The Royal Society is collaborating with JSTOR to digitize, preserve, and extend access to
Philosophical Transactions: Biological Sciences.
®
www.jstor.org
S56
\V. Neupert and N. Pfanner
Profcin tacqeting into nzitochondria
potential AY across the inner membrane is an absolute requirement for starting translocation into or
across the inner membranr:. I t is 11otentirely clear how
AY operates but a reasonable possihility is that the
positively charged targeting sequences respond to AY
in a n c1ectrophori:tic manner (Martin ef al. 19910).
This would explain why AY is only able (and
required) to trigger the translocation of the aminotermina1 part (i.e. the presequence) of'the precursor, hut
tioes not play a rolc in the fbllowing translocation of
the main part of the molecule.
After the precursor has partly or completely
reached the matrix, the cleavable targeting sequence
is removed by a soluhle proccssing enzyme. ' 1 ~ ~
components were found to be rccjuired for this reaction, the mitochondrial processing peptidase (YIPP)
and the processing enhancing protein (PEP) (I-Tawlitschek ef al. 1988).
'l'he proteins then undergo fblding in the rnatrix to
lnonomers and oligomers. In those cases xvhere filrther
sorting takes place, in particular to the inner membrane and the intermembrane space, reactions preventing folding must occur (see beloxv). An importarlt
problern in this regard is in what conformational state
precursors are crossing the txvo mitochondrial membranes. I t has become clear already some tirntx ago
that unfolding of precursors is a prerequisite for
transmernbrane 'transfer iEilers & Schatz 1986). l l o r e
recent studies have suggested that unfolding must be
rather extensi\-e. I t is a likely possihility that polypeptides tralversing the two membranes are in a f~llly
evtcnded state. Measurements of the length of the
segment OSa polypeptide chain that is spanning outer
and inner ~nernbranesshow that as little as 50--60
amino acid residues are sufficient to span the two
membranes oLrer a distance of about 15-20 nm (Kassolv el al. 1990). This conclusion led to the suggestion
that translocation occurs in a manner in xvhich the
polypeptide cllain is sliding through a proteinaceous
'translocation pore' af'ter rnore or less complete unfolding (Ncupert el al. 1990).
( 6 ) A role of cytosolic hsp 70s in preserving
precursor proteins in a transport-competent state
T h e conibrmational state of precursor proteins in
the cytosol is usually difyerent frorn that of the nati1.e
functional protein. I t had been noticed already quite
early in the study of mitochondrial protein irnport
that precursor proteins are extremely sensiti\-e to
added proteases under conditions where the functional
counterparts are resistant. Furthermore, a consistent
observation is the strong tendency of precursors
to\\-ards aggregation, both with preproteins synthesized in cell-free systems and with preproteins
rxpressed e.g. in E.rcherichia coli. I n h c t , precursor
proteins ~ n a d cin B. coli can be dissol\-ed in strongly
denaturing media such as 8 M urea, then diluted and
imported into mitochondria; in most cases, ho\\-e\-er,
such preparations are import-competent only for a
short period of tirne, then aggregation takes place
jsurnmarized in Becker el al. 1992). These observations
lrd to the conclusion that precursor proteins arc in a
conformational state that d i f i r s from that of their
mature forms, being more unfolded and exposing
hydrophobic 'sticky' patches.
must
I t is therefore generally assumed that d e lrices
'
exist in the cytosol to prevrnt misfolding and aggregation of preproteins. X number of components h a w
heen implied in such a function. I n particular,
cytosolic hsp 70s, termed Ssal-4p in yeast, ha~vebeen
Sound to he important for the translocation of precursors into mitochondria iand also into the endoplasmic
reticulum). In mutant strains in \\-hich three of the
fbur SS,! genes xvere deleted, precursors of certain
mitochondrial proteins accumulated in the cytosol
0 (Deshairs et al. 1988). Furthermore, experiments in
2,iti.o similarly suggested that cytosolic hsp 70 stimulates protein import into mitochondria (Murakami et
al. 1988).
fexv other cytosolic factors ha\-e also been described which hind to precursors in the cytosol but
which do not helong to the family of heat shock
proteins. Some of these factors halve heen implicated
in interacting specifically with the targeting sequences
(klurakami & Mori 1990). I t rnight well be that a
di~verseset of proteins acts on various parts of the
precursor proteins in the cytosol using difkrent strategies but halving a common aim, namely the preserlration of import-competence of precursors. I n this
respect rnitochondrial import may resernble protein
export frorn bacteria whcre hsp-type components,
such as I h a K and GroEJ,, xvere found to interact
xvith the presecretory proteins in the bacterial cytosol,
hut also more specialized chaperones such as SecB
that d o not belong to the group of heat shock proteins
(Wickner el al. 1991; Phillips Pr Silhavy 1990).
