The complete primary structure of rat chaperonin 10 reveals a putative βαβ nucleotide-binding domain with homology to p21ras

Biochimica et Biophysica Acta, 1164(1993) 219-222

219

© 1993Elsevier Science PublishersB.V. All rights reserved 0167-4838/93/$06.00

BBAPRO 30303

Rapid Report

The complete primary structure of rat chaperonin 10 reveals a putative flail nucleotide-binding domain with homology to p 2 1 ras Dallas J. Hartman, Nicholas J. Hoogenraad, Rosemary Condron and Peter Bordier Hcj Department of Biochemistry, La Trobe University, Bundoora, Victoria (Australia)

(Received 14 April 1993)

Keywords: Nucleotidebindingsequence;Chaperonin;Aminoacid sequence;(Rat) The first complete amino-acid sequence of a mitochondrial chaperonin 10 is reported. The amino-terminal alanine residue is acetylated, a modification that may be required for the interaction with heptameric chaperonin 60. Part of the sequence constitutes a potential dinueleotide binding motif and is identical with 7 out of 10 residues in the GTP-binding site of p21ras. This similarity may be the structural basis for the recently discovered complex between p21ras and chaperonin 60 in intact cells (Ikawa, S. and Weinberg, R.A. (1992) Proc. Natl. Acad. Sci. USA 89, 2012-2016).

Although many isolated polypeptides can fold and assemble in vitro in a spontaneous process dictated solely by their primary structure, a large number of nascent polypeptides require the presence of auxiliary proteins called molecular chaperones for successful folding in vivo (reviewed in Ref. 1). Most molecular chaperones are expressed constitutively at a basal level, but their synthesis increases upon treatments that perturb protein structure and hence they belong to a family of proteins termed 'stress proteins'. The chaperonin (cpn) class of molecular chaperones comprise a class of molecules that are homologous in primary structure to the Escherichia coli cpn60 (known as GroEL) and E. coli cpnl0 (known as GroES) proteins (reviewed in Ref. 2). In vitro experiments employing purified E. coli cpn60 and cpnl0 proteins have been successful in reconstituting the enzymatic activity of a number of chemically-denatured polypeptides [3,4]. Heterologous combinations of cpn60 and cpnl0 proteins have also been used to reconstitute the enzymatic activity of ornithine transcarbamylase and ribulose bisphosphate carboxylase/oxygenase [5,6]. Genetic and biochemical evidence indicate that E. coli cpn60 and cpnl0 proteins interact both in vitro and in vivo [7,8]. The tetradecameric E. coli cpn60 forms a complex with cpnl0 homologues from E. coli [7], pea chloroplasts [9] and mammalian mitochondria [10]. In conCorrespondence to: P.B. H~j,Departmentof Biochemistry,La Trobe University,Bundoora,Victoria, 3083, Australia. Abbreviations: cpn60, chaperonin 60; cpnl0, chaperonin 10; EIMS, electrosprayionisationmass spectrometry.

trast the heptameric mammalian cpn60 forms a complex with mammalian cpnl0 but not with E. coli cpnl0 [11]. Structural differences, that are important for an interaction between cpn60 and cpnl0 and thus protein folding, must therefore exist between E. coli cpnl0 and mammalian cpnl0. Here we report the first complete amino-acid sequence of a mitochondrial cpnl0 homologue and provide unequivocal evidence for a post-translational modification not found in the bacterial counterpart. Furthermore, sequence comparisons between rat cpnl0 and p21 ras suggest a structural basis for the formation of the p21ra~/cpn60 complex recently observed in the intact cell [12]. We previously reported an incomplete amino-acid sequence of rat cpnl0, including an unidentified amino-terminal modification [5]. To define this modification, cpnl0 was digested with clostripain and fractionated by reverse-phase HPLC to yield a peptide of 690.5 Da, which was modified at its amino-terminus (Fig. 1A) and had an amino-acid composition of 2 Ala, 1 Gly, 1 Phe, 1 Arg and 1 G l n / G l u (data not shown). This modified peptide was subjected to cleavage by the enzyme acylpeptide hydrolase (EC 3.4.19.1), which is capable of removing an acylamino acid from the amino-terminus of a peptide containing less than 20 to 30 amino acids [13]. The resulting digest was fractionated by reverse-phase HPLC and two peptide peaks were recovered (Fig. 1B). The larger of the two peaks was subjected to automated Edman degradation yielding the sequence Gly-Gln-Ala-Phe-Arg. The primary structure of the modified peptide representing the amino-terminus of cpnl0 is therefore Ac-Ala-Gly-Gln-

