ARCHIVES

OF BIOCHEMISTRY

Vol. 298, No. 2, November

AND

BIOPHYSICS

1, pp. 532-537,1992

Import of Mutant Aminotransferase Sergio

Giannattasio,

Ersilia

Forms of Mitochondriai Aspartate into Isolated Mitochondria Marra,’

Rosa Anna

Vacca, Genovina

Iannace,

and Ernest0

Quagliariello*

C.N.R. Centro di Studio sui Mitocondri e Metabolism0 Energetico, Bari and Sezione di Trani, Italy; and *Dipartimento di Biochimica e Biologia Molecolare, Universitci degli Studi di Bari, Bari, Italy

Received

February

14, 1992, and in revised

form

July 3, 1992

To gain some insight into the role played by certain protein domains in the import of mitochondrial aspartate aminotransferase in isolated mitochondria, three protein mutants were constructed by using the plasmid POTSmAspAT, which contains the nucleotide sequence encoding for the mature form of this enzyme. Two mutant proteins in which Cys- 166 was substituted with either serine or alanine and another protein lacking the nine N-terminal amino acids were all syntbesized in a cell-free transcription/translation system. Comparison was made among the newly synthesized mutant proteins and the newly synthesized wild type aspartate aminotransferase with respect to their capability to enter mitochondria. All the mutant proteins proved to be able to enter mitochondria even though with a lower efficiency than the wild type enzyme. Interestingly the thiol reagent mersalyl proved to inhibit import of both wild type enzyme and serine mutant, whereas import of alanine mutant was found to be insensitive to mersalyl, thus showing that Cys-166 is the unique -SH group involved in import. Import of mitochondrial aspartate aminotransferase by mitochondria is shown to involve certain protein domains present in the mature protein, two of them being the Cys166 and the N-terminal regions. o 1992 Academic PRESS, IN.

Like the majority of mitochondrial proteins, mitochondrial aspartate aminotransferase (EC 2.6.1.1) is synthesized on free cytoplasmic ribosomes as a precursor with a higher molecular weight than that of its mature form (1, 2). The precursor is then processed by cleavage of the amino-terminal extrasequence into the mature form upon translocation into mitochondria [for references see (3)].

i To whom correspondence should be addressed Studio sui Mitocondri e Metabolism0 Energetico, A, I-70126 Bari, Italy. Fax: 39-80-243317. 532

at C.N.R. Centro di Via Amendola 165/

Nevertheless, the purified mAspAT is imported into mitochondria in vitro (3, 4), thus strongly suggesting that the presequence of the precursor of mAspAT is not strictly required for the import process. Consistently, both precursor and mature mAspAT, as synthesized in a cell-free coupled transcription/translation system, enter isolated mitochondria with the same import efficiency sharing the same import pathway (5). Thus, targeting information for the correct mitochondrial localization of mAspAT must reside in the mature protein structure. Evidence in favor of such a conclusion has previously been shown by us for mature aspartate aminotransferase. In fact blockage of thiol group/s proved to prevent mAspAT uptake (6), thus suggesting the interaction of a protein domain including cysteine residue/ s in the import. Moreover evidence of a direct interaction of externally added purified mAspAT with the mitochondrial membrane was shown both kinetically and by testing the sensitivity of mAspAT uptake to certain compounds which can bind the mitochondrial membrane without affecting enzyme activity, namely 2-mercaptoethanol and tiron (7, 8). In another investigation trypsin-treated purified mAspAT, which lacks residues 1 to 31, fails to enter mitochondria in vitro (9), thus showing that enzyme uptake involves the N-terminal sequence. Consistently, two N-terminal cyanogen bromide peptides, l-9 and 10-31, proved to inhibit the rate of protein uptake into mitochondria, whereas other peptides distal from the N-terminus have no effect (10). To gain further insight into the mechanism of interaction of mAspAT with the mitochondria and more specifically into the role played by different enzyme domains, mutagenesis experiments appear to be worthwhile as minor changes can be caused in the protein structure without any chemical reaction with amino acid-specific reagents. ’ Abbreviations used: mAspAT, mitochondrial aspartate ferase; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide phoresis.

aminotransgel electro-

0003.9861/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

IN

VITRO

IMPORT

OF

MITOCHONDRIAL

ASPARTATE

AMINOTRANSFERASE

533

ligation

FIG. 1.