Protein import has a distinct rec~uirementfor A T P
in the cytosol. This requirement is, a t least in part,
due to the action of the heat shock proteins. I n
particular, A T P hydrolysis seems to be required to
release polypeptides from hsp 70s [ Rothman 1989).
This recluire~nentcan explain why hsp in\rol\,ement
and A ' l ~ P hydrolysis are reactions that are usually
found to parallel each other in experiments in rjitro.
.\
( c ) Hsp 70 in the mitochondrial matrix is essential
for driving the translocation of precursors into
the matrix space
O n e of the eight llsp 70s of the yeast Saccharon~yc~s
cei.evi.riaf, is localized within the mitochondria i Craig el
al. 1989). T h e product of this gene, called Ssclp or
mitochondrial (mt-) hsp 70, performs essential fun(.tions in the cell as deletion of the S S C l gene results in
cell death (Craig et al. 1987). Mt-hsp 70 appears to be
a general constituent of mitochondria since it xvas not
only fbund in yeast, but also in humans, plants and
protozoa.
:l terriperature-sensitive mutant afyectcd in SS(,,'/
accumulated ~nitochondrialprecursor forms (Kang et
al. 1990). iZfter a short time period following the shift
frorn permissiLve to non-permissi~~e
temperatures, thc
preproteins xvere found with the mitochondrial fraction, yet they were not translocated into the mitochondria sincc they \\-ere accessible to proteasc added
Protezn t a r g e l z y znlo mztochondrza
to the isolated mitochondria. Detailed biochemical
analysis of the mutant phenotype revealed a dual role
for this mt-hsp 70. I t is in\-ollved in tlle translocation of
precursors across the two mitochondrial membranes,
and it plays a n important role in the folding of
precursors inside tlle mitochondria.
We will first discuss the experimental evidence for
the role of m t hsp 70 in transmembrane movement of
preproteins. Lt'hen precursors were imported into
mitochondria isolated from cells groxvn a t a permissive
temperature and then exposed to a non-permissi~~e
temperature, processing to the mature-sized proteins
xvas obser~-edto almost the same degree as in identically treated mitochondria from wild-type cells. However, when the mutant mitochondria were then
probed as to whether they had imported these precursors into the matrix by treatment with protease, the
answer was no. 'The mutant mitochondria accumulated the precursors in a two-membrane spanning
fashion, i.e. the aminoterminus was able to enter the
matrix space but the mature parts of the precursors
were still exposed to the extramitochondrial space.
'These findings were made with matrix proteins as well
as with proteins of the mitochondrial inner membrane
and intermembrane space ( K a n g et al. 1990; Ostermann et al. 1990).
These data suggested that mitochondrial precursors
halve to interact with mt-hsp 70 to undergo transmembrane transfer, but that the onset fbr this requirement
was only after the precursors had completed the initial
translocation step that is triggered by the response of
the targeting sequences to A Y across the inner
membrane. Direct support for such a mechanism
came from the demonstration that in the temperaturesensitive .rscl-2 mutant (but not in the wild-type)
complexes between the precursors and mt-hsp 70
could be isolated by immunoprecipitation with antibodies against mt-hsp 70. This suggests that in the
mutant there may be a temperature-sensitive deficiency in the dissociation of mt-hsp 70 from the
imported segments of the precursor polypeptides.
Crosslinking experiments with a precursor protein
arrested in a membrane spanning fashion showed that
also in wild-type yeast precursors interact with mt-hsp
70 before translocation is completed [Scherer et al.
1990).
As discussed above, import of most precursor proteins not only recluires A T P in the cytosol, but also
I\TP in the matrix (Hwang & Schatz 1989). 'T'ranslocation across the inner membrane becomes arrested
when the intramitochondrial A T P level is drastically
reduced by incubating isolated mitochondria with
apyrase (an enzyme which degrades A T P and ADP)
and by inhibition of the mitochondrial A'T'P-synthase.
;\TP requirement in the matrix and the requirement
for functional mt-hsp 70 were found to be closely
correlated.
These findings are all compatible with a crucial role
of mt-hsp 70 in d r i ~ - i n g tlle translocation of the
precursors across the two mitochondrial membranes
by binding to the unfolded segments appearing on the
inside of the inner membrane (initially triggered by
AY), and thereby trapping the precursor chain step
I'hil. 'li.0171.
R. Soc. Lorrd. R (1993')
[ 1°1
\.S. Neupert and N. Pfanner
357
by step in the matrix space. A'I'P is required, according to such a mechanism, to confer a conformation
competent for binding to mt-hsp 70, and A T P
hydrolysis is required for the eventual release and
recycling of mt-hsp 70.
