. These NhaA (1zcd), zinc transporter YiiP (2qfi), a calcium pump
include porin OmpF (2omf), a trimeric protein that forms (1su4), and the two small molecule transporters men-
a pore through the membrane. Two other plentiful pro- tioned earlier. A secretory channel is shown on the right
teins, OmpA and lipoprotein, interact with the peptidogly- side of the image, including the SecY complex in the
can layer. OmpA has a large membrane spanning region membrane and SecA and SecB inside the cell (1rhz,
(1g90) and a periplasmic domain (1r1m). Lipoprotein 2fsg, 1ozb). The large complex that spans both mem-
(1eq7), on the other hand, is tethered on the inner face branes at the left side is a drug efflux pump AcrAB
of the membrane by covalently attached lipids. Estimates (1iwg) and TolC (1ek9).
place 200,000 Omp proteins and 720,000 lipoproteins Several enzymes of energy production are also
per cell [10]. Assuming a surface area of about 6 3 106 included, such as ubiquinol oxidase (1fft), nitrate reduc-
nm3, this places the lipopolysaccharide about 1.3 nm tase NarGHI (1y4z), tricarboxylic acid cycle enzyme
apart, Omp proteins about 9.4 nm apart, and lipoprotein succinate dehydrogenase (1nek), NADH dehydrogenase
(trimers) about 5 nm apart. Also included is an iron trans- (based on results from electron microscopy [12]), and
porter FhuA (1fcp) connected to the inner membrane ATP synthase (1c17, 1e79, 1l2p, 2a7u). A synthetic
through TonB and a fimbrial usher, described later. enzyme is also shown: PBP2 (penicillin binding protein 2,
The space between the membranes, the periplasmic 2olu) is involved in the synthesis of the peptidoglycan
space, is filled with peptidoglycan and a collection of sheath. It is associated with the structural protein MreC,
small soluble proteins. The peptidoglycan appears as a discussed later.
dense layer in electron microscopy, 2–3 nm thick, and Several of the cytoplasmic membrane proteins are
closer to the outer membrane than to the inner mem- involved in sensing. MscL (2oar) is one of several mecha-
brane. It is composed of polysaccharide chains com- nosensitive channels that monitor the internal pressure of
posed of about nine disaccharide repeats (each repeat the cell. On the right side of the illustration, a large as-
approximately 1 nm long), which are then extensively sembly of chemotaxis receptors are modeled after the
crosslinked by short peptides. The resultant network aspartate receptor. These include array of receptors,

FIG. 3. Key to Figure 2. Extracellular: 1, Enterotoxin. Outer membrane: 2, lipopolysaccharide; 3, lipoprotein; 4, porin; 5, OmpA; 6,
fimbrial usher; 7, pilus; 8, iron transport protein FhnA. Periplasm: 9, peptidoglycan; 10, periplasmic binding proteins; 11, beta-lacta-
mase; 12, superoxide dismutase; 13, heat shock protein/chaperone DegP; 14, proline isomerase FkpA. Inner membrane: 15, mag-
nesium transporter MglE; 16, vitamin B12 transporter BtuCD-F; 17, shape-determining proteins MreCD and penicillin-binding pro-
tein PBP2; 18, mechanosensory channel MscL; 19, molybdenum transporter ModBC-A; 20, drug efflux pump AcrAB and TolC; 21,
magnesium transporter CorA; 22, sodium/proton antiporter NhaA; 23, nitrate reductase NarGHI; 24, succinate dehydrogenase; 25,
ATP synthase; 26, ubiquinol oxidase; 27, aspartate receptor; 28, signaling proteins CheAY; 29, secretory channel SecAB; 30, NADH
dehydrogenase; 31, zinc transporter YiiP; 32, calcium pump. Flagellar motor: 33, flagellum; 34, flagellar hook; 35, rotor; 36, motor.