Construction of mitochondrial aspartate aminotransferase deletion mutant. The plasmid POTS-mAspAT, containing the nucleotide sequence encoding mature mitochondrial aspartate aminotransferase (filled box), was digested with B&II (B) and SphI (S). Both fragments obtained were purified by agarose gel electrophoresis, with the small fragment then digested with N&I (N) and Hind111 (H), and the Bgl/Nde and Hind/Sph fragments purified by agarose gel electrophoresis. Bgl/Nde and Hind/Sph fragments were ligated to the large Bgl/Sph fragment and a synthetic DNA (open box) with NdeI and Hind111 sticky ends encoding the desired mutation, obtaining the plasmid pOTS-mAspAT(Al9) encoding the N-terminal deleted form of mature mAspAT.

In this work three mutant proteins were obtained, two in which Cys-166 was substituted with either serine or alanine and another in which the nine N-terminal amino acids were deleted. The capability of these mutant proteins to enter mitochondria was tested and their import processeswere compared to that of the wild type protein. In the light of the obtained findings, the proposal is made that both the region containing Cys-166 and that containing the N-terminal are involved in the import process. EXPERIMENTAL

PROCIJDURES

Oligonucleotide-directed mutagenesis of POTS-mAspAT. In order to substitute mAspAT codon for cys-166 with the codon for serine or for alanine, the plasmid POTS-mAspAT was used (11). The 310-bp XmuI/ SphI restriction fragment, which contains Cys-166 codon, was subcloned into the replication form of M13mp18 DNA. Site-directed mutagenesis was carried out using the method described by Kunkel, which provides a very strong biological selection of the clones with nonphenotypically

selectable mutations against the original unaltered genotype (12). Two mutagenic oligonucleotides were synthesized and purchased from Pharmacia LKB Biotechnology: the 23-base oligonucleotide “AA GTC AAG GCA AGA CGT TTT GGGs’, used to change the Cys-166 codon TGT into the serine codon TCT, and the al-base oligonucleotide “TC AAG GCT AGC CGT TTT GGG3’, used to change the cysteine codon TGT into the alanine codon TGC. Both site-directed mutations were constructed by making use of the Muta-Gene Ml3 in uitro Mutagenesis Kit (Bio-Rad) essentially following the producer’s directions. DNA sequencing by the dideoxy method (13) was performed to confirm the identity of the mutations. The new XmoI/SphI DNA fragments obtained were separately used to reconstitute the plasmid POTS-mAspAT, with resulting constructions named POTS-mAspAT(C166S) and POTSmAspAT(C166A), respectively. Construction of POTS-mAspAT deletion mutant. In order to delete the DNA sequence encoding the nine N-terminal amino acids of mAspAT, the plasmid POTS-mAspAT was cleaved with BgnI and SphI restriction enzymes (Fig. 1). The two restriction fragments produced were a 6574-bp fragment, (the first) and a 1296-bp fragment, (the second). Both fragments were separated by agarose gel electrophoresis and purified from the gel. When the 1206-bp fragment was digested with N&I and HindIII restriction enzymes, three fragments were obtained: the

