precursor of mitochondrial aspartate aminotransferase synthesized in escherichia coli is complexed...

Eur. J. Biochem. 208,699-704 (1992) 0 FEBS 1992

Precursor of mitochondrial aspartate aminotransferase synthesized in Escherichia coli is complexed with heat-shock protein DnaK Daniel SCHMID, Rolf JAUSSI and Philipp CHRISTEN

Biochemisches Institut, Universitat Zurich, Switzerland

(Received May 22/July 9, 1992) - EJB 92 0717

On expression of the cDNA encoding the precursor of chicken mitochondrial aspartate aminotransferase (pmAspAT) in Escherichiu coli, the bulk of pmAspAT was found to be associated with the 70-kDa heat-shock protein DnaK which is closely related to mitochondrial 70-kDa heat- shock protein (HSP70). Purification protocols for the DnaK/pmAspAT complex and its individual components were elaborated. The complex dissociated on treatment with MgATP or at pH 5.5. Like the mature enzyme, pmAspAT is a dimer (2 x 47 kDa) and exhibits about a third of its enzyme activity. In the DnaK/pmAspAT complex, one DnaK molecule is bound to each subunit of pmAspAT; this tetramer may further aggregate to an octamer. The complex is catalytically almost as active as free pmAspAT. It could be reconstituted from isolated DnaK and pmAspAT. No complex was formed with mAspAT. Apparently, DnaK binds to the solvent-exposed presequence of folded pmAspAT without significantly changing the structure and functional properties of its mature moiety.

Constitutive heat-shock proteins of 70-kDa (HSP70) and 60-kDa (HSP60) act as chaperones by ensuring the correct folding and oligomeric assembly of other proteins (for a recent review, see Gething and Sambrook, 1992). Several chaperones are involved in the importation of precursor proteins into mitochondria: cytosolic HSP70 are thought to keep the precursors in a translocation-competent form, interaction with mitochondrial HSP70 appears to provide the driving force for their movement through import sites, and mitochondrial HSP60 assists in their intramitochondrial folding and assembly (Neupert et al., 1990; Baker and Schatz, 1991). Direct interaction of mitochondrial HSP70 with precursor proteins passing through contact sites was demonstrated by the iso- lation of chimeric precursors complexed with mitochondrial HSP70 of yeast (Sscl protein; Scherer et al., 1990; Ostermann et al., 1990).

The heat-shock protein DnaK is the single representative of the HSP70 family in E. coli and has been shown to be required for replication of bacteriophage ,I DNA. Like its eukaryotic homologs, it possesses a weak ATPase activity and, in addition, an autophosphorylation activity (Zylicz et al., 1983). Polypeptides bound to DnaK are released with con- comitant hydrolysis of ATP (Liberek et al., 1991). The amino acid sequence of DnaK is 58% identical with that of the Sscl protein of yeast (Craig et al., 1989). P71 protein, another mitochondrial HSP70, which has been purified from HeLa

Correspondence to P. Christen, Biochemisches Institut der Universitat Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland

Abbreviations. HSP70 and HSP60,70-kDa and 60-kDa heat-shock proteins, respectively; pmAspAT, precursor of mitochondrial aspartate aminotransferase; mAspAT, mitochondrial aspartate aminotransferase.

Enzyme. Aspartate aminotransferase (EC 2.6.1 .I).

cells, cross-reacts strongly with antiserum raised against DnaK (Leustek et al., 1989).

On synthesis of the precursor of chicken mitochondrial aspartate aminotransferase (pmAspAT) in E. coli, we found it to be strongly associated with DnaK. In this paper, we report protocols for the purification of the DnaK/pmAspAT complex and its two components and characterize the complex and isolated pmAspAT.

EXPERIMENTAL PROCEDURES

Expression system

The E. coli strain BL21(DE3) pLysS (F-ompT rsms; Grodberg and Dunn, 1988; Studier et al., 1990) was kindly supplied by Dr. W. Studier. The pGEMEX-1 vector was pur- chased from Promega and provided with a unique NdeI restriction site (Ziak et al., unpublished results). The full-length cDNA encoding pmAspAT was isolated from the expression vector pOTS/pmAspAT (Jaussi et al., 1987) by digestion with NdeI and XbuI (from Boehringer). The cDNA fragment was subcloned into the expression plasmid pGEMEX which had been digested with the same restriction endonucleases.