Cytoplasm: 37, cytoskeletal protein MreB; 38, ribosome; 39, transfer RNA; 40, elongation factor Tu; 41, elongation factor Ts; 42,
elongation factor G; 43, initiation factors; 44, aminoacyl-tRNA synthetase; 45, chaperone GroEL; 46, proteasome HslVU; 47, glyco-
lytic enzymes; 48, tricarboxylic acid cycle enzymes; 49, catalase; 50, Iron superoxide dismutase; 51, alkyl hydroperoxide reductase;
52, phosphoenolpyruvate: sugar phosphotransferase system; 53, nucleoside diphosphate kinase; 54, glycerol kinase; 55, acyl car-
rier protein system; 56, aspartate carbamoyltransferase; 57, aspartate aminotransferase; 58, glutamine synthetase. Nucleoid: 59,
DNA; 60, RNA polymerase; 61, messenger RNA; 62, catabolite activator protein; 63, lac repressor; 64, topoisomerase; 65, HU; 66,
H-NS; 67, IHF; 68, Fis; 69, Lrp; 70, condensin MukBEF; 71, RecA; 72, RecBCD; 73, DNA methyltransferase Hha1; 74, DNA glyco-
sylase MutM; 75, DNA polymerase; 76, single strand binding protein.
330 BAMBED, Vol. 37, No. 6, pp. 325–332, 2009
modeled after two structures (1vlt and 2ch7), and the cells, transcription and translation are performed in the
components of the Che system (1b3q), which deliver the same cellular compartment in prokaryotes, and often
messages inside the cell [13, 14]. occur simultaneously. I included one complex at the cen-
ter of the image where ribosomes have started translat-
FLAGELLAR MOTOR AND FIMBRIA ing an mRNA that is in the process of being transcribed.
The flagellar motor is arguably the most impressive RNA polymerase was modeled after the enzyme from
structure of the Escherichia coli cell. Typical cells have 5– Thermus thermophilus (1iw7). Based on information in
10 flagella scattered at random points around the cell. my 1991 article, there would be approximately 6 RNA
The flagellum is composed of roughly 20,000 subunits, polymerase molecules per 100 nm3—I included one in
extending 5–10 lm from the cell surface [15]. It is con- foreground and one in the background of this view.
nected to the motor with a tightly curved hook, which Amazingly, atomic structures are now available for
has roughly 130 subunits with a similar fibril structure as ribosomes in many different states. I used two PDB files
flagellin [16]. The motor is composed of a ring of MotA as the model for this illustration: 1yl3 and 1yl4. I used
and MotB proteins surrounding a rotor composed of several structures for the elongation factors, including
many different proteins. I based the structure on work the tRNA/EF-Tu complex in 1ttt, the EF-Tu/EF-Ts com-
from electron microscopy [17]. plex in 1efu, and EF-G in 1dar. Structures are available
Fimbria is important for the attachment of cells to for all classes of aminoacyl-tRNA synthetases—I used
hosts. Escherichia coli can express several different PDB entries 1asz, 1set, 1ffy, 1gax, 1euq, 1eiy, and 1qf6.
types with different properties. The one shown here is a As in my original TIBS report, I used a concentration of
type I pilus, with a diameter of roughly 2 nm and length 30 ribosomes, 30 aminoacyl-tRNA synthetases, 340
of 1–2 lm, composed of roughly 1,200–2,400 subunits tRNA per 100 nm3. EF-Tu is found to be one of the most
[18]. The tip, not seen in this picture, has specialized plentiful proteins in most proteomics studies [7, 8]—I
subunits involved in adhesion. The fimbriae are extruded included a number roughly equal to the number of tRNA
from the cell through an usher, shown here in the outer molecules. I also included several chaperonin proteins.
cell membrane [19]. The usher assembles new subunits GroEL and GroES show up as being particularly plentiful
to the end of the growing fiber, with the help of a FimC in proteomics studies. I included them based on PDB
chaperone protein (1qun). entry 1aon. A HslVU proteasome is also included from
PDB entry 1e94.