534

GIANNATTASIO

661-bp BglII/N&I fragment, the 64-bp NdeI/HindIII fragment (which contains the nucleotide sequence to be deleted) and the 461-bp HindIII/ SphI fragment. The two larger fragments were separated by agarose gel electrophoresis and purified from the gel. A new NdeI/HindIII DNA fragment was synthesized in which the initiator codon ATG is located immediately before the codon for glycine, the 10th amino acid of the wild type mAspAT. The sequence of the synthetic NdeI/HindIII fragment is shown in Fig. 1. The new DNA fragment is composed of two synthetic complementary oligonucleotides which were synthesized with an Applied Biosystem DNA Synthesizer. After removal of base protection groups, both oligonucleotides were purified by denaturing polyacrylamide gel electrophoresis (7 M urea, 12% polyacrylamide), followed by chromatography on Sephadex G-25 (Pharmacia LKB Biotechnology), after which the two oligonucleotides were mixed in an equimolar ratio, denatured at lOO”C, and allowed to anneal at room temperature. The DNA fragment obtained was phosphorylated by T4 polynucleotide kinase (Pharmacia LKB Biotechnology). The 6574-bp BglII/SphI, the BglII/NdeI, the HindIII/SphI, and the synthetic NdeI/HindIII fragments were then ligated, with the new plasmid used to transform Escherichiu coli MM294/cI+. Bacteria containing mutant plasmids were identified by BglII/HindIII restriction enzyme analysis. The 726-bp BglII/HindIII fragment, containing the initiator codon ATG, was subcloned into M13mp18. DNA sequencing was performed to confirm the identity of the mutation. The new plasmid was named POTS-mAspAT(Al-9). In vitro expression and immunoprecipitation of expression products. The POTS-mAspAT, POTS-mAspAT(ClBBS), POTSmAspAT(ClBBA), and POTS-mAspAT(Al-9) plasmids were amplified in E. coli strain MM294/& and purified as previously described (5). They were then expressed in a prokaryotic cell-free transcription/translation system (Prokaryotic DNA-Directed Translation Kit, Amersham Int.), in the presence of L-[3SS]methionine as a radioactive precursor. Radioactivity incorporation into newly synthesized polypeptides was measured by a trichloroacetic acid precipitation assay (5), with protein synthesis mixture analyzed by SDS-PAGE according to Laemmli (14). Following electrophoresis, the gel was treated for fluorography using the reagent Amplify (Amersham, Int.) following provided instructions. The gel was dried and a fluorograph made at -7O’C using Kodak XOmat AR film, after which the radioactive protein bands were excised from the gel. The gel slices were then rehydrated and incubated at room temperature overnight with 10 ml of a solution of Lipoluma:Lumasolve: water (l&1:0.2 v/v/v) (Lumac), with radioactivity counted in a liquid scintillation counter. Protease sensitivity tests of in uitro-synthesized wild type and mutant proteins were carried out in 10 mM Tris-HCl buffer (pH 7.5) by incubating the sample with pronase for 20 min at 23°C. Immunoprecipitation of transcription/translation mixture with antibody against mAspAT, was carried out according to (5). Import of wild type and mutant proteins into isolated mitochondria Mitochondria were isolated as described in (5), with measurements of mitochondrial import made as follows: the cell-free expression system containing the labeled protein was incubated with mitochondria (600 fig of proteins), in 100 ~1 TKS-medium (150 mM sucrose, 20 mM TrisHCl, pH 7.2,50 mM KCl) containing 5 mM succinate, 1 mM ADP, and 5 ~1 of cell-free expression system (about 5 X lo5 cpm incorporated radioactivity). After 15 min incubation at 37”C, the mixture was centrifuged and the mitochondrial pellet resuspended in TKS-medium and incubated for 20 min at 23°C in the presence of 20 ng/pl pronase, so as to digest externally hound protein. Mitochondria were then separated by centrifugation and the mitochondrial pellet immediately lysed in 2% SDS in which a mixture of protease inhibitors (antipain, pepstatin, leupeptin, and chymostatin) had been added (300 /.tM each) in order to prevent degradation of the imported protein after mitochondrial lysis. SDS-PAGE analysis and fluorography were carried out to monitor imported proteins, with radioactive protein bands excised from the gels, mixed with scintillation fluid, and counted in a liquid scintillation counter as described above. In every case the import level was measured as the

ET

AL.

ratio between the radioactivity associated with imported protein and that associated with the amount of protein incubated with mitochondria, the latter being virtually the same for all proteins analyzed.

