Cholesterol, It’s Not Just For
Heart Disease Anymore
Amy Kerzmann and Andrew L. Feig*
Department of Chemistry, Indiana University, 800 East Kirkwood Avenue, Bloomington, Indiana 47405
A
n insidious threat lurks in the
dark corners of hospital wards.
Because of the widespread use of
broad-spectrum antibiotics, C. difficile, an
opportunistic pathogen, has become one
of the most common hospital-acquired
infections in the United States and Canada
(1–3 ). It is also a growing threat to patients
in nursing homes and extended care facilities (2 ). C. difficile colonizes the underpopulated anaerobic niches in the GI tracts
of patients after their normal microflora
has been killed. This organism causes
pseudomembraneous colitis and severe
antibiotic-associated diarrhea, also called
C. difficile-associated diarrhea (CDAD) (1 ).
The organism secretes two toxins, Toxin A
(TcdA) and Toxin B (TcdB), that are the
virulence factors responsible for the cellular
damage (4 ). It was estimated in 2002 that
the U.S. medical community spends more
than $1.1 billion annually combating these
infections (5 ). Recent work by Giesemann
and colleagues shows that membrane
cholesterol levels play a significant role
in the transport of the C. difficile toxins
into eukaryotic cells (6 ). The data suggest
the intriguing possibility that cholesterol
may act as a small molecule chaperone to
facilitate the insertion of the protein into
the membrane, thus, generating the pore
necessary to translocate the catalytic component of the toxin into the cytoplasm.
TcdA and TcdB are very large proteins
(on the order of 300 kDa). They share
48% identity and in vitro behave in a very
similar fashion. Each toxin is comprised
of three functional domains (Figure 1,
www.acschemicalbiolog y.o rg panel a) (7 ). A C-terminal domain binds
cell surface receptors for target recognition, and an N-terminal domain carries
a glucosyl­transferase functionality that
targets a family of Ras-like G-proteins,
disrupting their function upon modification. The large central domain participates
in the translocation process. It creates the
pore necessary to transport the catalytic
domain across the endosomal membrane
after endocytosis. Once internalized, it
is believed that the catalytic domain is
proteolytically processed by a cellular protease around residue 543, freeing it from
the remainder of the toxin which, having
completed its function, remains in the
endosomal membrane (8 ). Interrupting any
of these three major steps (cellular recognition, internalization, or catalysis) could in
theory disrupt intoxication.
The Internalization Process. One of the
critical events during intoxication by the
bacterial toxin is a protein transduction
step, the movement of the catalytic domain
from outside the cell into the cytoplasm
(Figure 1, panel b, steps 3–6) (9, 10 ). In
this process, the toxin must trigger endo­
cytosis and then escape from the endosome into the cytoplasm. The toxins that
use endosome-mediated uptake can be
divided into two classes. Cholera and Shiga
toxins exemplify one class. They travel from
the endosome into the Golgi body and
eventually to the endoplasmic reticulum
before leaving the vacuole and entering the
cytoplasm. This mode of uptake is called
the long-trip model. TcdA and TcdB are
members of the other class of toxins that
A b s t r a c t Recent studies have shown that
cholesterol plays a significant role in the ability
of Toxin A from Clostridium difficile to enter
eukaryotic cells. The translocation process is
one of three major steps during intoxication that
could be targeted for intervention against the
severe antibiotic-associated diarrhea caused by
C. difficile.
*To whom correspondence
should be addressed.
E-mail: [email protected].
Published online April 21, 2006
10.1021/cb600133b CCC: $33.50
© 2006 by American Chemical Society
VOL.1 NO. 3 • ACS CHEMICAL BIOLO GY
141
Figure 1. a) The three functional regions of
TcdA: the N-terminal enzymatic domain (red),
the central translocation region (orange),
and the C-terminal repetitive oligopeptide
(CROP) domain (green). There is also a small
C-terminal hydrophobic region present
after the CROP region in TcdA (white), not
observed in TcdB. The highlighted residues are
essential for UDP-Glc binding (W102) and/or
catalysis (DxD motif at 286–288). b) Cellular
intoxication by C. difficile TcdA and TcdB.