CYTOPLASMIC PROTEINS
Much of the cytoplasmic region of the bacterial cell is DNA AND DNA PACKAGING
filled with molecules of protein synthesis. The remaining Escherichia coli cells do not have a discrete nucleus,
molecules are soluble enzymes performing diverse func- but they do have a loosely defined area at the center
tions. As with my previous TIBS work, I included 130 gly- termed the nucleoid that contains most of the DNA. It is
colytic enzymes and 100 enzymes from the tricarboxylic generally seen to be less densely packed with proteins
acid cycle in each 100 nm3 portion, which leaves about than the surrounding cytoplasm, presumably due to the
100 enzymes of other types in this same volume. I sieve-like exclusion of proteins by the DNA strands. A
included several of the most plentiful enzymes from pro- simple calculation was used to estimate the packing of
teomics studies in this collection, including aspartate car- DNA in the nucleoid. The Escherichia coli K-12 genome
bamoyltransferase (2atc), aspartate aminotransferase has 4,639,221 base pairs [22], which is about 1.5 mm in
(1ase), glycerol kinase (1bot), nucleoside diphosphate ki- length, and a typical growing cell has 2.3 genomes per
nase (2hur), glutamine synthetase (2gls), alkyl hydroper- cell. The volume of the nucleoid is roughly 0.14 lm3 [1].
oxide reductase (1n8j), catalase (1cf9), iron superoxide To get an estimate for the packing of the DNA, we may
dismutase (1isa), and several enzymes of the phosphoe- think of the nucleoid volume as a cube with sides of
nolpyruvate: sugar phosphotransferase system (2hwg, about 0.52 lm, then cut the DNA into 0.52 lm lengths
1cm3, 1wcr) and the acyl carrier protein system (1dd8, and stack them uniformly in the cube. From this approxi-
1i01). mation, we obtain a spacing of about 6.4 nm between
There is also increasing evidence that bacterial cells DNA strands.
have a significant cytoskeleton [20, 21]. I included the To calculate the number of repressors that we might
actin-like MreB protein (1jce), which is thought to form a expect to see in this slice, I assumed that most operons
large helical assembly inside the cytoplasmic membrane. were about the size of the lac operon. It is about 5,000
MreC (2qf4) links this filament to the membrane and may base pairs long or 1,700 nm. So, if our window is a 100
be important for control of cell shape. nm2, we might expect to see one repressor every 17
strands that pass through the view. I included one com-
PROTEIN SYNTHESIS plex of lac repressor and catabolite activator protein in
Protein synthesis is one of the major tasks of the Esch- the foreground of this illustration, and we might expect
erichia coli cell, which is reflected in the composition of to see one or two more in the background. The looped
the cell: over a third of the molecules are involved in one structure of the DNA is taken from the crystal structure
way or another with production of proteins. Remarkably, analysis [23].
atomic structures are available for most of the major mol- The bacterial genome is packaged by a variety of pro-
ecules involved in protein synthesis. Unlike in eukaryotic teins that bend DNA and bridge neighboring portions of
 331
the strand. Abundances for the DNA packaging proteins way that is true to the actual three-dimensional geometry
were taken from a review of the subject [24]: 50,000 his- of the object. In this illustration, I purposefully aligned all
tone-like protein HU, 20,000 H-NS (histone-like nucleoid of the DNA strands and the peptidoglycan strands
structuring protein), 15,000 IHF (integration host factor), approximately in the plane of the page, so that they
and 100,000 Fis (factor for inversion stimulation pro- would not be clipped. In reality, of course, we would
tein)—all values are the number of molecules per cell. expect a more tangled structure, with chains at all orien-
Assuming that there are 2.3 genomes per cell, a protein tations. The section was chosen perpendicular to the
with 10,000 copies per cell would be spaced about membrane, so show the cross section through the entire
1,000 base pairs, or 340 nm, apart. The structures of cell wall. This unfortunately hides the continuous planar
these proteins were taken from a recent review [25] and character of the membrane, but is necessary to reveal
several PDB entries (3fis, 1p71, 1lr1). I also include one the different compartments.