RESULTS

To gain some insight into the protein regions of mature mitochondrial aspartate aminotransferase which contain mitochondrial targeting information, either Cys-166 was substituted with serine or alanine, or the nine N-terminal amino acids of the enzyme were deleted. The rationale for this is derived from previous work in which the uptake of mature aspartate aminotransferase was found to be inhibited by enzyme treatment with the thiol reagent mersalyl(6) and by removal of the N-terminal aminoacids (9). Plasmids encoding substitutions of Cys-166 with serine or alanine were constructed by oligonucleotide-directed mutagenesis. The plasmid encoding the deletion of the nine N-terminal amino acids of mature mAspAT was created by substituting the wild type DNA sequence with a synthetic DNA fragment in which the initiator codon ATG is attached in frame to the codon for glycine, the 10th amino acid of the mature protein (Fig. 1). The identity of each mutation was confirmed by DNA sequencing. A schematic representation of the proteins encoded by the wild type and mutant plasmids is shown in Fig. 2. POTS-mAspAT, POTS-mAspAT(C166S), POTSmAspAT(C166A), and POTS-mAspAT(Al-9) were expressed in a cell-free coupled transcription/translation system in the presence of L-[35S]methionine. The radiolabeled expression products were analyzed by SDS-PAGE and fluorography. No change in the electrophoretic mobility was observed for either Ser-166 or Ala-166 containing expression products (Fig. 3, lanes b and c), whereas the POTS-mAspAT(Al-9)-derived polypeptide showed a higher electrophoretic mobility than the mature protein

MUTANT PROTEINS 1 1

401

10

401

1 S

1 FIG. 2. plasmids (C166A),

1

mAspAT

J

mAspATIAl-9)

I

mAspAT

(Cl66S)

]

mAspAT

(Ci66A)

401

401 Schematic representation of the polypeptides POTS-mAspAT, POTS-mAspAT(C166S), and POTS-mAspAT(Al-9).

encoded by the POTS-mAspAT-

IN

VITRO

IMPORT

OF

MITOCHONDRIAL

ASPARTATE

a a

b

c

d

FIG. 3. In uitro expression of wild type and mutant mAspAT. The prokaryotic cell-free coupled transcription/translation system was separately supplied with POTS-mAspAT, POTS-mAspAT(ClBBS), POTSmAspAT(ClBBA), or POTS-mAspAT(Al-9) plasmid DNA, and the expression mixture was incubated at 37°C for 1 h in the presence of L[36S]methionine. The mixture was then chased with cold methionine for 5 min, and the reaction was stopped by rapid cooling. The expression mixture was analyzed by 12% SDS-PAGE and fluorography. a, POTSmAspAT expression product; b, pOTSmAspAT(C166S) expression product; c, POTS-mAspAT(C166A) expression product; d, POTSmAspAT(Al-9) expression product.

as expected owing to the pre.dicted lower molecular weight (44 vs 45 kDa of the mature protein) (Fig. 3, lane d). All three in vitro-synthesized polypeptides were immunoprecipitated by the IgG fraction from antiserum against mAspAT (data not shown). In order to determine whether these mutations could affect protein capability to enter isolated mitochondria, newly synthesized wild type and mutant forms of mAspAT were incubated with isolated mitochondria at 37°C for 15 min, with the uptake measured as described under Experimental Procedures. Figure 4 shows that, like mAspAT, both the cysteine mutants and the deletion mutant are imported into isolated mitochondria (lanes a, c, e, g) where they become protease resistant. To confirm that all three mutant mAspATs had been internalized into mitochondria, mitochondrial membranes were disrupted by 2% SDS lysis of the mitochondrial pellet after in vitro import and subsequently treated with pronase. As expected, all proteins proved to be completely digested (Fig. 4, lanes b, d, f, h). Quantitative measurements of the amount of imported protein were performed by counting the radioactivity of the gel slices. It should be noted that in each experiment both wild type and mutant mAspATs were synthesized with the same specific activity. In the case of the point mutations involving Cys-166, the amount of imported protein was found to be 50% lower than in the case of mAspAT, while in the case of the deletion mutant the amount of imported protein was 40% of that of the wild type (Fig. 4, cfr. lanes a, c, e, g). To ascertain whether the mutant proteins had reached the correct submitochondrial localization, subfractionation of mitochondria was carried out after the in vitro import of the mutant proteins essentially as in (5): all the imported mutant proteins were found in the matrix fraction, with a large amount of protein bound to membranes, which is fairly similar to what was found for the wild type mAspAT (data not shown).

