вход по аккаунту


Escherichia colis How-To Guide for Forming Amyloid

код для вставки
Escherichia coli’s How-To Guide
for Forming Amyloid
Exploring how bacteria form and deploy amyloids might yield new ideas
for untangling the damage and diseases associated with them
M. R. Chapman, L. S. Robinson, and S. J. Hultgren
xamining a lineup card prior to an
eases. Abundant genetic and in vitro cell culture
afternoon game of baseball at Wrigdata indicate that amyloid fibers or the process
ley Field provides all of the informaby which they form contributes to cell death and
tion necessary to ascertain who will
tissue damage, providing evidence that such fibe playing in the game. However,
bers are a significant factor in the pathology of
from the lineup card, one cannot predict the
these diseases.
outcome of the day’s game or the wonders and
Understanding how soluble proteins become
marvels of enjoying a day of baseball at the
amyloid fibers could lead to the development of
oldest ballpark in America. Protein folding is
preventive treatments or therapies for these dismuch the same—simply undereases. One approach to developing
standing who the players are does
a better fundamental understandnot guarantee that one can predict
ing of these biochemical processes
When protein
how that protein might fold.
involves studying amyloid fibers
folding goes
Although the information reassembled naturally by Escheawry, proteins
quired for folding can be found in
richia coli and Salmonella. Using
can adopt
that protein’s amino acid sethese easily manipulable bacterial
quence, many factors, including
systems, it is possible to study
pH, ionic strength, temperature,
amyloid fiber formation with
and the presence of other proteins,
greater precision, and perhaps deforms that are
influence folding. Unfortunately,
rive paradigms that apply to other
harmful to the
these variables make it nearly imamyloid proteins. Additionally,
cell and to the
possible to predict how a linear
these bacterially produced amysurrounding
peptide folds into its final threeloid fibers may mediate biofilm
dimensional shape.
formation and could be important
When protein folding goes
disease determinants. Hence, a
awry, proteins can adopt nonnabetter understanding of how they
also help toward alleviating a
tive, aggregative forms that are harmful to the
diseases attributed to enteric
cell and to the surrounding tissue. Rudolph Virbacteria, including pneumonia, meningitis, diarchow observed such deposits in the 1850s while
rheal diseases, pyelonephritis, and cystitis.
staining diseased brain tissues. He coined the
term “amyloid” (Latin for starch-like) to deMorphology and Biophysical
scribe them because these deposits stained
Traits of Amyloids
strongly with iodine. For nearly a century, amyloid deposits were granted little clinical considAmyloidogenic proteins display no obvious simeration, but are now being recognized as the
ilarities to one another in either sequence or
hallmark of many diseases, including Alzheinative structure. However, the fibers they form
mer’s (AD), Parkinson’s, type 2 diabetes, sysshare several distinctive morphological and biotemic amyloidosis, and the prion-related disphysical characteristics. By electron microscopy,
M. R. Chapman is
is a postdoctoral
fellow, L. S. Robinson is a graduate
student, and S. J.
Hultgren is the
Helen L. Stoever
Professor of Molecular Microbiology in
the Department of
Molecular Microbiology and Microbial
Washington University School of Medicine, St. Louis, Mo.
Volume 69, Number 3, 2003 / ASM News Y 121
proteins, destabilizing the native fold,
for example by mutation, heat, or
chemical denaturants, is required to
initiate the process. Presumably, partially denatured globular proteins expose interactive surfaces, thus promoting fiber assembly. Soluble amyloid
precursors are often normal constituents of human biological fluids. A key
question in amyloid pathogenesis is
why these proteins are soluble under
some conditions but aggregate under
Several Gram-Negative Bacteria
Form Amyloids, called Curli
E. coli and Salmonella species assemble extracellular 4 – 6 nm-wide amyloid fibers, called curli. These fibers
form a tangled extracellular matrix
that can connect several neighboring
bacterial cells into small groups (Fig.
1). Curli can also bind host proteins
High-resolution deep-etch EM micrographs of curliated E. coli. (A-C). Curli fibers appear to
and might influence host immune reengulf and surround the bacteria, forming a meshwork that connects bacteria together.
