512
Chaperones, Molecular
major chaperone families.
In vitro
experi-
ments show that unfolded proteins can be
held in a nonaggregated but nonactive state
bound to the sHSPs, and subsequently
transferred to the DnaK and GroEL fam-
ilies for folding.
In vivo
experiments also
support the existence of such interactions,
although these interactions cannot be es-
sential for viability as even under fairly
extreme conditions, loss of the sHSPs from
E. coli
does not cause a severe phenotype.
It may be that, as was seen in the case
of trigger factor and DnaK, the ability for
the cell to ‘‘hold’’ transiently unfolded pro-
teins in a refoldable form is so important
that more than one protein can act as the
‘‘holder,’’ and to date the other proteins
that can do this have not been identiFed.
A third important interaction that exists
between chaperones in
E. coli
is that
between the Hsp70 protein DnaK and the
Hsp100 protein, ClpB. Here, the evidence
for an important interaction is much more
compelling. It is clear both from
in vitro
and
in vivo
experiments that proteins
aggregated by a stress such as heat shock
can be disaggregated by the combined
actions of DnaK and ClpB, and that the
two proteins have very different roles in
the process as they cannot substitute for
each other. It is likely that ClpB acts
upon protein aggregates Frst, altering their
structure to a point where DnaK is able to
promote their refolding. ClpB has thus
been classiFed by some workers as an
‘‘unfolder,’’ and it is not able to assist the
refolding of proteins on its own. Its mode
of action is not yet understood, but it is a
member of a large family of proteins called
the AAA proteins, many of which have
a major role in protein quality control.
This leads us into the complex area of
the overlap between protein refolding and
protein degradation.
E. coli
(and all other organisms) con-
tains a large number of different proteases
whose job it is to remove inactive or dam-
aged proteins from the cell. This process
becomes particularly important at high
temperatures. A large number of these pro-
teases are, along with ClpB, members of
the AAA family of proteins, and some have
features reminiscent of molecular chaper-
ones. ±or example, several of them have
an ATP-dependent ability to unfold pro-
teins, and exist as ring-shaped oligomers
(usually hexamers). ±or most of these, un-
folding occurs as an essential step prior to
degradation of the unfolded protein, either
by another site on the same polypeptide
that mediated the unfolding, or by another
pro
te
intha
tac
tsinconcer
tw
i
ththe‘
‘un
-
foldase.’’ The unfoldase activity can, in
some cases, give rise to a chaperone-like
activity in
in vitro
experiments, although
only in the case of ClpB does it appear to
haveatruero
leinhe
lp
ingpro
te
instore
-
fold
in vivo
. Degradation by these proteases
is reduced by mutations in the DnaK and
GroEL chaperone families, implying an
interaction, indirect or direct, between the
chaperone proteins and the proteases.
Thus, when proteins are unfolded in
the
E. coli
cell, they may be captured
by ‘‘holders’’ for subsequent refolding, or
form aggregates that can be disaggregated
by ClpB before refolding by DnaK (±ig. 7).
In addition, they may be targeted to cellular
proteases. Both these processes will help
ensure that only active proteins are present
in the cell, but the way the balance between
them is achieved (i.e. what determines
whether a given protein is degraded or
refolded) is not known. Clearly, refolding
is a preferential option to degradation if
the cell is not to lose the investment in
ATP that it has already made in producing
the protein; equally clearly, refolding is not
previous page 1186 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 1188 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off