535

AMINOTRANSFERASE

b

c

e

d

f

g

h

FIG. 4. Import into isolated mitochondria of in vitro synthesized wild type and mutant mAspAT. The cell-free expression products of POTSmAspAT, pOTSmAspAT(C166S), POTS-mAspAT(ClBBA), or POTSmAspAT(Al-9) were separately incubated with rat liver mitochondria. After incubation, the mixture was centrifuged and the mitocbondrial pellet treated with pronase to detect imported proteins, as described under Experimental Procedures. In a parallel experiment, after pronase treatment, mitochondria were lysed with 2% SDS and subsequently retreated with pronase (20 ng/pl). In each case proteins were analyzed by SDS-PAGE and fluorography. (a, c, e, g) Mitochondria after treatment with pronase; (b, d, f, h) mitochondria after pronase treatment followed by SDS lysis and pronase treatment; (a, b) POTS-mAspAT expression product; (c, d) POTS-mAspAT(C166S) expression product; (e, f) POTS-mAspAT(C166A) expression product; (g, h) POTSmAspAT(Al-9) expression product.

In another experiment the import of the mutant proteins was investigated as a function of time. Each of the transcription/translation mixtures containing [35S]methionine labeled protein was incubated with isolated mitochondria, with import detected after different incubation times by analysis through SDS-PAGE and fluorography. Figure 5 shows the electrophoresis patterns of the time course of the uptake of mAspAT, mAspAT(C166S), mAspAT(C166A), and mAspAT(Al9). Import of wild type mAspAT was shown to be more efficient than that of deleted and cysteine substituted proteins. Quantitative measurements were carried out by counting the radioactivity of the protein bands. A graphic representation of the data obtained is shown in Fig. 6 where the percentage of import of mutant proteins relative to that of the wild type protein taken up by mitochondria

mA spAT

mAspAT(C166

S)

mAspAT(C166A)

mAspAT(Al-9) 2’

4’

6’

10’

15’

FIG. 5. Time course of the uptake into mitochondria of in oitro-synthesized wild type and mutant forms of mAspAT. In vitro-synthesized mAspAT, mAspAT(ClBBS), mAspAT(C166A), and mAspAT(Al-9) were separately incubated for different times with isolated mitochondria under the conditions previously described to measure import. After incubation the mixture was centrifuged and the mitochondrial pellet analyzed by SDS-PAGE and fluorography.

536

GIANNATTASIO 4 loo-

T--

90 BO7060-

,

0

2

/

,,

4

incubation

,,

6

,

,,(

8

((,

10

time

12

,*

14

(min)

FIG. 6. Quantitative analysis of the time course of wild type and mutant mAspAT import. The time course of the import into mitochondria of either wild type or mutant forms of mAspAT was performed as described in the legend to Fig. 5 and analyzed by SDS-PAGE and fluorography. The amounts of wild type and mutant mAspAT were quantified by radioactivity counting of the protein bands excised from the gel. Results are expressed as the percentage of the import, measured as described under Experimental Procedures, as compared with that of wild type mAspAT after 15 min incubation, to which the value of 100% was given. (0) mAspAT, (*) mAspAT(C166S), (B) mAspAT(ClBGA), (A) mAspAT(Al-9).