(Step 1) TcdA and TcdB are exported from
the bacterium; (step 2) the C-terminal CROP
motif binds to cell-surface carbohydrates;
(steps 3 and 4) the toxin–receptor complex
is internalized through receptor-mediated
endocytosis; (step 5) acidification of the
maturing endosome by V-type ATPases
drives a pH-dependent conformational
change of the central translocation domain,
resulting in insertion into cholesterolcontaining endosomal membranes; (step 6)
the N‑terminal catalytic fragment is released
into the cytosol; (step 7) once in the cytosol, the toxin fragment catalyzes the transfer of glucose from UDP-glucose to a conserved threonine
residue of specific Ras-like GTPases; and (step 8) monoglucosylation of the G-proteins blocks the conformational changes that normally occur
as the protein switches between GDP- and GTP-bound states. By preventing these conformational changes, these proteins are effectively
“turned off” and are unable to interact with their effector proteins leading to depolymerization of the actin cytoskeleton, cellular rounding, and
ultimately cell death.
also includes the diphtheria and anthrax
toxins. These toxins escape directly from
the early endosomes (Figure 1b, step 6).
The endosomal compartment is acidified
by vacuolar ATPases as it begins to travel
toward the Golgi body. The toxin uses the
acidification process as its cue to insert into
the membrane, forming a pore and extruding its catalytic domain into the cytoplasm.
Giesemann and colleagues have
explored the translocation phase of TcdA
intoxication and found it to be highly
dependent upon the presence of membrane
cholesterol (6 ). The authors preloaded cells
with 86Rb+. They then used 86Rb+ efflux
measurements and single channel conductance to probe whether the toxin success­
fully inserted into the membrane and
formed a channel. By analyzing intoxication
of cell lines that differed in membrane
cholesterol content, they could compare
the relative susceptibility to the toxin. They
also treated cells with methyl-b-cyclodextrin
142
VO L .1 N O. 3 • 1 4 1 –144 • 2006
(MbCD), a reagent that binds to and
depletes cholesterol from the plasma membrane. Cells treated with MbCD showed
marked reductions in their mortality after
exposure to the toxin. The toxin still bound
to the surface of the cholesterol-depleted
membranes, but failed to form pores, and
the catalytic domain was incapable of being
transported into the cytosol.
One possible reason for cholesteroldependent toxin uptake would be involvement of lipid raft structures or the receptors associated with them. Lipid rafts are
specialized microdomains in the membranes that have a high concentration of
both cholesterol and membrane proteins
(11, 12 ). By depleting cells of cholesterol,
one might disrupt these rafts and thereby
disperse the receptors contained therein.
To address this possibility, Giesemann and
colleagues treated cells with a phosph­a­­tidylinositol-specific phospho­lipase C (PI-PLC).
This enzyme hydrolyzes the proteins and
Kerzmann and Feig
polysaccharide receptors from glycosyl­
phosphatidylinositol (GPI)-anchored structures, leaving the remainder of the lipid raft
intact. HT-29 cells treated with PI‑PLC were
just as susceptible to intoxication as the
untreated cells. This result demonstrated
that GPI-anchored receptors are not required
for uptake, but the remainder of the lipid
raft might still play a role.
So what is the function of cholesterol in
toxin translocation? Are there direct interactions between cholesterol and the toxin or
does cholesterol simply affect the physical properties of the membrane making
it more susceptible to protein insertion?
A clue may come from examining the role
of cholesterol in cytolysins, another class of
bacterial toxins exemplified by perfringolysin O (PFO) (13 ). Cytolysins kill eukaryotic
cells by forming oligomeric structures
that breach the plasma membrane with
large pores up to 300 Å across (13 ). It has
been hypothesized that stable folding of
w w w. a c s c h e m i ca l biology.org
Figure 2. Structural biology of C. difficile toxin fragments. a) Model resulting from the X-ray
crystal structure of a catalytically active N-terminal fragment of TcdB consisting of residues
1–543 (reprinted with permission from ref 16, Copyright 2005 Elsevier B.V.). The conserved
GT-A fold is shown in blue. The catalytic DxD motif is shown in ball-and-stick mode, as well as
UDP, glucose, and the catalytic Mn2+ ion. b) A structural model of the CROP domain from TcdA.