220 Ala-Phe-Arg with a calculated molecular mass of 690.8 Da. We were interested to see if E. coli cpnl0, like rat cpnl0, was post-translationally modified. Electrospray ionisation mass spectrometry (EIMS) of purified E. coli cpnl0 gave a molecular mass of 10386.72 + 0.66 Da (data not shown). The molecular mass calculated 0.200

o

A

0.100

<

B G-Q-A-F-R 0.050

2

x._.,

0.025' <

2'5 3'0 % CH3CN Fig. 1. Identification of an amino-terminal acetylalanine residue in cpnl0. Rat cpnl0 was purified from isolated rat liver mitochondria and subjected to chemical and proteolytic digestion followed by HPLC fractionation, essentially as described previously [5,22]. A 690.5-Da modified peptide (peak 1, panel A) was recovered and subjected to digestion with acylpeptide hydrolase (Takara Biomedicals) in 50 mM sodium phosphate (pH 7.2), 1 mM EDTA and 1 mM DTT for 12 h at 37°C. Following digestion, the mixture was refractionated under identical conditions (panel B), the largest peak recovered (peak 2) and subjected to automated Edman degradation yielding the sequence G-Q-A-F-R.

from the published primary structure of E. coli cpnl0 [14] is 10386.94 Da. E. coli cpnl0 is, therefore, not post-translationally modified. N ~ acetylation of proteins is a wide spread phenomenon, being found in animals, plants, viruses, bacteria and fungi, but most often the function of this post-translational modification is not known [15]. /3Endorphin from porcine pituitary glands is found both in an N'~-acetylated form and a non-acetylated form. Only the non-acetylated form has opiate receptor binding activity and analgesic properties. N ~ acetylation in this case most probably modulates the activity of/3-endorphin to maintain normal levels of opiate activity [16]. Our studies indicate rat cpnl0 exists permanently in the N~-acetylated form. This seems to rule out any involvement of reversible N ~ acetylation in the regulation of cpnl0 activity. The significance of the N ~ acetylation of rat cpnl0 is therefore not clear. Recently Viitanen et al. [11] demonstrated that mammalian cpn60 could not utilise E. coli cpnl0 as a co-chaperonin in an in vitro ribulose bisphosphate carboxylase/oxygenase folding assay. Substitution of the bacterial cpnl0, which we show not to be N ~ acetylated, with mammalian cpnl0 resulted in a functional chaperonin folding system. Possibly, the N ~ acetylation of rat cpnl0 is a requirement for a functional interaction with mammalian cpn60. N ~ acetylation of proteins has previously been shown to affect the interaction of a protein with its ligand. The cellular retinol-binding protein type-II exists in approximately equal proportions of an N ~ acetylated and a non-acetylated form. The N ~acetylated form is isolated with 10-fold less endogenously-bound retinol [17]. The intracellular location of cpnl0 N ~ acetylation is not known, however, since most N" acetylation of proteins takes place in the cytosol [18] this is the likely compartment for cpnl0 acetylation as well. This in turn implies that cpnl0, unlike most other matrix proteins in mitochondria, is synthesised without a cleavable presequence and that targeting information must therefore be present in the mature protein (see Ref. 19 for a recent review). The importance, if any, of N ~ acetylation for cpnl0 targeting remains to be investigated. To complete the primary structure determination of cpnl0 we digested the protein with endoproteinase GIu-C. The sequence information gained from selected peptides combined with previous data [5] and data generated by ELMS, allowed all sequencing entities to be arranged in their correct alignment to reveal the complete primary structure of rat cpnl0 (Fig. 2A, B). The amino-acid sequence was in excellent agreement with an amino-acid analysis performed on purified rat cpnl0 (data not shown) and a high degree of positional identity (about 40%) was exhibited between cpnl0 homologues from a variety of sources (Fig. 3).

221

A acAGQAFRKFLPLFDRVLVERSAAETVTKGGIMLPEKSQGKVLQATVVAVGSGGKGKGGEIQPVSVKVGDKVLLPEYGGTKVVLDDKDYFLFRDGDILGKYVD D

!