Expression of pmAspAT

Luria-Bertani medium (5 l), containing ampicillin (100 pg/ ml) and chloramphenicol (34 pg/ml), was inoculated with 50 ml of a preculture of E. coli BL21 pLysS transformed with pGEMEX/pmAspAT. The cells were incubated at 37°C with continuous shaking. On reaching an Asno of 0.5-1.0, expression was induced with 0.5 mM isopropyl-thio-a-D- galactoside, and incubation was continued for 3 h.

700

Purification of DnaK/pmAspAT complex

Cells were collected by centrifugation (5500 x g for 30 min at 4°C) and the pellet (23 g wet cells) was suspended in 100 ml 50 mM Trischloride, pH 8.0, containing 1 mM 2-0x0- glutarate, 10 mM EDTA, 1 mM EGTA, 1 mM 1,Cdithio-DL- threitol, 10 pM pyridoxal S'-phosphate (pyridoxal-P) and the proteinase inhibitors 6-amino-n-hexanoic acid (50 mM), benzamidine ( 5 mM), 3,4-dichloroisocoumarin (0.23 mM). Cells were disrupted by sonication (sonifier model 250 from Branson Sonic Power; 4 x 2 min at OOC). Cell debris was removed by centrifugation (100 000 x g, 30 min). These and all subsequent steps were performed at 4°C. The supernatant was diluted with the same volume of distilled water and loaded onto a DEAE column (Fractogel TSK DEAE-650(M), 2.2 cm x 21 cm), equilibrated with 20 mM Trischloride, 1 mM 2-oxoglutarate, 1 mM EDTA, 1 mM EGTA and 1 mM 1,4- dithio-DL-threitol, pH 8.0 (buffer A). After washing with the same buffer, the DnaKIpmAspAT complex was eluted with 400 ml of a linear gradient of 0 - 500 mM NaCl. Fractions containing the complex were identified by measuring AspAT activity and by SDS/PAGE, then pooled, concentrated by ultrafiltration to 30 ml and transferred into 20 mM sodium phosphate, 1 mM EDTA, 1 mM EGTA and 1 mM 1,Cdithio- DL-threitol, pH 7.0 (buffer B) by gel filtration (Sephacryl S- 200 from Pharmacia). Fractions with AspAT activity were collected and loaded onto a cation-exchange column (S Sepharose Fast Flow from Pharmacia; 1.8 cm x 15 cm) equilibrated with buffer B. After washing with the same buffer, the complex was eluted with 200 ml of a linear gradient of 0- 500 mM NaCl. The activity-containing fractions were collected, pooled, concentrated by ultrafiltration, and loaded onto a gel-filtration Superose 12 FPLC column (HR 10/30, Pharmacia) equilibrated with 50 mM sodium phosphate and 100 mM NaCl, pH 7.5. Fractions containing pure complex, i.e. giving only pmAspAT and DnaK bands on SDS/PAGE, were collected. The final yield was 3 - 5% of the AspAT activity in the cell homogenate. After concentration by ultrafiltration, the DnaK/pmAspAT complex was incubated with 0.2 mM pyridoxal-P for 20 min at room temperature in the dark. Excess pyridoxal-P was removed by Sephadex G- 25 chromatography in 50 mM sodium phosphate, pH 7.5. pmAspAT complexed with DnaK prepared in this way was exclusively in its pyridoxal-P form.