I\ salient feature of this proposed mechanism is the
passage of the precursor into the matrix in an
unfolded confbrmation. Only then would there be
sunicient binding sites a,~ailable for dri~vingtranslocation in the direction of the matrix. W e want to
emphasize here that such a driving T-iahsp 70 binding
may not necessarily be the only possible driving
mechanism. In fact, we will discuss below that, with
some kinds of precursors of intermembrane space
proteins, the driving force may be p r o ~ ~ i d eby
d the
export reaction from the matrix into the intermembrane space.
( d ) Hsp 70 in the matrix has a role in facilitating
the unfolding of polypeptide chains on the
mitochondrial surface
If transmembrane mo~-erntmtrequires an extensive
unfolding of the polypeptide chain one may ask what
the mechanisms are that fjcilitate unfolding on the
cytosolic side. As discussed abolre, precursors are
usually in a 'loosely fblded' conformation but these
conformations still contain abundant secondary and
tertiary structure. lloreo\,er, it has been found that in
some cases domains of precursors assume a nati~ve
conformation when they are in the cytosol. Various
speculations have been made on the existence of
unfolding components or 'unfoldases' that might interact lvith folded proteins in such a manner that the
nati~ve folding state is disturbed. Sor far, no clear
experimental e~-idencein h \ - o u r of the existence of
such unfoldases has been obtained.
A mechanism for unfolding can be proposed, howelver, relying on further data obtained through the
analysis of the temperature-sensiti~-e.rscl-2 mutant.
T h e step of i2'I'P-dependent unfolding on the mitochondrial surhce can be bypassed by artificially
unfolding precursors before import. I n particular, this
was done with f~lsionproteins consisting ofaminotermina1 segments of the ATPase subunit 9 precursor and
of complete mouse cytosolic dihydrofblate reductase
( D H F R ) . Upon unfolding (in 8 IVI urea followed by
rapid dilution), these SUS-DHFR precursors can be
imported into isolated mitochondria without addition
of ,4'I'P, in contrast to import of the same precursors
out of reticulocyte lysate where they presumably are
cornplexed with hsp 70.
Import of these unfolded fusion proteins into isolated mitochondria from the .r.rcl-2 mutant was found
to occur with almost the same kinetics as import into
wild-type mitochondria, in sharp contrast to the
situation with import from the reticulocyte lysate
system (Kang et al. 1990). This obser\ration demonstrated that the mutant rnitochondria mainly had
problems with importing precursors into the matrix
when unfolding was a critical factor. Unfolding on the
rnitochondrial surface, i.e. of those parts of a precursor
that are still present in the cytosolic compartment, is
I
358
\V. Neupert and N. Pfhnner
Protezn targetzng znto mzlochondrza
apparently fjcilitated by mitochondrial hsp 70 together with matrix ATP.
Hoxv could a component in the matrix then facilitate tlle unfolding of a protein on the other side of two
membranes? As a n answer to this question we propose
that unfolding in the cytosol is essentially a spontaneous process. \Ye suggest that unfolcling occurs in
very small steps that require a very low amount of
energy. Each time such a small segment has assumed
an extended conformation it could slide in the 'translocation pore' and mt-hsp 70 on the other side I\-ould
have tlle opportunity to bind and make the process
eventually irreversible. Thereby, the ecluilibriu~riprocess of folding and unfolding could be turned into a
\sectorial reaction.
Experimental support for this pathway of' unfolding
comes from the following obser\rations. Membrane
spanning intermediates of cyt b2-1)HFR fusion proteins (aminoterminal parts of tlle cytocllrome h2
precursor filsed to complete D H F K ) can be accurnulated in the presence of rnethotrexate which stabilizes
the DHF'R domain. Import occurs until the folded
D H F R domain 'hits' the outer nlemhrane (Rassow et
al. 1989). Stabilization by the ligand clearly reduces
spontaneous unfolding strongly. 'The free energy of
stahilization is still rather low, being in the range of a
few kcal per mole. This is a \-er). small amount of
energy compared with that needed for the ATPdependent release of the bound mt-hsps. ~ I l t h o u g hit is
not clear how many molecules of matrix A T P are
required for importing an a\-erage-sized mitochondrial
matrix protein, the free energy needed for prel~enting
unfolding
., !e.g. hv, stabilization through a lieand), is
almost certainly negligible compared to the free
energy of ,4'TP hydrolysis needed fbr recycling the mthsps. Also in agreement with the proposed unfolding
mechanism is the fact that the translocation of the
folded D H F R domain (when methotrexate is remo\red
after accumulating transmembrane cyt b2-DHFR) is
much more dependent on temperature as compared to
the import of the urea-unfolded precursor (Eilers et al.