The inset in panel B is shown enlarged in panel C. (D) Amyloid fibers formed by the in vitro
sponses. Like eukaryotic amyloid fipolymerization of purified CsgA. Scale bars are 200 nm.
bers, curli resist protease digestion, remain insoluble when boiled in 1%
sodium dodecyl sulfate, and bind to
dyes, including CR and ThT.
amyloids consist of straight, unbranched fibers
in E. coli are dedicated to
that are 5–13 nm in diameter. X-ray diffraction
cell surface (Fig. 2).
analysis reveals that these fibers are rich in
by the divergently
вђ¤-strand secondary structure, with an unusual
cross-вђ¤ conformation in which the strands run
perpendicular to the fiber axis while the sheets
run parallel to the fiber axis.
cues, including osmolarity, temperature, growth
Amyloid fibers bind Congo red (CR) and inphase, and protein aggregation in the periplasm.
duce a “red shift” with a maximum absorbance
The csgBA operon encodes two homologous
difference between CR alone and CR bound to
proteins (CsgA and CsgB) that are secreted into
amyloid at 541 nM. Fibers bound to Congo red
the extracellular environment. The major com(CR) also display red-green birefringence under
ponent of E. coli curli is the 13-kDa CsgA procross-polarized light. Amyloids can also bind
tein. CsgB, the minor curli subunit, is required
thioflavin T (ThT), resist proteases, and withfor CsgA polymerization at the cell surface.
stand depolymerization by detergents.
In the absence of the CsgB nucleator (B-/AП©),
Unlike the вђ¤-sheet rich fibers, amyloid proteins in their native conformation may be either
CsgA is secreted from the cell in a soluble, unasunstructured or contain varying proportions of
sembled state. This soluble CsgA can polymerize
вђЈ-helices and вђ¤-strands. Because non-diseaseinto curli fibers if it contacts an adjacent cell
related proteins such as myoglobin and
expressing the CsgB nucleator, and not CsgA
acylphosphatase can form amyloid fibers under
(Bϩ/A–), in a process called interbacterial
appropriate conditions, some researchers becomplementation (Fig. 3). We propose that
lieve that amyloid formation is a general propCsgB induces a conformational change in CsgA
erty of polypeptides. In the case of globular
that nucleates its assembly into fibers. Polymer-
122 Y ASM News / Volume 68, Number 3, 2003
Integrating Teaching and Research
Scott Hultgren hopes that his
team’s work in understanding
how amyloid fibers are formed in
bacteria will lead to a better grasp
of the process in humans, insights
that could point to preventing or
treating a wide range of devastating brain-wasting diseases. “It’s a
beautiful model to study the molecular details of protein-protein
interaction necessary for amyloid
formation,” he says. “We’ve discovered that E. coli bacteria have
an assembly machine—a whole
gene cluster— dedicated to amyloid assembly.”
Certain illnesses affecting the
central nervous system, particularly Alzheimer’s disease and
Creutzfeldt-Jakob disease, are
characterized by clusters of such
fibers being formed throughout
the brain. Hultgren and his colleagues at the Washington University School of Medicine are
studying similar fiber clusters that
are associated with enteric bacteria, including E. coli and Salmonella. The development of these
bacterial fibers likely is no accident, according to Hultgren.
“These fibers may play an important role in human disease,” he
says. “If we have a molecular
snapshot of the subunit interactions during [bacterial] amyloid
assembly, we should be able to
design compounds that would interrupt this assembly.”
Hultgren, 43, is the Helen L.
Stoever professor of molecular
microbiology at the Washington
University School of Medicine.
He has been there since 1989, after finishing his postdoctoral
training at Umea University in
Umea, Sweden. Earlier, he started
as a chemistry major at Indiana
University in Bloomington, but
later switched to microbiology
and completed his Ph.D. in 1988
at the Northwestern University
medical school campus in Chicago, Ill.