Purification of free pmAspAT

Free pmAspAT was purified by the same procedure as the DnaK/pmAspAT complex until anion-exchange chromatography. The pooled gradient fractions with AspAT activity were then transferred into 20 mM sodium acetate, 1 mM EDTA, 1 mM EGTA and 1 mM 1,4-dithio-DL-threitol, pH 5.5 (buffer C), by gel filtration and loaded onto the cation- exchange column. For both types of chromatography, the same gel materials and columns were used as for purification of the DnaK/pmAspAT complex. After washing with buffer C, pmAspAT was eluted with 200 ml of a linear salt gradient of 0 - 500 mM NaC1. The activity-containing fractions at the end of the gradient were collected and dialyzed overnight against buffer A. The dialysate was loaded onto a DEAE column (2.2 cm x 2.5 cm) and the run-through fraction collected. After concentration by ultrafiltration, free pmAspAT was brought completely into its pyridoxal-P form in the same way as that described for the DnaK/pmAspAT complex.

Mature chicken mAspAT was isolated as reported (Gehring et al., 1977). Its specific activity at 25°C was 220 U/ mg.

Purification of DnaK

The run-through fraction from S Sepharose cation-exchange chromatography in buffer C (see purification of free pmAspAT above) was passed over a 5-ml column of ATP- agarose (Welch and Feramisco, 1985; C8 linkage; Sigma) at a flow rate of 7 ml/h. The column was washed with 1 M NaCl in buffer C, and DnaK was eluted with 5 mM ATP, 7 mM MgC12 and 10mM KC1 in buffer C. Fractions containing DnaK were dialyzed overnight against 50 mM Trischloride, 100 mM NaCl and 1 mM EDTA, pH 8.0 (buffer D). The dialyzed solution was concentrated by ultrafiltration and loaded onto a Superose 12 FPLC column (HR 10/30, Pharmacia) equilibrated with buffer D. Fractions were analyzed with SDS/PAGE, and those containing pure DnaK were collected.

Characterization of proteins

The concentrations of purified pmAspAT or mAspAT and of DnaK were determined photometrically using a molar ab- sorption coefficient of the subunit &280 = 70 000 M- l cm-' (Gehring and Christen, 1978) and E~~~ = 14 500 M - l , cm-' (Hellebust et al., 1990), respectively. Enzyme activity was determined in a coupled assay with malate dehydrogenase and 20 mM aspartate plus 20 mM 2-oxoglutarate as substrates (Gehring and Christen, 1978). The polyclonal antiserum against mAspAT was prepared according to the procedure of Sonderegger et al. (1982). The immune complexes were precipitated with Staphylococcus aureus cells. NH2-terminal sequence analysis was performed with a sequencer model 477A or model 470A/900A sequencer from Applied Biosys- tems equipped with an on-line phenylthiohydantoin- derivatized amino acid analyzer model 120A. Polyacrylamide gels were scanned with a laser densitometer (2202 Ultroscan from LKB). Circular dichroism was measured with a spectropolarimeter model J-500 from Jasco.

RESULTS

Detection, purification and characterization of the DnaK/pmAspAT complex

The pGEMEX-1 vector provided high-level expression of pmAspAT (10 mg/l culture); however, degradation of the prepiece of pmAspAT by E. coli proteinases posed a serious problem. It was overcome by using the E. coli strain BL21(DE3) pLysS, which, as a B strain, is deficient in the lon proteinase and in the ompT outer membrane proteinase (Grodberg and Dunn, 1988), and by adding proteinase inhibitors to the lysis buffer. On anion-exchange chromatography, the bulk of AspAT activity of the cell lysate was eluted in the first third of the salt gradient (Table 1); only 5 - 10% of the total activity was found in the run-through fraction and corresponded to free pmAspAT, as identified by immunoprecipitation and SDS/PAGE. Immunoprecipitation of pmAspAT in the gradient-eluted fractions showed a

1 2 3 4 5 6

Fig. 1. SDS/PAGE analysis of steps in the purification of pmAspAT and DnaK. The 10% polyacrylamide gel was stained with Coomassie blue. The AspAT-activity-containing fractions (lanes 1 - 4) analyzed are described in Table 1 and in more detail in Experimental Procedures. Lane 1, DEAE gradient fractions; lane 2, immunopreci- pitate of DEAE gradient fractions obtained with antiserum against mAspAT (the DnaK band is relatively weak in comparison to that of pmAspAT due to its faster destaining), the intermediary thick band corresponds to the heavy chains of the antibodies; lane 3, S Sepharose gradient fractions; lane 4, DEAE run-through fractions containing pure pmAspAT. The steps in the purification of DnaK (lanes 5 and 6) are described in Experimental Procedures. Lane 5, S Sepharose run-through fractions containing DnaK; lane 6, Superose- 12 fractions with pure DnaK.