Lrp (leucine-responsive regulatory protein) in the view I chose the overall view at the rounded end of the cell
(2gqq), as well as a large star-shaped complex of con- to give a feeling for the finite size of the cell. If I had cho-
densin MukBEF [26]. sen a view along one of the straight sides, there would
I included two types of topoisomerases: a type I topo- not be any hints about the diameter of the cell. Of
isomerase next to a transcribing RNA polymerase (1i7d) course, the relative size of molecules and cells is appa-
and a DNA gyrase caught in the middle of passing one rent in Fig. 1, which shows the whole cell, but I also
DNA strand through another (1bgw, 1ei1). I also included wanted to capture this relationship in the enlarged cross-
several molecules involved in DNA repair. These include section. Figure 2 is a water color painting, created at
a long helical complex of RecA (2reb) and two RecBCD twice the printed size. Figure 1 was created digitally
complexes acting on a DNA break (1w36). Also included entirely in Photoshop.
are enzymes that correct damaged bases, such as the
DNA glycosylase MutM (1kfv). Acknowledgment— This is manuscript 20312 from the
Scripps Research Institute.
At the center of the nucleoid region, I included a repli-
cation fork. The DNA polymerase complex is based on REFERENCES
several structures of the isolated components [27]—but [1] D. S. Goodsell (1991) Inside a living cell, Trends Biochem. Sci. 16,
the overall association of the components is still a matter 203–206.
[2] D. S. Goodsell (2009) The Machinery of Life, 2nd ed., Springer,
of debate. The transient single-stranded regions are New York.
bound to SSB (single stranded DNA binding protein), [3] M. Schaechter (2001) Escherichia coli and Salmonella 2000: The
modeled after the structure of the DNA-binding domain view from here, Microbiol. Mol. Biol. Rev. 65, 119–130.
[4] N. Nanninga (1998) Morphogenesis of Escherichia coli, Microbiol.
of the protein (1eyg). Mol. Biol. Rev. 62, 110–129.
[5] D. M. Morris, G. J. Jensen (2008) Toward a biomechanical under-
standing of whole bacterial cells, Annu. Rev. Biochem. 77, 583–613.
[6] I. M. Keseler, C. Bonavides-Martinez, J. Collado-Vides, S. Gama-
Castro, R. P. Gunsalus, D. A. Johnson, M. Krummenacker, L. M.
PEDAGOGIC AND ESTHETIC CONSIDERATIONS
Nolan, S. Paley, I. T. Paulsen, M. Peralta-Gil, A. Santos-Zavaleta, A.
I wanted to capture several concepts in this illustration: G. Shearer, P. D. Karp (2009) EcoCyc: A comprehensive view of
Escherichia coli biology, Nucleic Acids Res. 37, D464–470.
that the cell is a crowded environment, that many differ- [7] R. A. VanBogelen, K. Z. Abshire, B. Moldover, E. R. Olson, F. C.
ent processes are intermingled and occur simultaneously, Neidhardt (1997) Escherichia coli proteome analysis using the gene-
and that there is compartmentation even in bacterial protein database, Electrophoresis 18, 1243–1251.
[8] R. W. Corbin, O. Paliy, F. Yang, J. Shabanowitz, M. Platt, C. E. , Jr.,
cells. To highlight compartmentation, I chose a coloring Lyons, K. Root, J. McAuliffe, M. I. Jordan, S. Kustu, E. Soupene, D.
scheme that separated each compartment. The two F. Hunt (2003) Toward a protein profile of Escherichia coli: Compari-
membranes are colored greens, with the periplasmic son to its transcription profile, Proc. Natl. Acad. Sci. USA 100,
9232–9237.
space in turquoise. The cytoplasm is in blues and pur- [9] H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T. N. Bhat, H.
ples, and the nucleoid is in yellows and oranges. This Weissig, I. N. Shindyalov, P. E. Bourne (2000) The protein data
bank, Nucleic Acids Res. 28, 235–242.
coloring scheme highlights the physical compartmenta- [10] H. Nikaido (1996) Outer Membrane. Escherichia coli and Salmo-
tion—the separation of spaces by membranes and the nella: Cellular and Molecular Biology, F. C. Neidhardt, Ed., ASM
sieve-like separation provided by the DNA—but it hides Press, Washington, DC, pp. 29–47.
[11] H. Nikaido (2003) Molecular basis of bacterial outer membrane per-
the functional localization, for instance, the fact that the meability revisited, Microbiol. Mol. Biol. Rev. 67, 593–656.
cytoplasmic membrane plays multiple roles in energy [12] T. Friedrich, B. Bottcher (2004) The gross structure of the respira-
production, nutrient transport, and signaling. To show tory complex I: A Lego System, Biochim. Biophys. Acta 1608, 1–9.