Hultgren grew up in Michigan
City, Ind., a small town where
almost everyone went to the same
high school and many were
taught by the same teachers—
among them his father, who
taught high school chemistry. The
elder Hultgren introduced his son
to science. “He was the best
teacher I ever had,” Hultgren says
of his father, who won numerous
teaching awards. “He was so
good that my first year in college
chemistry was a total review, because I’d already had everything
from him.”
This lasting appreciation for his
father’s prowess as a teacher is
probably one of the reasons why
Hultgren today is willing to push
the boundaries in his own teaching, an attitude unusual in a competitive environment like that
found at Washington University
School of Medicine and similar
academic settings, where research
tends to be more readily rewarded
than is teaching. “The influence of
a good teacher can make such a
difference in somebody’s life,” he
says. “It can be the spark that
sends you in a certain direction.”
At the medical school, Hultgren
recently revamped a microbiology
course for medical students, with
the goal of putting disease pathogenesis into a framework that begins with describing symptoms
and follows with detailed information providing an understanding of the microorganisms responsible. The idea is to give medical
students the entire picture of infectious diseases, enabling them
to recognize telltale signs from the
moment a patient walks into the
clinic. He describes this approach
as “a conceptual course in microbial pathogenesis, taught by experts in the field.
“What we teach in this course is
the molecular basis of the microbe, but we try to integrate that
into a framework that relates all
the way back to when someone
comes into the clinic with a stomachache,” Hultgren explains. Every lecture lasts two hours. “We
have a clinician give a talk on the
clinical syndrome during the first
hour,” he says. “Then the next
hour is a basic scientist who does
research in that area, so that medical students can learn about the
microbe itself and the mechanism
by which it causes these symptoms.
“In the medical community, the
more you understand the whole
framework, the better physician
you will be,” Hultgren says.
“When you hear a symptom, you
need to be able to work it all the
way back through your mind to
the microbe itself. That will make
you a better physician—and that’s
our philosophy.”
In some ways, this approach to
teaching medical students about
microbial pathogens parallels
what microbiologists need to
learn as they attempt to design
new drugs and vaccines for a particular infectious disease. They
first need to understand the nature of the responsible organism
before they can figure out ways to
treat or prevent the disease that it
Marlene Cimons
Marlene Cimons is a freelance writer
in Bethesda, Md.
Volume 68, Number 3, 2003 / ASM News Y 123
Many amyloid proteins assemble by
nucleation-dependent polymerization,
where the rate-limiting step is nucleus
formation. When the monomer concentration is only slightly higher than
the critical concentration (the concentration below which polymerization
cannot take place), nucleus formation
occurs slowly due to the entropic cost
of subunit association. However, once
an ordered nucleus is assembled, further subunit addition becomes favorable, and polymerization proceeds rapidly. The concentrations required for
nucleation of amyloid proteins in vitro
often far exceed their observed physiological concentrations, raising the
question as to how amyloid formation
occurs in vivo. Detailed analysis of the
process by which CsgB converts CsgA
from soluble to fibrous protein may
provide insight into how amyloid proteins are nucleated in humans.
The Curli Assembly Machinery
Organelles assemble on the surfaces of
gram-negative bacteria in multiple
steps involving both protein synthesis
and subunit secretion across two membranes. The type II secretion machinery is responsible for transporting curli
subunits across the inner membrane,
but the mechanism of secretion across the outer
membrane is not known.
Other bacterial secretion systems, such as the
chaperone-usher pilus assembly pathway, feature an outer membrane-localized pore-forming
protein that serves dually as an assembly platform and as the subunit secretion apparatus.
CsgG, a 30-kDa lipoprotein that localizes to the
periplasmic face of the outer membrane, may
perform an analogous task during curli biogenesis.
Transposon insertions in csgG abolish curli
production. Furthermore, csgG mutants cannot
donate CsgA subunits or act as acceptors during
interbacterial complementation, indicating that
CsgG is not only required for fiber formation
but also for nucleation. We recently found that
when CsgA and CsgG are coexpressed in a
strain lacking the other curli proteins, CsgA is
secreted into the extracellular space in a CsgG-
Current model of curli regulation and assembly. All curli subunits (excluding CsgD) have
signal sequences (indicated in yellow) for translocation across the inner membrane. The
major subunit protein (CsgA) and the nucleator (CsgB) are secreted to the cell surface in
a CsgG-dependent fashion. CsgE is a periplasmic protein required for the stability of CsgA.