is plotted as a function of the incubation time. As shown in (5) the import of the newly synthesized mAspAT is complete in about 15 min, whereas this is not true for the three mutant proteins. A lower efficiency of import was found at all the times investigated, further confirming that the capability of the mutant proteins to enter mitochondria is lower than that of the wild type protein. To gain some insight into the role played by Cys-166 in protein import the capability of the thiol reagent mersalyl to impair the uptake of the newly synthesized wild type and mutant mAspATs was tested (Fig. 7). Mersalylaspartate aminotransferase adduct enters mitochondria with an efficiency 60% lower than the control in the presence of 100 PM mersalyl, with uptake impairment increasing to 70% with 250 pM mersalyl (lanes a-c). Interestingly, externally added mersalyl fails to affect import of the Ala-166 mutant (lanes g-i), whereas, as expected by virtue of the capability of mersalyl to bind with serine (15), a 55% reduction in mitochondrial uptake of the Ser166 mutant was found only at the maximum concentration of added mersalyl (lanes d-f). DISCUSSION Since the amino-terminal extrasequence of the precursor of mitochondrial aspartate aminotransferase was

ET

AL.

found not to be strictly required for import into mitochondria in vitro (5), determinants for mitochondrial localization must lie in the structure of the mature protein. Such a conclusion has here the added dimension given by the investigation of the import carried out by adding the newly synthesized mutant mAspAT to mitochondria. In this way the chemical reaction with any externally added reagents is completely prevented, thus allowing for a clear-cut interpretation of the results. In fact in this work two of the proteins investigated differ from the wild type mAspAT only in that the -SH moiety of residue 166 is replaced by either -H or -OH. Interestingly the newly synthesized Ser-166 and Ala-166 mAspAT mutants both enter mitochondria but with a significantly lower efficiency than that of the wild type. This clearly indicates that -SH group has a significant role in the import process, although this role remains matter of speculation. In fact both binding with the mitochondrial membrane and a function of the thiol in determining the enzyme conformation necessary for the uptake must be taken into consideration. Mersalyl studies show that the thiol group previously suggested to be involved in purified enzyme uptake (6) is located on Cys-166 residue. In fact, externally added mersalyl doesimpair the uptake of the newly synthesized wildtype mAspAT, whereas in proteins in which the Cys-166 is absent the situation is different: no inhibition is found when alanine is present in position 166, whereas significant inhibition is found when investigating serine mutant, as expected since mersalyl can bind with an -OH group even if with a lower efficiency than with -SH (15). Consistently the thiol moiety of Cys-166 is the only thiol group exposed in the spatial structure of the protein (16, 17) and it is conserved in the primary structure of the mAspAT from different species (18). The minor changes in the tertiary structure due to the substitution of cysteine with either serine or alanine could be responsible for the changes in the effectiveness of uptake exhibited by the mutated proteins. As far as the role in protein import of the N-terminal region of mAspAT is concerned, previous data show that when the 31 N-terminal amino acids of the purified enzyme are removed by treatment with trypsin, the protein neither binds to mitochondria nor is imported into the

a

b

c

d

e

f

9

h

i

FIG. 7. Effects of mersalyl on the incorporation of wild type and mutant mAspAT into mitochondria. In vitro-synthesized mAspAT, mAspAT(C166S), and mAspAT(C166A) were separately preincubated in the absence (a, d, g) or in the presence of 100 pM (b, e, h) and 250 gM (c, f, i) mersalyl at 37°C for 2 min, and import was subsequently measured in each case as described under Experimental Procedures. (a, b, c) mAspAT; (d, e, f) mAspAT(C166S); (g, h, i) mAspAT(C166A).