The model is based on a crystal structure of a 127 amino acid fragment (residues 2573–2709)
revealing stacked pairs of b-hairpins in a b-solenoid fold (reprinted with permission from
ref 17, Copyright 2005 National Academy of Sciences, U.S.A.).
the monomeric PFO prevents premature
aggregation of the cytolysins in solution. In
cases where cholesterol has been depleted
from target membranes, PFO oligomerizes
on the membrane surface, but fails to insert
into the membrane (14 ). Thus, cholesterol
then may help to unfold the preinsertion
structure of the toxin during the initial
stages of membrane interaction (15 ).
C. difficile toxins do not need to aggregate the way PFO and related cytolysins do.
However, the translocation domains of TcdA
and TcdB do need to refold from their initial
solution conformation to their membraneinserted pore conformation at the proper
time during entry into the cytosol (Figure 1,
panel b, step 5). If the toxin refolds too early,
it may be subject to aggregation and precipitation. If it occurs too late, it will have lost its
chance to escape from the endosome. It is
possible that acidification alone is insuffi­
cient to destabilize the preinsertion structures of TcdA and TcdB. Giesemann’s results
www.acschemicalbiolog y.o rg are consistent with a mechanism whereby
cholesterol may chaperone the membrane
insertion process. The search for the mechanism by which cholesterol facilitates intoxication may provide a detailed window into the
nature of protein insertion into membranes
in addition to providing a potential mode for
therapeutic intervention for CDAD.
Receptor Binding and Catalysis. Recent
work has also expanded our understanding
of the two other steps in the intoxication
process. A high-resolution crystal structure
of the catalytically active N-terminal domain
consisting of residues 1–543 of TcdB was
recently reported (Figure 2, panel a) (16 ).
The model shows an extensive network
of b structure at its core surrounded by a
cluster of a helices. The core of the fold is
homologous to the GT-A family of glycosyltransferases that also includes glycogenin,
a3-GalT, and LgtC. Whereas the basic
glucosyl­transferase behavior is probably
quite similar to that of other members
of this family of enzymes, the toxins are
unique in their exquisite selectivity for their
protein acceptors. Besides TcdA and TcdB,
there are several other large clostridial
toxins, and each has a unique set of cellu­
lar targets (4 ). The potential to use this
selectivity in specifically targeting the toxins
for inhibition is so far underutilized.
A 127 amino acid fragment from the
C‑terminus of TcdA has also been structurally characterized (Figure 2, panel b)
(17 ). This fragment derives from the CROP
region of the protein responsible for binding receptors on the surface of colonic
epithelial cells (Figure 1, panel b, step 2).
This repetitive sequence folds into a series
of b hairpins that stack on top of one
another to form an extended filamentous
assembly with a significant helical twist.
This domain most likely protrudes from the
body of the toxin in search of an appropriate cell-surface receptor, believed to be a
short oligosaccharide motif. Several candi­
date trisaccharides have been reported
as potential targets for TcdA binding,
including Gal-a1,3-Galb-1,4-GlcNAc (18 )
and GalNAc-b1,3-Gal-b1,4-GlcNAc (19 ), but
it remains unclear whether these are in fact
the biologically relevant motifs. Much work
remains to be done on this aspect of the
cellular recognition problem.
Together, these most recent studies on
the role of cholesterol and the mechanism
by which the C. difficile toxins recognize,
penetrate, and kill host cells will facilitate
a host of additional experiments on these
systems. Scientists have been working
tirelessly to develop novel antibiotics and
immunization against C. difficile, and this
route still holds significant promise for
long-term efficacy. The deeper understanding of the molecular details of intoxication,
however, may allow direct targeting of the
toxins in the ongoing battle against CDAD.
Acknowledgment: The authors acknowledge
support from NIH Training Grant GM-007757 (to
A.K.) and a Women in Science Fellowship (to A.K.).
A.L.F. is a Cottrell Scholar of Research Corporation.
VOL.1 NO. 3 • 141—144 • 2 0 0 6
143
References
1. Kelly, C. P., Pothoulakis, C., and LaMont, J. T. (1994)
Clostridium difficile colitis, N. Engl. J. Med. 330,
257–262.
2. Lamont, J. T. (2002) Theodore E. Woodward Award.
How bacterial enterotoxins work: Insights from
in vivo studies, Trans. Am. Clin. Climatol. Assoc. 113,
167–180.
3. Valiquette, L., Low, D. E., Pepin, J., and McGeer, A.
(2004) Clostridium difficile infection in hospitals:
A brewing storm, Can. Med. Assoc. J. 171, 27–29.