1988; Rassow el al. 1989). I n summary, mt-hsp 70 in
a n indirect manner appears to play a decisive role in
the unfolding of precursor proteins in the cytosol.
\,
result \\-as found both in z~ie~oand after import of
proteins into isolated mitochondria. Further hiochemical investigation then showed that hsp 60 is
directly in\-olved in the folding of individual polypeptide chains. For its activity in facilitating folding, llsp
60 recluires A'TP hydrolysis. Upon reduction of the
iZTP level in the matrix, in the presence of a nonhydrolysahle A'TP analogue, ;\MP-PNP, or a t
reduced temperature, a sloxv-down of fhlding in the
matrix was obser\-ed and the imported precursors
\\-ere largely found in association with the hsp 60
complex (Ostermann el al. 1989). These results gatre
the first experimental indication that folding of individual polypeptide chains in ixivo is mediated by a
chaperone.
H o ~ v e ~ ~not
e r , only hsp 60 is in\vol\-ed in facilitating
folding in tlle matrix. \f:ith the .r.rcl-2 mutant it was
found that the refolding of urea-unfolded precursors
imported a t non-permissi~~e
temperature was strongly
reduced. T h e precursors remained associated with mthsp 70 and were presumably not transferred to hsp 60
( K a n g el al. 1990). I t was therefore suggested that
precursors ha\ve to be passed on from mt-hsp 70 to llsp
60 to become folded (Neupert et al. 1990; Langer el al.
1992). 'This \view has been questioned, however, and it
has been claimed that pools of precursors not complexed with either hsp 70 or hsp 60 exist jhfanningKrieg et al. 1991).
( f ) Hsp 70 and hsp 60 play important roles in
intramitochondrial sorting to the inner m e m b r a n e
and the intermembrane space
k ,
Some proteins that are transported to the matrix do
not remain there but are sorted to other suhmitochondrial compartments. Se\-era1 lines of elridence suggest
that molecular chaperones in the matrix are in\-ol\-ed
in these processes. 11s a n example, the precursor of the
Rieske F e e s protein ( a component of complex 111 of
the respiratory chain anchored to the inner membrane
and exposing a hydrophilic domain into the intermembrane space) requires both mt-hsp 70 and llsp 60
fhr becoming localized to its functional site. I n the
sscl-2 mutant as well as in the m@ mutant its assembly
was found to be impaired i Kang el al. 1990; Cheng et
01. 1989). A similar situation has been described for
( e ) Hsp 70 and hsp 60 in the m a t r i x are involved in
cytochrome bi! ( a soluble component of the intermernfolding of polypeptide chains
brane space) (Koll et al. 1992), although this pathway
has recently been challenged (Click el al. 1992).
Mitochondria horn all sources analysed so far
What is the role of the mitochondrial chaperones in
contain the chaperonin hsp 60 which is structurally
these cases? T h e role of mt-hsp 70 appears to be to
and functionally closely related to GroEI, of prokaryofacilitate the entry of the polypeptides into the matrix
tic organisms [Reading et al. 1989; Ellis Pr Hemmin the same way as it facilitates entry of matrix
ingsen 1989). This similarity appears to reflect the
enclosymbiotic origin of mitochondria although hsp 60
c.omponents. A function of hsp 60 was proposed to lie
of mitocllondria is encoded by a nuclear gene. I,ike
in the protection of the intermediates in the matrix
mt-hsp 70, hsp 60 is essential for the life of yeast cells
against misfolding or aggregation befbre the ensuing
i Cheng et al. 1989).
export reaction [Koll et al. 1992). T h e absence of
T h e analysis of the yeast mutant tniJl, which is
folding by hsp 60 in these cases was thought to be the
deficient in hsp 60 function in a temperature-sensitive
result of the presence of sequences in the intermemanner, revealed a n unimpeded import of precursors
diates, the so-called sorting secluences, which would
into the matrix, hut a lack of formation of a c t i ~ ~ e prelrent accluisition of a native folding state. I t should
be pointed out that this mode of sorting is likely to be
matrix enzymes such as ornithine transcarbamoylase,
the preferred one when import into the matrix is \,cry
F!-ArT'Pase or hsp 60 itself (Cheng el al. 1989). This
Prot~zntarg~tinginto mztochondria
14'. Neupert and N. Pfbnner
359
triggering
translocation
'+pi
ATP
fold~ng
/
Figure 1. Hypothetical model on the roles of molecular chaperones in mitochondrial protein import (modified after
Neupert et al. 1990). C, carboxyltermiriu~of preproteiri; Ihl, iriner memhrarie; N, aminoterminus of preproteiri:
O M , outer membrane.
rapid. I n the experiments described, in fact, import
was perfbrmed using urea-unfblded precursors whose
import is 10-100-fold faster than import out of
reticulocyte lysate. I n the fbrmer case, export may not
keep u p in pace with import, and therefbre hsp 70 and
hsp 60 come into play.