1

g

2 I

|

3

|

4

|

5 12

P

|

13

16

|

6

|

7

|

9

t

14

|

10

P

11

|

15

D

B

10,813.4 + 0.4 Da (10,812.3 Da)

acAGQAFRKFLPLFDRVLVERSAAETVTKGGIMLPEKSQGKVLQATVVAVGSG GKGKGGEIQPVSVKVGDKVLLPEYGGTKVVLDDKDYFLFRDGDILGKYVD 690.5 De (690.8 De)

1,078.2 Da 0,076.2 Da)

3,092.4 + 0.1 Da (3,091.7 Da)

1,529.7 Da (1,529.7 Da)

3,402.77 + 0.09 Da (3,402.77 Da)

7,379.0.2 + 0.5 Da (7,379.0 Da) Fig. 2. Strategy for the sequencing of cpnlO. (A), Overlapping sequences of peptides derived from digestion of cpnl0, or peptides thereof, with CNBr (1-2), clostripain (3-11), endoproteinase glu-C (12-15) and acylpeptide hydrolase (16). (B), EIMS of selected peptides derived from digestion of cpn10. The observed average molecular masses are compared to those calculated from the sequence and given in brackets. The molecular mass of the fragment comprising residues 1-30 refers to the homoserine lactone obtained following CNBr cleavage.

1

i0

20

30

40

50

Rattus norvegicus Pseudomonas aeruginosa Thermophilic

P3

Legionella micdadei Chlamydia psittaci Chlamydia trachomatis Escherichia coli Mycobacterium tubercolosis Coxiella burnettii Synechococcus 6301 Spinacea oleracea 1 - 1 0 3 spinacea oleracea 1 0 4 - 2 0 2

60

70

80

90

i00

Rattus norvegicus Pseudomonas aeruginosa Thermophilic

P3

Legionella micdadei Chlamydia psittaci Chlamydia trachomatis Escherichia coli Mycobacterium tubercolosis Coxiella burnettii Synechococcus 6301 Spinacea oleracea 1 - 1 0 3 spinacea oleracea 1 0 4 - 2 0 2 F~. 3. Amino-acid sequence sfmi~ri~ ~ e e n cpnlO homologues. The ~mplete rat ~ n l 0 ~quence was ~mpared tot~h~-prim~ st~ctures of cpnl0 homologues from the ~llowing species: Pseudomonas aemginosa [~], Thermophdic strain PS3 [~], Legmnella micdadei [25], Ch~mydia psitmci [26], Ch~mydia trachoma~ [27], Es~edchm coli [14], Myco~cWnum t~ercolos~ [28], Coxiella burnet~ [~], ~nechococc~ strain 6301 [29], Spmachea oleracea (amino acids 1-103) [9] and Spmachea oleracea (amino acids 105-202) [9]. Identical residues and conse~atwe replacements defined as K / R , T / S , D / E , Q / E , D / N and I / L are shaded grey.

222

p21 rcpnl0

(>D D • • • [] [] D D e K~SALTIQLIQNH--~DEY~PT T ~ ~ ~ G G E~QPVSWKVG~P~ YG

Fig. 4. Rat cpnl0 has a potential ~ot¢l nucleotide binding domain. (A), Comparison of sequence of the ~a¢l nucleotide binding domain of LDH and p21TM with a proposed nucleotide-binding domain in cpnl0. The binding motif consists of an invariant sequence of three glycine residues (o) in a GXGXXG configuration followed by an invariant negatively-charged amino acid (e) 20 residues carboxyterminal to the last G. An invariant hydrophilic amino acid (¢) is found 5 residues prior to the first G, whilst relatively conserved hydrophobic or neutral residues ([]) are situated along the length of the motif [20,21]. Identical residues and conservative replacements, defined as D / E and I / L , are shaded grey. LDH, pig lactate dehydrogenase; p21TM,human p21~°'; rcpnl0, rat cpnl0.

Upon examination of the primary structure of cpnl0 a consensus sequence for a /3a/3 nucleotide-binding domain homologous to the one found in lactate dehydrogenase and p2V as [20] (Fig. 4) was identified. The pattern of this motif contains the sequence Gly-X-GlyX-X-Gly followed by a negatively-charged amino acid about 20 residues carboxy-terminal to the last Gly [21]. Invariant hydrophilic amino acids are found 5 residues prior to the first Gly and relatively conserved hydrophobic residues are found along the stretch of the motif. Further examination of the glycine-rich portion of the putative flail nucleotide-binding motif of rat cpnl0 indicates that it is identical in 7 amino acids out of 10 when compared to p21 ~as in this region. Interestingly, Ikawa and Weinberg have detected an interaction between cpn60 and p2V as [12]. That p21TMand rat cpnl0 are highly conserved at the important residues of the/3a/3 motif, suggests a possible structural explanation for the origin of the interaction. The interaction between p21 ~s and cpn60 may be of physiological importance, or alternatively, a consequence of the homology in the flaj8 motif between p21TMand rat cpnl0. This research was supported in part by grants from the Australian Research Council and from the National Health and Medical Research Council of Australia. References 1 Gething, M.-J. and Sambrook, J. (1992) Nature 355, 33-45. 2 Zeilstra-Ryalls, J., Fayet, O. and Georgopoulos, C. (1991) Annu. Rev. Microbiol. 45, 301-325.