Dna K- p m As PAT 270 kDa

- 2 2 -1 +1 P~ASPAT A L L Q s R'L L L s A P R.R.A A A T A RIA s s w Found A L L Q S R L L L S A P R R A A A T A R A S S W

1 1

DnaK G K I I G I D L G T T N S C V A I Found G K I I G I D L - T T N S W V - I

Fig. 2. Identification of the isolated components of the DnaK/pmAspAT complex by NH2-terminal amino acid sequence analysis. The reported sequences have been deduced from the cDNA of chicken pmAspAT (Jaussi et al., 1985) and DnaK (Bardwell and Craig, 1984). pmAspAT was isolated by anion-exchange chromatography after dissociation of the complex with MgATP. The NH,-terrninal methionine from recombinant chicken pmAspAT had been removed by the bacterial processing system. DnaK, separated from pmAspAT by SDSjPAGE of the DnaK/pmAspAT complex (250 pmol), was transferred from the gel onto a poly(viny1 difluoride) membrane by electroblotting using a MilliBlot-SDE transfer system (from Millipore; Chaps, pH 11 .O, 20 min, 2.5 mA/cmZ). The band was cut out and directly subjected to Edman degradation. (-) No clear assignment possible; ( W ) cysteine residue which is generally not detectable; (V) potential trypsin-cleavage sites in pmAspAT.

coprecipitating 70-kDa protein (Fig. 1, lane 2) . This protein was identified as the E. coli heat-shock protein DnaK (see below). The isolated DnaK/pmAspAT complex was dissociated into free DnaK and pmAspAT by incubation with 5 mM MgATP for 30 min at 4°C (Liberek et al., 1991). The two components were separated by DEAE anion-exchange chromatography in 20 mM Trischloride, pH 8.0, whereby free pmAspAT ran through and DnaK bound to the column (not shown).

Both pmAspAT and DnaK isolated from the pure complex were identified by NH2-terminal sequence analysis (Fig. 2). The complex itself appeared inhomogeneous on re- chromatography on a Superose-12 gel-filtration column. It eluted in two peaks with molecular masses of approximately 270 kDa and 520 kDa (Fig. 3). Both peaks were analyzed with SDS/PAGE and showed the DnaK and pmAspAT subunit bands in a ratio of approximately 1 : 1, as determined by scan- ning of the silver-stained gel. The same stoichiometry was also found by NH2-terminal sequence analysis of the DnaK/ pmAspAT complex, the ratio of the extrapolated yields of the

70 1

I

0.8

E 0

0.6 c 0

al

C 0 0.4 -? VI n <

0.2

0.0 0 5 10 15

Elution volume (mi)

Fig. 3. Molecular mass of DnaK/pmAspAT complex. A Superose-12 FPLC column was used for gel filtration. SDSjPAGE analysis showed that the minor peaks a and b corresponded to aggregated DnaK and free pmAspAT, respectively. The calibration proteins were ferritin, catalase, aldolase (all from Pharmacia) and mAspAT.

c)nn L"" I I I I I I I I I I I I I I I ]

A

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6

Cycle

Fig. 4. Stoichiometry of the DnaK/pmAspAT complex. The yields of the amino acids obtained in the subsequent cycles of NH2-terminal sequence analysis of the complex were assigned to pmAspAT ( 0 ) and DnaK (A) according to their known sequences (see Fig. 2). The intersects of the two regression lines (top, pmAspAT; bottom, DnaK) with the ordinate indicate the amounts of the two proteins present in the complex. The following amino acids were not taken into account: identical residues at the same position in the two proteins (leucine in cycle 8); glutamine, asparagine, serine, threonine and arginine because of partial degradation and low detection yield.

amino acids corresponding to the two sequences being close to unity (Fig. 4).