I t appears that in some cases both mt-hsp 70 and
hsp 60 can be bypassed, in particular when export is
fister than import. With certain precursor proteins, in
particular with cyt D2-DHFR fusion proteins in which
the pre-cyt bz part is shorter than approximately 170
amino acid residues, a participation of mt-hsp 70 and
hsp 60 was not observed. I n agreement with this, there
was also no requirement for matrix A T P (unpublished
results). I t appears likely that under such conditions
import into the matrix and export into the iiitermembrarle space are coupled events. T h e driving fbrce fbr
Phll T r a n ~ R.
. Soc. Lund R (1993 I
the overall reaction might then be provided by the
transport into the intermembrane space, rather than
by mt-hsp 70 binding. This specialized pathway is,
however, apparently not possible when longer segments of pre-cyt b2 precede the D H F R domain,
possibly because the energy of the export reaction is
not sufficient to drive import.
3. CONCLUSIONS AND PERSPECTIVES
O u r present knowledge and ideas about the roles of
molecular chaperones in mitochondrial protein sorting
are summarized in a working hypothesis which is
depicted in figure 1. T h e overall reaction is divided
into three consecutive steps.
1. Triggering step. T h e hsp 70-complexed preproteins
[ 103 ]
360
\V. Neupert and N. Pfanner
P ~ o t ~ 2t na t g ~ t ~ n~tn t ornztociiondi~a
interact with the receptor machinery in the outer
membrane and translocation of the aminoterminal
targeting signal into the mitochondrial matrix takes
place.
2. Translocation step. Spontaneous unfolding on the
cytosolic side, sliding of the extended precursor chain
in a 'translocation pore', and binding of mt-hsp 70 on
the matrix side lead to gradual movement of the
precursor into the matrix. A T P is required on both
sides of' the two mitochondrial membranes.
3. Folding step. The precursors are passed on fi-om
mt-hsp 70 to hsp 60 and become folded on the chaperonin. IZTP is required for release of mt-hsp 70 and for
repeated cycles of binding and release o f t h e polypeptide on hsp 60 in the course of the folding process.
the same pathway and the same components in a n
identical manner. Alternative pathways, using
unknown components or bypass pathways, should
always be considered as being possible. This may be
particularly true because the fblding behaviour and
the binding capacity fbr molecular chaperones may
not be unifbrm properties of all precursor proteins.
Future investigations will have to aim at a critical
e x a m i ~ ~ a t i oofn our prcsent working hypothesis. They
will have to detect a presumably large number OS
additional chaperones and related components, they
must dcfine the molecular mechanisms of' chaperone
action in mitochondria to provide a much more
precise picture, and they will have to search for
additional functions of the known mitochondrial chaperones.
It may be useful to look at the functions of
mitochondrial molecular chaperones from a n evolutionary angle. T h e prokaryotic origin of the mitochondria is reflected in the amazing similarities of the
molecular roles of hsp 60 and grol<L (l<llistL Hemmingsen 1989; l'lartin et al. 199 16). llitochondrial hsp
70 is a very close relative of bacterial I>naI< (Craig et
al. 1989). \;t'hy should the mitochondrial equivalent of
I>liaK have acquired such a specific role in driving the
import OSproteins which in the prokaryotic ancestors
of the mitochondria were encoded by the DNA o f t h e
endosymbiont:' T h e main function of DnaKjhsp 70
(.haperones apprars to be binding to unfolded segments OS polypeptide chains in a reversible fishion
lFlynn et a1 1989). O n e can easily imagine that such a
rather simple and general tool can be used fbr
controlli~ig a large host of cellular reactions. l'litoc:hondrial import may have evolved taking advantage
of two basic devices that rould have existed already in
the prokaryotic ancestors. This first device would be a
controllable 'pore' that allows the passage of polypeptides through the envelope in an extended state; the
existence of pores, Sor example for the uptakc of
polypeptides such as colicin A (Bourdineaud ei al.
1989). might indicatc that such an assumption is not
elitirely unreasonable. T h e l)naI< in the prokaryotic
ancestor would then be a complementary tool as it
was especially designed to bind to extended parts of
polypeptide chains. These two de\.ices together would
then havc: served to set up the basic machinery fbr
protein import into endosymbiotic organelles.
It should be made clear at this point that the
working hypothesis illustrated in figure 1, if true in all
parts, is certainly f i r from being complete. I t is very
likely that additional protein components are involved
in tlle complex reaction scheme. 1,'rom bacteria we
know that, for instance, in addition to I h a K the heat
shock proteins IlnaJ and GrpL exert important
functions in chaperone action (Georgopoulos et al.