3 Martin, J., Langer, T., Boteva, R., Schramel, A., Horwich, A. and Hartl, F.-U. (1991) Nature 353, 36-42. 4 Viitanen, P., Lubben, T.H., Reed, J., Goloubinoff, P., O'Keefe, D.P. and Lorimer, G.H. (1990) Biochemistry 29, 5665-5671. 5 Hartman, D.J., Hoogenraad, N.J., Condron, R. and Hoj, P.B. (1992) Proc. Natl. Acad. Sci. USA 89, 3394-3398. 6 Goloubinoff, P., Christeller, J.T., Gatenby, A.A. and Lorimer, G.H. (1989) Nature 342, 884-889. 7 Chandrasekhar, G.N., Tilly, K., Woolford, C., Hendrix, R. and Georgopoulos, C.P. (1986) J. Biol. Chem. 261, 12414-12419. 8 Tilly, K. and Georgopoulos, C.P. (1982) J. Bacteriol. 149, 10821088. 9 Bertsch, U., Soil, J., Seetharam, R. and Viitanen, P.V. (1992) Proc. Natl. Acad. Sci. USA 89, 8696-8700. 10 Lubben, T.H., Gatenby, A.A., Donaldson, G.K., Lorimer, G.H. and Viitanen, P.A. (1990) Proc. Natl. Acad. Sci. USA 87, 7683768. 11 Viitanen, P.V., Lorimer, G.H., Seetharam, R., Gupta, R.S., Oppenheim, J., Thomas, J.O. and Cowan, N.J. (1992) J. Biol. Chem. 267, 695-698. 12 Ikawa, S. and Weinberg, R.A. (1992) Proc. Natl. Acad. Sci. USA 89, 2012-2016. 13 Krishna, R.G., Chin, C.Q.C. and Wold, F. (1991) Anal. Biochem. 199, 45-50. 14 Hemmingsen, S.M., Woolford, C., Van der Vies, S.M., Tilly, K., Dennis, D.T., Georgopoulos, C.P., Hendrix, R.W. and Ellis, R.J. (1988) Nature 333, 330-334. 15 Tsunusawa, S. and Sakiyama, F. (1984) Methods Enzymol. 106, 165-170. 16 Smyth, D.G., Massey, D.E., Zakarian, S. and Finnie, M.D.A. (1979) Nature 279, 252-254. 17 Schaefer, W.H., Kakkad, B., Crow, J.A., Blair, I.A. and Ong, D.E. (1989) J. Biol. Chem. 264, 4212-4221. 18 Bradshaw, R.A. (1989) Trends Biochem. Sci. 14, 276-279. 19 Schatz, G. (1993) Protein Sci. 2, 141-146. 20 Wierenga, R.K. and Hol, W.G.J. (1983) Nature 302, 842-844. 21 Sternberg, M.J.E. and Taylor, W.R. (1984) FEBS Lett. 175, 387-392. 22 Hoj, P.B., Condron, R., Traeger, J.C., McAuliffe, J.C. and Stone, B.A. (1992) J. Biol. Chem. 267, 25059-25066. 23 Sipos, A., Klocke, M. and Frosch, M. (1991) Infect. Immunol. 59, 3219-3226. 24 Tamada, H., Ohta, T., Hamamoto, T., Otawara-Hamamoto, Y., Yanagi, M., Hiraiwa, H., Hirata, H. and Kagawa, Y. (1991) Biochem. Biophys. Res. Commun. 179, 565-571. 25 Hindersson, P., H¢iby, N. and Bangsborg, J. (1991) FEMS Microbiol. Lett. 77, 31-38. 26 Morrison, R.P., Belland, R.J., Lyng, K. and Caldwell, H. (1989) J. Exp. Med. 170, 1271-1283. 27 Cerrone, M.C., Ma, J.J. and Stephens, R.S. (1991) Infect. Immunol. 58, 79-80. 28 Baird, P.N., Hall, L.M.C. and Coates, A.R.M. (1988) Nucleic Acids Res. 16, 9047. 29 Cozens, A.L. and Walker, J.E. (1987) J. Mol. Biol. 194, 359-383.