Purification of free DnaK and pmAspAT

The DnaK/pmAspAT complex in the DEAE gradient fractions dissociated, not only upon addition of MgATP, but also upon exposure to sodium acetate, pH 5.5 (Hellebust et al., 1990). Its components were isolated by cation-exchange chromatography in the same buffer. Free pmAspAT appeared at the end of the salt gradient in a distinct peak (Fig. 1, lane

702

Table 1. Purification of free pmAspAT from E. coli BL21 pLysS transformed with pGEMEX-l/pmAspAT. 23 g wet cells from 5 1 of culture were used.

Purification step Activity Protein Specific activity

U mg U/mg Cell lysate 3680 900 4 100000 x g 2410 490 5

S Sepharose gradient 540 9.5 57 DEAE gradient 1750 120 14

DEAE run-through 160 2.1 76

240 H

c / /

I I I I I I

0 20 40 60 80 100

Time (min)

Fig. 5. Partial digestion of pmAspAT (0) and mature mAspAT ( A ) with trypsin. The proteins (0.1 mg/ml) were incubated in 50 mM sodium phosphate, pH 7.5, at 37°C. At each time point, AspAT activity was measured and then freshly prepared trypsin solution added (0.2 pg trypsin/lO pg protein).

3; Table 1) and could be easily purified by repeating the anion- exchange chromatography at pH 8.0. Contaminating proteins bound to the column and pmAspAT ran through (Fig. 1, lane 4; Table 1). The predominant proteins in the run-through fraction of the cation-exchange chromatography at pH 5.5 were acidic DnaK and E. coli AspAT (Fig. 1, lane 5 ; molecular mass 44 kDa; Kondo et al., 1987; 350 U of the total 3680 U in the cell lysate of Table 1). DnaK could be isolated by affinity chromatography on ATP-agarose and selective elution with ATP. After gel filtration on a Superose 12 column, it was pure (Fig. 1, lane 6; final yield 1.2 mg from the cell lysate of Table 1).

Characterization of pmAspAT

Pure pmAspAT exhibits a specific activity of 70-80 U/ mg, which is about one third of that of the mature enzyme. The precursor in the purified DnaK/pmAspAT complex is catalytically almost as active as free pmAspAT (approximately 85%). The specific activity of free pmAspAT almost doubled on partial trypsin digestion (Fig. 5). SDSjPAGE showed a time-dependent band shift of pmAspAT to mainly mature enzyme, and at the longest times of incubation a few faint lower-molecular-mass degradation bands (not shown). Two distinct intermediate bands between full-length pmAspAT

l-4

E, m 0 TI v

(D

I 0 v

X

L 0 0 I -

I I I I I 200 210 220 230 240 250

Wavelength (nm)

Fig. 6. Circular-dichroism spectra of pmAspAT (-) and mature mAspAT t e e ) . The protein concentration was 70 pg/ml in 25 mM sodium phosphate, pH 7.5.

D n a K

pmAspAT

1 2 3 Fig. 7. Reconstitution of DnaK/pmAspAT complex. pmAspAT (lane 1) or mAspAT (lane 2) (each 70 pg/ml in 300 p1 50 mM sodium phosphate, pH 7.5) were added to DnaK (100 pg/ml in 200 pl buffer C; see Experimental Procedures). The reconstitution assays were incubated for 3 h at 37°C. After addition of antiserum against chicken mAspAT, the incubation was continued overnight at 4°C. The major bands in lane 1 and 2 correspond to the heavy chains of the antibodies. DnaK alone, treated the same way, showed no cross- reaction with the antiserum (lane 3). The immunoprecipitates were analyzed by 10% SDS/PAGE. The gel was silver-stained.

and the mature enzyme appeared (for potential trypsin-cleavage sites, see Fig. 2). K', of pmAspAT for the substrates L- aspartate (0.6 mM) and 2-oxoglutarate (2.3 mM) proved simi- lar to those of the mature enzyme (X', 0.3 mM and 1.1 mM, respectively; Kirsten et al., 1983). The circular-dichroism spec- trum of pmAspAT reproducibly showed a slightly less pro- nounced negative molar ellipticity in the range 200 -230 nm than that of mAspAT (Fig. 6). Apparently, the mature moiety of pmAspAT adopts somewhat less secondary structure than the mature enzyme.