1990; 1,anger et al. 1992). I t is not unlikely that related
components also function in mitochondria. I n yeast, a
llnqJ homologue has already been described which
c,ould have such a role, although a mitocl~ondrial
location in this case has not beer1 proven (Ulumberg 8i
Silver 1991). The proposed reaction scheme is also not
meant to indicatc that all precursor proteins rnust use
\.VC arc g r a t r f ~ ~tol Alexaridra \.Vrinzirrl for preparing thr
drawing. Studies from the authors' laboratorirs wrrc supportcd by tfie Ilentschc Forscfiurigs~en~eillscl~aft,
tfic
Hurnari I'rontier Scirncc Program and thc Fonds der
Chrmischrn Industrie.
REFERENCES
Baker, K.P. & Schatz, G. 1991 blitochondrial protrills
essential for viability mcdiate protrin import into yeast
mitocllondria. .Vatni-r, Lond. 349, 205-208.
lsecker, K . , Gniard, l%.,Rassow, J., Sbllnrr, 'l'. & Pfaririer,
N. 1992 Targetirig of a chrmically pure preprotcin to
mitochondria does not require tfie addition of a cytosolic
signal recopition factor. J. biol. C%eni. 267. 5637-.i643.
t$lumbcrg, H . & Silver, P.,4. 1991 11 homologue of the.
bacterial heat-shock gcnc Dnci.J that alters protein sorting
in yrast. hhtui-e, Lond. 349, 627-630.
Rourdirieand, J.-l'., Howard, S.P. & Lazdunski, C. 1989
Localization and assembly into thr L'sch~ricttia cull cnvclope of a protrill required for mtry of colicin A. .J. Wact.
171: 21.58-2463.
Chcrig, hl.Y., Hartl, l:.-L!., hlartin, J., Pollock, 11.1\.,
Kalousek, F., Ncnpcrt, \V., Hallberg, E.hI., Hallberg,
R.L. & Horwich, A.L. 1989 hlitocholldrial heat-shock
protein hsp60 is cssrntial for assemblb- of proteins
irnportcd irito );cast mitochondria. 'Vaturr, Land. 337, 620625.
Craig, Ii.i\., Krarner, J. Kr. Kosic-Smithers, J . 1987 S.SC1, a
member of thr 70-klla hrat shock proteiri multigene
family of Saccharornycrs cerez:iszae, is essential for growth.
Proc. natn. ilrad. Sci. L;.S.A. 84, 41 56-41 60.
Craig, E A . , Kramer, ,J., Shilling, J., \.Vernrr-\Vashburne,
LT.,Holmes, S., Kosic-Smithers, .J. & Xicolet, C.M. 1989
S S C I , a n essential member of'the yeast HSP 70 multigene
family, encodes a mitochondrial protein. Allolrr. Cell. Biol.
9, 3000-3008.
Craig,
Deshaies, R.J., Koch, B.D., LVerner-LVashburne, M,.
E.A. 8L Sheckman, R. 1988 A subfamily oi'stress proteins
facilitates translocation of secretory and mitochondrial
precursor polypeptides. :Lr(itzlre,Lond. 332, 800-805.
Eilers, M.& Schatz, G . 1986 Binding of a specific ligatld
inhibits import of a purified precursor protein into
mitochondria. .Vature, I,ond. 322, 228-232.
Eilers, h[.,Hwang, S. & Schatz, G. 1988 Unfolding and
refblding of a purified precursor protein during import
into isolated mitochondria. E A l B O J. 7, 1 1 3 9 1145.
Ellis, J. 1987 Proteins as molecular chaperones. &lure,
Lond. 328, 378-379.
Ellis, R.J. & Hemmingsen, S.RI. 1989 hlolecular chaper-
Protezn targetzng into mitochondrza
ones: proteins essential fbr the biogenesis of some macromolecular structures. Trends Biochem. Sci. 14, 339-342.
F1)-nn, G.C., Chappell, 'I'.G. & Rothman, ,J.E. 1989
Peptide binding and release by proteins implicated as
catalysts of protein assembly. Scienrr, W a s h . 245, 385-390.
Georgopoulos, C., Ang, D., Liberek, K. & ZJ-licz, M. 1990
Properties of the Escherichiu coli heat shock proteins and
their role in bacteriophage h growth. I n Slres.c proteins in
biolog,y and medzcine (ed. 'l'. X4orimot0, A. 'l'issiers &
C. Georgopoulos), pp. 191-221. New York: Cold Spring
Harbor Laborator). Press.
Glick, B.S., Brandt, A., Cunningham, K., Muller, S.,
Hallberg, R.L. & Schatz, G. 1992 Cytochromrs cl and hn
are sorted to the intermembrane space of yeast mitochondria by a stop-transfer mechanism. C'ell 69, 809-822.