CsgB nucleator activity is drastically reduced without CsgF.
ized CsgA would then induce a similar change in
the next incoming soluble CsgA subunit, and
repetition of this process would drive curli assembly. Such a model assumes that conformationally altered CsgA would have a nucleator
activity similar to that of CsgB. Consistent with
this idea, CsgA and CsgB are proteins of identical predicted size, are built up of similar repeat
motifs, and exhibit 49% sequence similarity.
CsgA can be purified in a soluble, unassembled state from cells lacking CsgB. Soluble
CsgA does not adopt the вђ¤-strand-rich structure
or fibrous appearance that typify curli. However, after prolonged incubation, purified CsgA
spontaneously assembles into fibers (Fig. 1D)
that bind CR, demonstrating that under favorable conditions CsgA can guide its own conversion into amyloid fibers and that CsgB is not
absolutely required for nucleation and polymerization of CsgA.
124 Y ASM News / Volume 68, Number 3, 2003
dependent manner. However, CsgA
does not accumulate within the cell
when CsgG is absent. In fact, both
CsgA and CsgB are rapidly degraded in
the absence of CsgG. Conversely, the
steady-state levels of polymerized
CsgA and CsgB exceed wild-type levels
when CsgG is overexpressed from an
inducible promoter.
Taken together, these results suggest
that CsgG might bind to curli subunits
to stabilize them in the periplasm. Alternatively, CsgG might form pores
through which subunits escape the
periplasm, where they are unstable,
into the extracellular milieu, where
they can assemble into durable fibers.
The expression of at least one periplasmic protease, DegP, increases when
CsgA and CsgB subunits are present.
The ability of the bacteria to sense and
respond to a buildup of curli subunits
might keep CsgA from polymerizing
within the cell.
CsgE and CsgF
Interbacterial complementation of CsgA polymerization. Cells lacking the CsgB nucleator
Wild-type E. coli stains red after 30
(donor) will secrete CsgA molecules that can be assembled on the surface of bacteria
expressing the CsgB nucleator (acceptor). The inset depicts a donor (top) and acceptor
hours of growth on plates containing
strain (bottom) grown adjacent to each other on YESCA plates amended with CR. The
CR, indicating that these bacteria are
zone of CR binding on the recipient strain corresponds with CsgA polymerization.
producing curli. In the absence of
CsgF, however, the bacteria remain
unstained after 30 hours of growth and
stain only light red after 48 hours, suggesting
in an assembly-competent state. Furthermore,
that some curli are being assembled. Indeed, we
neither csgF– nor csgF–B– mutants are able to
can detect fibers by EM, and, although less
accept CsgA from a csgB– (CsgA-donating)
abundant, they are otherwise indistinguishable
strain. This result confirms a nucleation defect,
from those produced by wild-type bacteria. In
leading us to believe that CsgF may function as a
the absence of CsgF, the majority of CsgA is
molecular chaperone for CsgB. We are currently
released into the extracellular environment as
investigating CsgB localization and steady-state
soluble unpolymerized subunits.
levels in csgFПЄ mutants.
We can account for the nucleation defect of
A nonpolar csgE- deletion mutant produces
csgF mutants in one of two ways. First, CsgF
non-CR-binding colonies akin to strains harbormight be required for full CsgB activity, possibly
ing a mutation in the curli subunit gene, csgA.
by stabilizing or correctly positioning CsgB, or
Despite the pale colony phenotype, this csgEПЄ
by facilitating the folding of a nucleation commutant produces extracellular fibers that depetent CsgB molecule. Second, CsgF could modpend upon an intact csgBA operon and react
ify CsgA, making it polymerization-competent.
with CsgA antibodies. CsgE- fibers are less
We favor the former possibility because CsgA
abundant than wild-type curli, and they tend to
secreted from a csgF– mutant can be assembled
arrange into ring-like shapes, whereas wild-type
on the surface of a CsgB acceptor cell, indicating
curli are invariably straight and rigid.