[13] G. H. Wadhams, J. P. Armitage (2004) Making sense of it all: Bacte-
this intermingled function, I created a series of ‘‘keys’’ for rial chemotaxis, Nat Rev Mol Cell Biol 5, 1024–1037.
the book that isolated just the molecules involved in a [14] G. L. Hazelbauer, J. J. Falke, J. S. Parkinson (2008) Bacterial che-
particular function. These keys are similar to Fig. 3, but moreceptors: High-performance signaling in networked arrays,
Trends Biochem. Sci. 33, 9–19.
show the molecules of a particular function in color and [15] K. Yonekura, S. Maki-Yonekura, K. Namba (2003) Complete atomic
the rest in gray tones. As the reader progresses through model of the bacterial flagellar filament by electron cryomicroscopy,
the chapter, they can compare several of these keys and Nature 424, 643–650.
[16] F. A. Samatey, H. Matsunami, K. Imada, S. Nagashima, T. R. Shaikh,
explore the mixture of different functions occurring in D. R. Thomas, J. Z. Chen, D. J. Derosier, A. Kitao, K. Namba (2004)
each space. Structure of the bacterial flagellar hook and implication for the molec-
There is also a healthy dose of artistic license applied ular universal joint mechanism, Nature 431, 1062–1068.
[17] G. E. Murphy, J. R. Leadbetter, G. J. Jensen (2006) In situ structure
to this type of illustration. In particular, it is always a chal- of the complete Treponema primitia flagellar motor, Nature 442,
lenge to represent fibrous and planar components in a 1062–1064.
332 BAMBED, Vol. 37, No. 6, pp. 325–332, 2009
[18] G. Capitani, O. Eidam, R. Glockshuber, M. G. Grutter (2006) Struc- [23] M. Lewis, G. Chang, N. C. Horton, M. A. Kercher, H. C. Pace, M. A.
tural and functional insights into the assembly of type 1 pili from Schumacher, R. G. Brennan, P. Lu (1996) Crystal structure of the
Escherichia coli, Microbes Infect. 8, 2284–2290. lactose operon repressor and its complexes with DNA and inducer,
[19] H. Li, L. Qian, Z. Chen, D. Thibault, G. Liu, T. Liu, D. G. Thanassi Science 271, 1247–1254.
(2004) The outer membrane usher forms a twin-pore secretion com- [24] J. Stavans, A. Oppenheim (2006) DNA-protein interactions and bac-
plex, J. Mol. Biol. 344, 1397–1407. terial chromosome architecture, Phys. Biol. 3, R1–R10.
[20] Y. L. Shih, L. Rothfield (2006) The bacterial cytoskeleton, Microbiol. [25] M. S. Luijsterburg, M. C. Noom, G. J. Wuite, R. T. Dame (2006) The
Mol. Biol. Rev. 70, 729–754. architectural role of nucleoid-associated proteins in the organization
[21] K. A. Michie, J. Lowe (2006) Dynamic filaments of the bacterial cy- of bacterial chromatin: A molecular perspective, J. Struct. Biol. 156,
toskeleton, Annu. Rev. Biochem. 75, 467–492. 262–272.
[22] F. R. Blattner, G., III, Plunkett, C. A. Bloch, N. T. Perna, V. Burland, [26] R. B. Case, Y. P. Chang, S. B. Smith, J. Gore, N. R. Cozzarelli, C.
M. Riley, J. Collado-Vides, J. D. Glasner, C. K. Rode, G. F. Mayhew, Bustamante (2004) The bacterial condensin MukBEF compacts
J. Gregor, N. W. Davis, H. A. Kirkpatrick, M. A. Goeden, D. J. Rose, DNA into a repetitive, stable structure, Science 305, 222–227.
B. Mau, Y. Shao (1997) The complete genome sequence of Esche- [27] J. L. Keck, J. M. Berger (2000) DNA replication at high resolution,
richia coli K-12, Science 277, 1453–1474. Chem. Biol. 7, R63–R71.