Ha~vlitschek, G., Schneider, H., Schmidt, B., Tropschug,
hl., Hartl, F.-U. & Xeupert, \ \ . 1988 hlitochondrial
protein import: identification of processing peptidase and
of PEP, a processing enhancing protein. Cell 53, 795 806.
Horwich, A. 1990 Protein import into mitochondria and
peroxisomes. Curr. Opin. Crll Biol. 2, 625-633.
Hwang, S.T. & Schatz, G. 1989 'l'ranslocation of proteins
across the mitochondrial inner membrane, but not into
the outer membrane, requires nucleoside triphosphates in
the matrix. Proc. nutn. ilrud. Sci. LT.S.A. 86, 8432-8436.
Hwang, S.'I'., \Vachter, C. & Schatz, G. 1991 Protein
import into the yeast mitochondrial matrix: a new
translocation intermediate between the two mitochondrial membranes. J. biol. Chrm. 266, 21083-21089.
Kang, P;J., Ostermann,,J., Schilling,,J., Neupert, \V., Craig,
E.A. & Pfanner, X. 1990 Requirement for hsp 70 in the
mitochondrial matrix for translocation and folding of
precursor proteins. Nuturr, Lond. 348, 137-143.
Koll, H., Guiard, B., Rasso~v,,J., Ostermann, ,J., Horwich,
A.L., Neupert, \V. & Hartl, F.-U. 1992 Antifblding
activity of' hsp 60 couples protein import into the mitochondrial matrix ~vith export to the intermembrane
space. C'ell 68, 1163-1 175.
Langer, 'l'., Lu, C., Echols, H., Flanagan, J., Ha)-er, M . K .
& Hartl, F.-U. 1992 Successive action of DnaK, DnaJ
and GroEL along the pathwa)- of chaperone-mediated
protein folding. ~Vuture,Lond. 356, 683-689.
Llanning-Krieg, U.C., Scherer, P.E. & Schatz, G. 1991
Sequential action of mitochondrial chaperones in protein
import into the matrix. E,WBO J . 10, 3273-3280.
Llartin, J., Mahlke, K . & Pf'anner, N. 1 9 9 1 ~Role of an
energized inner membrane in mitochondrial protein
import: AY drives the movement of presequences. J. biol.
Chem. 266, 18051-18057.
Martin, ,J., Langer, T., Boteva, R., Schramel, A., Horwich,
A.L. & Hartl, F.-U. 19916 Chaperonin-mediated protein fblding at the surface of groEL through a 'molten
globule'-like intermediate. l k l u r r , Lond. 352, 36-42.
hlurakami, H., Pain, D. & Blobel, G. 1988 70-kD heat
shock-related protein is one of' at least two distinct
c)-tosolic hctors stimulating protein import into mitochondria. .J. Cell Biol. 107, 2051-2057.
Murakami, K. & hlori, M. 1990 Purified presecluence
binding factor (PBF) fbrms an import-competent complex
with a purified mitochondrial precursor protein. E A f B O
.I. 9, 32013208.
Ncupcrt, \V., I-Iartl, I.'.-U,, Craig, E.A. 8r Pfanner, K. 1990
How do polypeptides cross the mitochondrial membranes!' Cell 63, 447-4.50.
Ostermann,,J., Horwich, A.L., Neupert, \V. & Hartl, F.-U.
1989 Protein fblding in mitochondria requires complex
fbrmation with hsp 60 and A'I'P hydro1)-sis. ~Vuturr,Lond.
341, 12.5-130.
Ostermann,,J., Voos, \V., Kang, P:J., Craig, E.A., Neupert,
Piiil. Tmrzj. K . Soc. Lorid. B (l993
8
[ l05
M'. Neupert and N. Pfanner
361
\V. & Pfanner, X. 1990 Precursor proteins in transit
through mitochondrial contact sites interact with hsp 70
in the matrix. F E U S Lett. 277, 281-284.
Phnner, N. & Neupert, \ \ . 1990 The mitochondrial
protein import apparatus. A. Rev. Biorhem. 59, 331-353.
Phillips, G.,J. & Si1hav)-, 'I'.J. 1990 Heat-shock proteins
DnaK and GroEL facilitate export of LacZ hybrid
proteins in E . coli. ~Vuturr,Lond. 344, 882-884.
Rasso~v,J . & Pfanner, X. 1991 hlitochondrial preproteins
en route from the outer membrane to the inner membrane
are exposed to the intermembrane space. F E B S Lell. 293,
85-88.