IN

VZTRO

IMPORT

OF

MITOCHONDRIAL

organelles (9). Consistently, CNBr peptide fragments l9 and lo-31 specifically inhibit import of the purified enzyme, respectively in purely competitive and noncompetitive ways (10). The results obtained with the deletion mutant, in which only nine N-terminal amino acids were deleted, show a strong decrease in the import efficiency as compared to that of the wild-type proteins, although the mutant protein still retains some import competence. Though the experimental systems used to study the effect of the 31 and 9 amino acid deletions on protein import are different, the picture that emerges is one in which the N-terminal portion of mature mAspAT interacts with the mitochondrial membrane during transport. The fact that a longer stretch of N-terminal amino acids is needed for import of the purified enzyme than in the case of in vitro-synthesized protein could be the effect of the different protein conformation of mAspAT in the import assay. These results are consistent with the finding that the amino-terminal region of mAspAT was more conserved than the corresponding region of the cytosolic isoform. In addition, the degree of interspecies identity at the N-terminal segment between two mitochondrial enzymes exceeds the average degree of identity of the two polypeptide chains, whereas the degree of interspecies identity between two cytosolic AspATs is lower than the average of the polypeptide chain (18). These findings strongly substantiate our previous proposal that a chemical interaction occurs between different protein domains and the mitochondrial receptor area (7). The conclusions reported in this paper strongly confirm that the mAspAT receptor binding and the following import is dependent on a variety of chemical interactions between protein domains and the mitochondrial membrane. ACKNOWLEDGMENTS We are grateful to Prof. nucleotides for the deletion

ASPARTATE

537

AMINOTRANSFERASE

ulating discussions, to Prof. J. Credico (Lord Byron College, Bari) for linguistic consultation, and to Mr. R. S. Merafina for technical assistance. This work was supported by C.N.R. Target Project on Biotechnology and Bioinstrumentation.

REFERENCES 1. Sonderegger, P., Jaussi, R., and Christen, Res. Commun. 94,1256-1260. 2. Sakakibara, H. (1980)

P. (1980) BiocZzem. Biophys.

N., Huynh, Q. K., Nishida, Y., Watanabe, T., and Wada, Biochem. Biophys. Res. Commun. 95, 1781-1788.

3. Doonan, S., Marra, E., Passarella, S., Saccone, E. (1984) Znt. Reu. Cytol. 91, 141-186.

C., and Quagliariello,

4. Marra, E., Doonan, S., Saccone, C., and Quagliariello, E. (1977) B&hem. J. 174,685-691. 5. Giannattasio, S., Marra, E., Abruzzese, M. F., Greco, M., and Quagliariello, E. (1991) Arch. B&hem. Biophys. 290, 528-534. 6. Marra, E., Passarella, S., Doonan, E. (1979) Arch. Biochem. Bbphys.

S., Saccone,

7. Marra, E., Passarella, S., Casamassima, and Quagliariello, E. (1985) Biochem. 8. Passarella, S., Marra, E. (1985) FEBS L&t. 9. O’Donovan, Quagliariello,

E., Atlante,

C., and Quagliariello,

195,269-279. E., Perlino, E., Doonan, J. 228, 493-503.

A., Doonan,

S., and Quagliariello,

189,235-240.

K. M. G., Doonan, E. (1985) Biochem.

S., Marra, E., Passarella, J. 228, 609-614.

10. Barile, M., Giannattasio, S., Marra, E., Passarella, Sannia, G., and Quagliariello, E. (1990) Biochem. Commun. 170,609-615. 11. Jaussi, R., Behra, R., Giannattasio, (1987) J. Biol. Chem. 262,12434-12437. 12. Kunkel,

T. A. (1985)

Proc. N&l.

U. K. (1970)

Nature

15. Pearson,

R. G. (1968)

16. Gehring, 3163.

H., and Christen,

S., Flura,

Acad.

13. Sanger, F., Nicklen, S., and Coulson, Sci. USA 74,5463-5467. 14. Laemmli,

S.,

S., and

S., Pucci, P., Biophys. Res.

T., and Christen,

Sci. USA A. R. (1977)

P.

82.488-492. Proc. Natl.

Acad.

227,680-685.

J. Chem.

Educ.

P. (1978)

17. Kirsch, J. F., Eichele, G., Ford, J. N., Gehring, H., and Christen,

45,

581.

J. Biol.

Chem.

253,

3158-

G. C., Vincent, M. G., Jansonius, P. (1984) J. Mol. Biol. 174,497-

525. A. Simeone mutagenesis,

for synthesizing the two oligoto Prof. S. Passarella for stim-

18. Graf-Hausner, U., Wilson, Chem. 268,8813-8826.

K. J., and Christen,

P. (1983)

J. Biol.