Reconstitution of the DnaKlpmAspAT complex The DnaK/pmAspAT complex could be reconstituted

from its individual components. The complex formed was detected with immunoprecipitation under non-denaturing conditions and SDSjPAGE of the precipitate. DnaK did not form a complex with mature mitochondria1 AspAT (Fig. 7).

DISCUSSION After high-level expression of the cDNA of pmAspAT in

E. coli, the bulk of pmAspAT was found to exist as a complex

703

with the heat-shock protein DnaK of the host cell. No DnaK/ pmAspAT complex was observed after synthesis of rat pmAspAT in E. coli (Altieri et al., 1989), the discrepancy probably being due to the 400-fold-higher expression rate achieved by the T7 promoter in pGEMEX. The formation of a complex of the acidic DnaK (pZ 5.0-6.5 is a general characteristic of HSP; Lindquist, 1986) with pmAspAT be- came evident by the anomalous chromatographic behavior of pmAspAT. About 90% of the AspAT activity adsorbed to the anion-exchange column, equilibrated at pH 8, at which free pmAspAT with pZ > 9.5 ran through (PI of mAspAT 9.0- 9.5; Sonderegger et al., 1982; pmAspAT possesses four ad- ditional arginine residues). Mature mAspAT, produced with the same expression system, did not form a complex with DnaK (C. Jakob and E. Sandmeier, unpublished observation).

DnaK accounts for 1.4% of the total cellular protein in another E. coli B strain (Herendeen et al., 1979; B66 was later identified as DnaK). In the present study, pmAspAT represented 5% of the total protein in the cell lysate, most of it being complexed with DnaK. Apparently, high-level expression of pmAspAT resulting in increased amounts of unfolded polypeptide segments in the cells brings about enhanced synthesis of DnaK. DnaK has been reported to be involved in the rapid degradation of mutant alkaline phosphatase, probably by binding and transferring it to the proteolytic machinery, in which the ATP-dependent proteinase La (product of the lon gene) is considered to play an important role (Sherman and Goldberg, 1992). The strain used for this study, BL 21, as a lon- strain, may therefore be a prerequisite for obtaining undegraded pmAspAT. An attempt of expression in AR 120 cells (lon'; Mott et al., 1985) failed, as about half of the pmAspAT was already degraded to mature enzyme in the cell homogenate (unpublished results). Even in BL 21 cells, the prepiece of pmAspAT turned out to be very sensitive against proteolysis throughout the purification procedure.

Free pmAspAT chromatographed on gel filtration as homogeneous material corresponding to the molecular mass of a dimer ( 2 x 47 kDa). The mature enzyme is also a dimer; dimerization is essential for enzyme activity because the active site is composed of residues from both subunits (Kirsch et al., 1984). The specific activity of free pmAspAT was found to be about one third of that of the mature enzyme. This difference in activity is significantly larger than that reported for rat liver pmAspAT (Martinez-Carrion et al., 1990). Upon partial digestion of pmAspAT with trypsin, enzyme activity was increased to more than 50% of that of the mature enzyme. Mild trypsin digestion of pmAspAT probably generated a protein with an NH2-terminal alanine residue at position -1 (for potential trypsin cleavage sites in the prepiece, see Fig. 2). The production of quasi-mature enzyme by trypsin digestion has also been observed with the precursor of rat mAspAT (Martinez-Carrion et al., 1990). It is counteracted by further degradation which is also observed in the control with mature enzyme. A main degradation product of 42 kDa appeared that results from cleavage after Arg26 or after Lys31 and possesses 3% of the initial activity (Sandmeier and Christen, 1980). Both the reduced specific activity of pmAspAT and its slightly lower secondary structure content in comparison with mature enzyme (Fig. 6) indicate that the mature moiety of the pmAspAT subunit, although folded in a way which allows both dimerization and enzyme activity, is not completely inert toward interaction with the attached prepiece.