Rasso~v,,J., Guiard, B., \Vienhues, U , , Herzog, \'., Hartl,
F.-U. & Neupert, \ \ . 1989 Translocation arrest by
reversible fblding of' a precursor protein imported into
mitochondria. A means to quantitate translocation contact sites. J . fill Biol. 109, 1421-1428.
Rassow, ,J., Hartl, F.-U,, Guiard, B., Pfanner, N. &
Neupert, \V. 1990 Po1)-peptides traverse the mitochondrial envelope in an extended state. F E R S Lett. 275, 190
194.
Reading, D.S., Hallberg, R.L. & Llyers, A.hl. 1989
Characterization of the )-east HSP 60 gene coding for a
mitochondrial assemb1)- factor. Al'ulure, Land. 337, 655659.
Rothman, ,J.E. 1989 Polypeptidc chain binding proteins:
catalysts of' protein fblding and related processes in cells.
C'ell 59, 591-601.
Scherer, P.E., Krieg, U.C., Hwang, S.T., Vestweber, D. &
Schatz, G. 1990 A precursor protein partly translocated
into yeast mitochondria is bound to a 70 kd mitochondrial
stress protein. E M B O J. 9, 431.5-4322.
Schleyer, M . & Neupert, \\. 198.5 Transport of proteins
into mitochondria: translocational intermediates spanning
contact sites between outer and inner membranes. Cell 43,
339-350.
\Vickner, \V., Driessen, A.J.14. & Hartl, F.-U. 1991 'l'he
enzymology of protein translocation across the Escherichiu
coli plasma membrane. 11. Re!;. Riorhem. 60, 101-124.
Discussion
\V. J. \VELCH (Department of Medicine and Physiolop.y,
University of C'alfornia, San Franscisco, U . S . A . ) . \$Thy do
proteins destined for the intermembrane mitochondrial space have to interact with hsp 60? If hsp 60
assists protein folding does this mean that the ironsulphur protein is active in the matrix?
\V. NEUPERT. I imagine that hsp 60 interacts with
most proteins that enter the matrix space and tries to
assist its folding: it certainly does this even with
fbreign proteins like DHF'R. But the protein may not
be capable of' being fblded because a signal sequence is
still present, or because of'the presence of' other fictors
which prevent folding. So the extent of interaction
with hsp 60 will depend on these fkctors, and on the
kinetic situation, and may not be obligatory for
proteins destined fbr the intermembrane space. T h e
situation could resemble that suggested for the role of
GroEI, in protein transport in bacteria.
12'. J. ~ V E L C H .DO the authors think there may be a
role for T C P l in binding to mitochondrial precursor
proteins in the cytosol?
1
362
It'. Neupert and N. PfBnner
Protein targeting into mitochondria
12'. NEUPERT.I have no evidence on which to assess
this possibility.
A. BAKER (Department of Biochemistry, University of
Cambridge, U . K . ) . The authors suggest that the two
different models fbr protein transport into the intermembrane space can be reconciled by taking into
account kinetic factors. It is possible to change the
pathway of transport of, fbr example, cytochrome b2,
by altering the rate of transport?
It:. NEUPERT.Kinetic fkctors will not be able to
reconcile the two models, as they are mechanistically
different, however, they may explain controversial
interpretations of' experimental data. Since the two
models are not mutually exclusive it cannot be
excluded that different pathways might be favoured
by different experimental conditions.
A. BAKER.I n the experiments with protein import
into mitochondria fkom the hsp 60 mutant where the
authors fhund no mature form of the iron-sulphur
protein, did they observe any mature protease-protected form of cytochrome b2?
\V. NEUPERT.There could be some present, but
clearly less than in the control mitochondria.
11.-J. CETHING
( H o w a ~ d Hughes Medical Institute.
Unizlersity of Texas, Dallas, Lr.S.A.). If, as the authors
suggest, the driving force fbr protein import into the
matrix space is binding to the matrix hsp 70, what is
the driving force for the transport fkom the matrix to
the intermembrane space? This situation seems to
parallel the transport of proteins into the periplasmic
space of' bacteria.
It'. NEUPERT.It'e do not know the answer to this
interesting question. At the moment we can study only
the import and export steps at the same time, and
what we require for such studies is a vesicle slstem
which will transport from the matrix. There is of
course A T P in the intermembrane space, but how
proteins fold in this space is an open question.
R. JAENICKE (Department of Biophysics and Physical
Biochemistry. Lrniversity of Regensburg, F . R . G . ) . How long
does the precursor protein on the cytosolic side of' the
mitochondrial membrane have to be to interact with
hsp 70?
\V. NEUPERT.There is no systematic study of this
aspect. Cytochrome c is a small mitochondrial protein
and we have been unable to see any interaction
between this protein and hsp 70; the transport of' this
protein across the outer membrane into the intermembrane space does not require A T P on the cytosolic
side.