4. Just, I., and Gerhard, R. (2004) Large clostridial
cytotoxins, Rev. Physiol. Biochem. Pharmacol. 152,
23–47.
5. Kyne, L., Hamel, M. B., Polavaram, R., and Kelly, C. P.
(2002) Health care costs and mortality associated
with nosocomial diarrhea due to Clostridium difficile, Clin. Infect. Dis. 34, 346–353.
6. Geisemann, T., Jank, T., Gerhard, R., Maier, E.,
Just, I., Benz, R., and Aktories, K. (2006) Cholesteroldependent pore formation of Clostridium difficile
toxin A, J. Biol. Chem. 281, 10808–10815.
7. Moncrief, J. S., Lyerly, D. M., and Wilkins, T. D.
(1997) Molecular biology of the Clostridium difficile toxins. In The Clostridia: Molecular Biology
and Pathogenesis (Rood, J. I., McClane, B. A.,
Songer, J. G., and Titball, R. W., Eds.), pp 369–392,
Academic Press, San Diego, CA.
8. Rupnik, M., Pabst, S., Rupnik, M., von EichelStreiber, C., Urlaub, H., and Soling, H. D. (2005)
Characterization of the cleavage site and function
of resulting cleavage fragments after limited proteo­
lysis of Clostridium difficile toxin B (TcdB) by host
cells, Microbiology 151, 199–208.
9. Lord, J. M., and Roberts, L. M. (1998) Toxin entry:
Retrograde transport through the secretory pathway,
J. Cell Biol. 140, 733–736.
10. Lord, J. M., Smith, D. C., and Roberts, L. M. (1999)
Toxin entry: How bacterial proteins get into mammalian cells, Cell. Microbiol. 1, 85–91.
11. Munro, S. (2003) Lipid rafts: Elusive or illusive?
Cell 115, 377–388.
144
VO L .1 N O. 3 • 1 4 1 –144 • 2006
12. Pike, L. J. (2004) Lipid rafts: Heterogeneity on the
high seas, Biochem. J. 378, 281–292.
13. van der Goot, F. G. (2003) Membrane-damaging
toxins: Pore formation. In Bacterial Protein Toxins
(Burns, D. L., Barbieri, J. T., Iglewski, B. H., and
Rappuoli, R., Eds.), pp 189–202, ASM Press,
Washington, DC.
14. Giddings, K. S., Johnson, A. E., and Tweten, R. K.
(2003) Redefining cholesterol’s role in the mechanism of the cholesterol-dependent cytolysins, Proc.
Natl. Acad. Sci. U.S.A. 100, 11315–11320.
15. Heuck, A. P., Hotze, E. M., Tweten, R. K., and
Johnson, A. E. (2000) Mechanism of membrane
insertion of a multimeric beta-barrel protein:
Perfringolysin O creates a pore using ordered and
coupled conformational changes, Mol. Cell 6,
1233–1242.
16. Reinert, D. J., Jank, T., Aktories, K., and Schulz, G. E.
(2005) Structural basis for the function of
Clostridium difficile toxin B, J. Mol. Biol. 351,
973–981.
17. Ho, J. G., Greco, A., Rupnik, M., and Ng, K. K. (2005)
Crystal structure of receptor-binding C-terminal
repeats from Clostridium difficile toxin A, Proc. Natl.
Acad. Sci. U.S.A. 102, 18373–18378.
18. Krivan, H. C., Clark, G. F., Smith, D. F., and
Wilkins, T. D. (1986) Cell surface binding site for
Clostridium difficile enterotoxin evidence for glycoconjugate containing the sequence Gala1-3Galb14GlcNAc, Infect. Immun. 53, 573–578.
19. Teneberg, S., Lonnroth, I., Torres Lopez, J. F.,
Galili, U., Halvarsson, M. O., Angstrom, J., and
Karlsson, K. A. (1996) Molecular mimicry in the
recognition of glycosphingolipids by Gal a3Galb4GlcNAcb-binding Clostridium difficile toxin A, human
natural anti a-galactosyl IgG and the monoclonal
antibody Gal-13: Characterization of a bindingactive human glycosphingolipid, non-identical with
the animal receptor, Glycobiology 6, 599–609.
Kerzmann and Feig
w w w. a c s c h e m i ca l biology.org