Since the ratio of DnaK/pmAspAT-subunit proved to be 1 : 1 in both the 270-kDa and the 520-kDa DnaK/pmAspAT complex, we assume the 270-kDa material to be composed of

one catalytically active pmAspAT dimer (2 x 47 kDa) and two DnaK molecules (2 x 70 kDa), and the 520-kDa material to represent a dimer of this complex.

Complex formation of DnaK with expressed heterologous products has been already observed with the nuclear oncogene p53 (Clarke et al., 1988), truncated protein A (Hellebust et al., 1990) and a fusion protein composed of part of the cro repressor, truncated protein A and 14 residues of p- galactosidase (Sherman and Goldberg, 1991). The structural basis for complex formation with the oncogene product was not determined, but in the other cases unfolded parts of the expressed proteins were identified as the binding sites for DnaK. In the fusion protein which contains several sections without defined secondary structure, the cro region was found to bind to DnaK.

In the case of pmAspAT, DnaK binds to the prepiece. This conclusion is based on the following evidence. The prepiece of pmAspAT can be assumed to be exposed to the solvent, because the NH2-terminus of mAspAT forms an extra contact with the adjacent subunit, which is not contiguous with the main intersubunit contact area (Ford et al., 1980), pmAspAT is enzymically active, and the prepiece is removable by mild trypsin digestion under non-denaturing conditions. In reconstitution experiments with isolated components, DnaK binds folded, dimeric and enzymically active pmAspAT. The precursor in complex with DnaK remains enzymically active, regard- less of whether the complex has been formed in E. coli or has been reconstituted from its isolated components. DnaK does not bind mAspAT. The failure to reconstitute the complex with native mature enzyme agrees with the observation that mAspAT expressed in E. coli was not found complexed with DnaK. The prepiece may be assumed to possess a random-coil structure, because the synthetic prepeptide in solution does not adopt a defined secondary structure (J.-M. Lin- denmann and P. Christen, unpublished data).

What structural features make the presequence of pmAspAT a target for DnaK? As precursors in translocation interact with mitochondrial HSP70, presequences of imported mitochondrial proteins in general must contain a recognition site for this chaperone and probably also for its prokaryotic homolog DnaK. In view of the endosymbiotic origin of mitochondria, this conclusion implies that the mechanism for importation of mitochondrial proteins employs an evolutionarily pre-existing molecular device. A recent study on the binding specificity of BiP, another member of the HSP70 family, with a series of random oligomeric peptides, showed that the binding site accommodates at least seven amino acid residues, and selects in all positions for apolar aliphatic residues (Flynn et al., 1991). Amino acids of this type are mostly located in the interior of proteins; this situation may explain why HSP with chaperone character recognize only segments of unfolded polypeptide chains. The prepiece of pmAspAT, presumably existing in aqueous environment as permanently unfolded polypeptide segment, exposed on the surface of the protein and containing more than 50% of apolar aliphatic amino acids (Fig. 2), thus appears to meet the recognition criteria for HSP70.

This work was supported in part by Swiss National Science Foun- dation Grant 31-27975.89. R. J. was the recipient of a stipend from the Cloetta-Stiftung, Zurich.

REFERENCES Altieri, F., Mattingly, J. R., Rodriguez-Berrocal, F. J., Youssef, J.,

Iriarte, A,, Wu, T. & Martinez-Carrion, M. (1989) J. Biol. Chem. 264,4182-4186.

704

Baker, K. P. & Schatz, G. (1991) Nature 349, 205-208. Bardwell, J. C. A. & Craig, E. A. (1984) Proc. Nut1 Acad. Sci. USA

Clarke, C. F., Cheng, K., Frey, A. B., Stein, R., Hinds, P. W. & Levine, A. J. (1988) Mol. Cell. Biol. 8, 1206-1215.

Craig, E. A,, Kramer, J., Shilling, J., Werner-Washburne, M., Holmes, S., Kosic-Smithers, J. & Nicolet, C. M. (1989) Mol. Cell. Biol. 9,

Flynn, G. C., Pohl, J., Flocco, M. T. & Rothman, J. E. (1991) Nature

Ford, G. C., Eichele, G. & Jansonius, J. N. (1980) Proc. Nut1 Acad. Sci. USA 77,2559-2563.