The total amount of CsgA is dramatically
that even in the absence of CsgF, CsgA remains
Volume 68, Number 3, 2003 / ASM News Y 125
reduced in csgE– mutant bacteria, and a csgE- or
a csgB-E- double mutant is unable to donate
CsgA subunits for assembly on CsgB acceptor
cells. CsgB is also reduced in csgEПЄ mutant
bacteria, but these cells can still act as an acceptor (albeit less well than CsgEП© strains), suggesting that CsgB is at least partly functional in the
absence of CsgE. CsgE could act directly on the
CsgB and CsgA proteins, or it might contribute
to CsgA and CsgB stability indirectly by interacting with CsgG.
Neither CsgE nor CsgF display (significant)
homology to other proteins in the National Center for Biotechnology Information database. We
are currently conducting additional experiments
to better understand the structure and function
of CsgE and CsgF.
Host-Pathogen Interaction
and Amyloid Formation
Eukaryotic amyloids are generally thought to
represent a biological accident—the result of
proteins adopting nonnative, yet stably folded
structures. By contrast, curli formation is not a
mistake, but the result of an elaborate and finely
tuned assembly system.
Why do bacteria produce amyloid fibers?
Curli apparently are important mediators of
biofilm formation and could also be important
disease determinants. Curli bind eukaryotic extracellular matrix proteins, such as fibronectin
and laminin. Moreover, human macrophages
recognize and respond better to curliated bacteria than to noncurliated strains, and the mouse
immune response to curliated strains is similarly
more robust.
Bacterial amyloids could also play a direct
role in certain human neurodegenerative and
amyloid-related diseases. During an infection,
bacterial curli would encounter many different
types of host proteins, including some that
might be predisposed to form amyloid. Because
amyloid fibers of one type can seed amyloid
development of another type, curli might induce
host proteins to form pathogenic amyloid fibers.
Such cross-species nucleation has been demonstrated among prions from different genera of
yeast. Exploring such ideas might lead to new
ways for combatting amyloids and the damage
and diseases associated with them.
Chapman, M. R., L. S. Robinson, J. S. Pinkner, R. Roth, J. Heuser, M. Hammar, S. Normark, and S. J. Hultgren. 2002. Role
of Escherichia coli curli operons in directing amyloid fiber formation. Science 295:851– 855.
Collinson, S. K., J. M. Parker, R. S. Hodges, and W. W. Kay. 1999 Structural predictions of AgfA, the insoluble fimbrial
subunit of Salmonella thin aggregative fimbriae. J. Mol. Biol. 290:741–756.
Hammar, M., Z. Bian, and S. Normark. 1996. Nucleator-dependent intercellular assembly of adhesive curli organelles in
Escherichia coli. Proc. Natl. Acad. Sci. USA 93:6562– 6566.
Loferer, H., M. Hammar, and S. Normark. 1997. Availability of the fiber subunit CsgA and the nucleator protein CsgB during
assembly of fibronectin-binding curli is limited by the intracellular concentration of the novel lipoprotein CsgG. Mol.
Microbiol. 1:11–23.
Sunde, M., and C. Blake. 1997. The structure of amyloid fibrils by electron microscopy and X-ray diffraction. Adv. Protein.
Chem. 50:123–159.
True, H. L., and S. L. Lindquist. 2000. A yeast prion provides a mechanism for genetic variation and phenotypic diversity.
Nature 404: 477– 483.
Sauer, F. G., M. Barnhart, D. Choudhury, S. D. Knight, G. Waksman, and S. J. Hultgren. 2000. Chaperone-assisted pilus
assembly and bacterial attachment. Curr. Opin. Struct. Biol. 10:548 –556.
126 Y ASM News / Volume 68, Number 3, 2003
Без категории
Размер файла
1 414 Кб
Пожаловаться на содержимое документа