Gehring, H., Christen, P., Eichele, G., Glor, M., Jansonius, J. N., Reimer, A.-S., Smit, J. D. G. & Thaller, C. (1977) J . Biol. Chem.

81,848-852.

3000-3008.

353.726 -730.

115,97-101. Gehring, H. & Christen, P. (1978) J . Biol. Chem. 253, 3158-3163. Gething, M.-J. & Sambrook, J. (1992) Nature 355, 33-45. Grodberg, J. & Dunn, J. J. (1988) J . Bacteriol. 170, 1245-1252. Hellebust, H., Uhlen, M. & Enfors, S.-0. (1990) J . Bacteriol. 172,

Herendeen, S. L., van Bogelen, R. A. & Neidhardt, F. C. (1979) J . Bacteriol. 139, 185 - 194.

Jaussi, R., Cotton, B., JuretiC, N., Christen, P. & Schiimperli, D. (1985) J . Biol. Chem. 260, 16 060-16 063.

Jaussi, R., Behra, R., Giannattasio, S., Flura, T. &Christen, P. (1987) J . Biol. Chem. 262,12 434- 12 437.

Kirsch, J. F., Eichele, G., Ford, G. C., Vincent, M. G., Jansonius, J. N., Gehring, H. & Christen, P. (1984) J . Mol. Biol. 174,497- 525.

Kirsten, H., Gehring, H. & Christen, P. (1983) Proc. Natl Acad. Sci.

5030 - 5034.

USA 80, 1807-1810.

Kondo, K., Wakabayashi, S. &Kagamiyama, H. (1987)J. Biol. Chem.

Leustek, T., Dalie, B., Amir-Shapira, D., Brot, N. & Weissbach, H.

Liberek, K., Skowyra, D., Zylicz, M., Johnson, C. & Georgopoulos,

Lindquist, S. (1986) Annu. Rev. Biochem. 55, 1151 -1191. Martinez-Carrion, M., Altieri, F., Iriarte, A. J., Mattingly, J.. Youssef,

Mott, J. E., Grant, R. A,, Ho, Y.-S. & Platt, T. (1985) Proc. Nut1

Neupert, W., Hartl, F.-U., Craig, E. A. &Pfanner, N. (1990) CeN63,

Ostermann, J., Voos, W., Kang, P. J., Craig, E. A,, Neupert, W. &

Sandmeier, E. & Christen, P. (1980) J. Biol. Chem. 255, 10 284-

Scherer, P. E., Krieg, U. C., Hwang, S. T., Vestweber, D. & Schatz,

Sherman, M. Y. & Goldberg, A. L. (1991) J . Bacteriol. 173, 7249-

Sherman, M. Y. & Goldberg, A. L. (1992) EMBO J. 11, 71 -77. Sonderegger, P., Jaussi, R., Christen, P. & Gehring, H. (1982) J . Biol.

Studier, F. W., Rosenberg, A. H. & Dunn, J. J. (1990) Methods

Welch, W. J. & Feramisco, J. R. (1985) Mol. Cell. Biol. 5, 1229-

Zylicz, M., LeBowitz, J. H., McMacken, R. & Georgopoulos, C.

262,8648 - 8659.

(1989) Proc. Natl Acad. Sci. USA 86,7805-7808.

C. (1991) J . Biol. Chem. 266,14 491 - 14 496.

J. & Wu, T. (1990) Ann. N . Y. Acad. Sci. 585, 346-356.

Acad. Sci. USA 82, 88 - 92.

447 - 450.

Pfanner, N. (1990) FEBS Lett. 277, 281 -284.

10 289.

G. (1990) EMBO J. 9,4315-4322.

7256.

Chem. 257,3339-3345.

Enzymol. 185, 60-89.

1237.

(1983) Proc. Natl Acad. Sci. USA 80,6431 -6435.

precursor of mitochondrial aspartate aminotransferase synthesized in escherichia coli is complexed...

Documents